model creative problem solving adalah

KajianPustaka.com

• metode pembelajaran

Model Pembelajaran Creative Problem Solving (CPS)

Model pembelajaran Creative Problem Solving (CPS) adalah suatu metode pembelajaran yang pemusatannya tertuju pada keterampilan pemecahan masalah melalui teknik sistematik dalam mengorganisasian gagasan-gagasan kreatif. Siswa tidak hanya diajarkan cara menghafal tanpa berpikir, namun dituntut untuk memilih dan mengembangkan suatu tanggapan untuk memperluas proses berpikir.

Creative problem solving merupakan teknik pembelajaran dalam penyelesaian suatu permasalahan berkaitan dengan pemecahan masalah yang melalui teknik sistematik dan mengorganisasikan gagasan kreatif. Melalui model pembelajaran creative problem solving, siswa dapat memilih dan mengembangkan ide dan pemikirannya. Munculnya solusi kreatif sebagai upaya pemecahan masalah akan menumbuhkan kepercayaan diri, keberanian menyampaikan pendapat, berpikir devergen, dan fleksibel dalam upaya pemecahan masalah.

Creative problem solving dibangun atas tiga macam komponen, yaitu; ketekunan, masalah dan tantangan. Komponen tersebut dapat diimplementasikan secara sistematik dengan berbagai komponen pembelajaran. Model pembelajaran creative problem solving berusaha mengembangkan pemikiran divergen, berusaha mencapai berbagai alternatif dalam memecahkan suatu masalah.

Berikut definisi dan pengertian model pembelajaran creative problem solving dari beberapa sumber buku: 

• Menurut Shoimin (2014), creative problem solving adalah model pembelajaran yang pemusatannya pada pengajaran dan keterampilan dalam memecahkan masalah. Ketika dihadapkan dengan suatu pernyataan, siswa dapat melakukan keterampilan dalam memecahkan masalah untuk memilih dan mengembangkan tanggapannya. Tidak hanya dengan cara menghafal tanpa berpikir, keterampilan memecahkan masalah dapat memperluas proses berpikir. 
• Menurut Baharudin (2010), creative problem solving adalah variasi dari pembelajaran dengan pemecahan masalah melalui teknik sistematik dalam mengorganisasikan gagasan kreatif untuk menyelesaikan suatu permasalahan.
• Menurut Cahyono (2009), creative problem solving adalah suatu metode pembelajaran yang melakukan pemusatan pada pengajaran dan ketrampilan memecahkan masalah, yang diikuti dengan penguatan ketrampilan.

Karakteristik Model Pembelajaran Creative Problem Solving 

Menurut Imam (2010), model pembelajaran creative problem solving memiliki tiga karakteristik yang menjadi prosedur dalam proses pembelajarannya, yaitu sebagai berikut: 

• Menemukan fakta, melibatkan penggambaran masalah, mengumpulkan, dan meneliti data dan informasi yang bersangkutan. 
• Menentukan gagasan, berkaitan dengan memunculkan dan memodifikasi gagasan tentang strategi pemecahan masalah. 
• Menemukan solusi, yaitu proses evaluasi sebagai puncak pemecahan masalah.

Menurut Menurut Suryosubroto (2009), karakteristik dari model pembelajaran creative problem solving adalah sebagai berikut: 

• Melatih siswa untuk berpikir divergen dalam memecahkan masalah dengan berbagai cara, mampu memberikan berbagai alternatif pemecahan atas sebuah masalah dan kemampuan mengemukakan berbagai gagasan baru, dengan cara-cara baru yang jarang dilakukan oleh orang lain.
• Peran pendidik lebih banyak menempatkan diri sebagai fasilitator, motivator dan dinamisator belajar bagi peserta didik.

Tujuan Metode Creative Problem Solving 

Menurut Shoimin (2014), melalui model pembelajaran creative problem solving siswa diharapkan mampu:

• Menyatakan urutan langkah-langkah pemecahan masalah dalam creative problem solving. 
• Menemukan kemungkinan-kemungkinan strategi pembelajaran.
• Mengevaluasi dan menyeleksi kemungkinan-kemungkinan tersebut kaitannya dengan kriteria-kriteria yang ada. 
• Memilih suatu pilihan solusi yang optimal. 
• Mengembangkan suatu rencana dalam mengimplementasikan strategi pemecahan masalah. 
• Mengartikulasikan bagaimana creative problem solving dapat digunakan dalam berbagai bidang/situasi.

Langkah-langkah Model Pembelajaran Creative Problem Solving 

Menurut Huda (2013), sintak atau tahapan proses dalam model pembelajaran Creative Problem Solving menurut model Osborn-Parnes dikenal dengan istilah OFPISA, yaitu Objective, Finding, Fact Finding, Idea Finding, Solution Finding, dan Acceptence Finding. Adapun penjelasannya adalah sebagai berikut:

a. Objective Finding 

Siswa dibagi ke dalam kelompok-kelompok. Siswa mendiskusikan situasi permasalahan yang diajukan guru dan membrainstroming sejumlah tujuan atau sasaran yang bisa digunakan untuk kerja kreatif mereka. Sepanjang proses ini siswa diharapkan bisa membuat suatu konsensus tentang sasaran yang hendak dicapai kelompoknya.

b. Fact Finding 

Siswa membrainstroming semua fakta yang mungkin berkaitan dengan sasaran tersebut. Guru mendaftar setiap perspektif yang dihasilkan oleh siswa. Guru memberi waktu kepada siswa untuk berefleksi tentang fakta-fakta apa saja yang menurut mereka paling relevan dengan sasaran dan solusi permasalahan.

c. Problem Finding 

Salah satu aspek terpenting dari kreativitas adalah mendefinisikan kembali perihal permasalahan agar siswa bisa lebih dekat dengan masalah sehingga memungkinkannya untuk menemukan solusi yang lebih jelas. Salah satu teknik yang bisa digunakan adalah membrainstroming beragam cara yang mungkin dilakukan untuk semakin memperjelas sebuah masalah.

d. Idea Finding 

Pada langkah ini, gagasan-gagasan siswa didaftar agar siswa bisa melihat kemungkinan menjadi solusi atas situasi permasalahan. Ini merupakan langkah brainstorming yang sangat penting. Setiap usaha siswa harus diapresiasi sedemikian rupa dengan penulisan setiap gagasan, tidak peduli seberapa relevan gagasan tersebut akan menjadi solusi. Setelah gagasan-gagasan terkumpul, cobalah meluangkan beberapa saat untuk menyortir mana gagasan yang potensial dan yang tidak potensial sebagai solusi. Tekniknya adalah evaluasi cepat atas gagasan-gagasan tersebut untuk menghasilkan hasil sortir gagasan yang sekiranya bisa menjadi pertimbangan solusi lebih lanjut.

e. Solution Finding 

Pada tahap ini, gagasan-gagasan yang memiliki potensi terbesar dievaluasi bersama. Salah satu caranya adalah dengan membrainstroming kriteria-kriteria yang dapat menentukan seperti apa solusi yang terbaik itu seharusnya. Kriteria ini dievaluasi hingga ia menghasilkan penilaian yang final atas gagasan yang pantas menjadi solusi atas situasi permasalahan.

f. Acceptance Finding 

Pada tahap ini, siswa mulai mempertimbangkan isu-isu nyata dengan cara berpikir yang sudah mulai berubah. Siswa diharapkan sudah memiliki cara baru untuk menyelesaikan berbagai masalah secara kreatif. Gagasan-gagasan mereka diharapkan sudah bisa digunakan tidak hanya untuk menyelesaikan masalah, tetapi juga untuk mencapai kesuksesan.

Kelebihan dan Kekurangan Model Pembelajaran Creative Problem Solving 

Setiap model pembelajaran pada umumnya memiliki kelebihan dan kekurangan masing-masing begitu juga dengan model pembelajaran creative problem solving. Menurut Istarani dan Ridwan (2014), kelebihan dan kekurangan creative problem solving adalah sebagai berikut:

a. Kelebihan 

Kelebihan atau keunggulan model pembelajaran creative problem solving yaitu:

• Berpikir dan bertindak kreatif.
• Dapat membuat pendidikan sekolah lebih baik relevan dengan kehidupan, khususnya dunia kerja. 
• Memecahkan masalah yang dihadapi secara realistis. 
• Merangsang perkembangan kemajuan berpikir siswa untuk menyelesaikan masalah yang dihadapi dengan tepat.
• Melatih siswa untuk mendesain suatu penemuan. 
• Mengidentifikasikan dan melakukan penyelidikan.
• Menafsirkan dan mengevaluasi hasil pengamatan.
• Memilih fakta aktual sebagai dasar dan landasan untuk membahas pembelajaran.
• Pembelajaran ini melatih dan menumbuhkan orisinalitas ide, kreativitas kognitif tinggi, kritis, komunikasi-interaksi, sharing keterbukaan, dan sosialisasi.
• Menumbuhkan rasa kebersamaan siswa melalui diskusi akhir dari pemecahan masalah.

b. Kekurangan 

Kekurangan atau kelemahan model pembelajaran creative problem solving yaitu:

• Memerlukan waktu yang lebih banyak dibandingkan dengan metode pembelajaran yang lain. 
• Beberapa pokok bahasan sangat sulit dalam menerapkan sebuah metode pembelajaran ini. Sehingga menyebabkan siswa sulit untuk melihat, mengamati, dan menyimpulkan kejadian atau konsep tersebut. 
• Sulit mencari masalah yang benar-benar aktual dalam pembelajaran.
• Adanya masalah yang tidak relevan dengan materi pembelajaran.
• Menentukan suatu masalah yang tingkat kesulitannya sesuai dengan tingkat berpikir siswa memerlukan kemampuan dan keterampilan guru.
• Mengubah kebiasaan siswa belajar merupakan kesulitan tersendiri bagi siswa untuk menerima informasi dari guru.

Creative Problem Solving (CPS) untuk Meningkatkan Keterampilan Berpikir Kreatif Siswa

World Economic Forum mengatakan bahwa kreativitas menjadi keterampilan terbaik yang dibutuhkan oleh tenaga kerja di abad ke-21 agar berhasil dalam persaingan. Sebagai pendidik, mendorong dan mengembangkan kreativitas siswa melalui model pembelajaran yang dipilih. Sebagai contoh, guru bisa menerapkan model Creative Problem Solving (CPS) untuk meningkatkan keterampilan berpikir kreatif siswa di dalam kelas.

Creative Problem Solving (CPS) adalah model pembelajaran yang berfokus pada keterampilan memecahkan masalah dan tantangan dalam menemukan solusi terbaik dengan cara berpikir kreatif, inovatif, dan imajinatif. Memiliki keterampilan berpikir kreatif sangat penting bagi siswa untuk mempersiapkan diri dalam perubahan yang terjadi begitu cepat di masa mendatang. Dengan kreativitas yang tinggi, bukan hal mustahil bagi siswa untuk menjadi penemu-penemu baru di dunia.

Creative Problem Solving (CPS) untuk meningkatkan keterampilan berpikir kreatif siswa dalam pembelajaran menjadi sangat penting untuk membentuk siswa menjadi pemikir yang mampu berkolaborasi dengan ide-ide yang kompleks. Siswa akan dilatih untuk menumbuhkan intuitif dalam analisis kritis dan imajinasi untuk mengungkap ide atau solusi-solusi yang baru dalam memecahkan masalah.

Karakteristik Model Pembelajaran Creative Problem Solving

Model pembelajaran Creative Problem Solving digunakan untuk meningkatkan kemampuan siswa dalam berpikir secara kreatif karena di dalamnya terdapat proses identifikasi masalah hingga bagaimana cara penyelesaiannya dan penarikan kesimpulan. Model ini berpusat kepada peserta didik. Namun, model ini memerlukan bimbingan guru karena terdapat banyak kegiatan yang harus dilalui.

Dalam pendekatan Creative Problem Solving , aspek komunikasi, interaksi sosial antar-peserta didik dan sikap kooperatif, menjadi dimensi penting yang mendukung implementasinya dalam proses pembelajaran di dalam kelas.

Peran guru dalam pembelajaran Creative Problem Solving ialah memberikan arahan kepada peserta didik dalam memecahkan masalah secara mandiri, kreatif, dan menstimulasi mereka agar mampu berimajinasi. Selanjutnya, guru juga perlu menyediakan materi atau topik pembelajaran yang dapat merangsang pemikiran peserta didik untuk memecahkan masalah dengan berpikir kreatif ketika proses belajar berlangsung.

Model pembelajaran CPS dikembangkan pertama kali oleh Alex Osborn sehingga model ini sering disebut The Osborn-parnes Creative Problem Solving Models .

Terdapat dua asumsi dalam Creative Problem Solving , yaitu sebagai berikut.

1.  Setiap orang kreatif di dalam berbagai bidang.

2.  Keterampilan berpikir kreatif dapat dipelajari dan ditingkatkan.

Tahapan CPS

Berikut adalah tahapan dari CPS , yaitu sebagai berikut:

• Mess-finding, yaitu guru mengidentifikasi masalah yang harus dipecahkan oleh peserta didik.
• Fact-Finding, yaitu peserta didik menemukan fakta yang berhubungan dengan permasalahan untuk mencari informasi esensial terhadap masalah yang sedang diidentifikasi.
• Problem-finding, yaitu peserta didik mengidentifikasi kemungkinan penting yang mendasari permasalahan.
• Idea-finding, yaitu menemukan ide dan gagasan yang mungkin bisa dijadikan solusi dalam memecahkan masalah. Pada tahap ini, guru perlu memberikan apresiasi terhadap gagasan dan ide yang diajukan oleh peserta didik dan menyortir ide yang paling berpotensi dijadikan solusi terbaik.
• Solution-finding, yaitu melalukan evaluasi terhadap ide atau gagasan final yang memiliki potensi terbesar dan paling tepat dijadikan solusi untuk memecahkan permasalahan.
• Acceptance-finding, yaitu mengimplementasikan cara berpikir yang baru dalam memecahkan isu-isu dalam kehidupan sehari-hari secara kreatif.

Implementasi proses model pembelajaran Creative Problem Solving dapat dilakukan dengan cara berikut.

• Memberi apersepsi dengan menyampaikan tujuan pembelajaran dan memotivasi peserta didik akan pentingnya pembelajaran yang akan mereka ikuti.
• Membentuk peserta didik ke dalam kelompok kecil untuk melakukan diskusi
• Membagikan lembar kerja untuk setiap kelompok yang telah berisi masalah beserta arahan pengerjaannya.
• Klarifikasi masalah, yaitu memberikan penjelasan terhadap masalah yang diajukan. Dengan demikian, peserta didik memahami dengan jelas dan mudah untuk merumuskan langkah penyelesaian yang harus diambil.
• Pengungkapan Pendapat, yaitu peserta didik diberikan kebebasan untuk mengungkapkan pendapat, ide dan gagasan mereka secara kreatif dan divergen.
• Evaluasi dan Pemilihan, peserta didik mempertimbangkan secara kritis dan selektif terhadap strategi yang dianggap kurang relevan dan paling potensial untuk dijadikan alternatif terbaik sebagai solusi dalam memecahkan masalah.
• Implementasi, di sini peserta didik bersama kelompoknya menentukan solusi terbaik dalam memecahkan permasalahan dan mempresentasikan gagasan mereka kepada kelompok lain serta guru. Selanjutnya, guru memberikan konfirmasi maupun penegasan dan secara bersama menyimpulkan materi pembelajaran. Sebagai pemantapan materi, guru bisa membagikan kuis untuk dikerjakan peserta didik secara individu.

Kelebihan dan Kelemahan Creative Problem Solving

Kelebihan pendekatan creative problem solving.

• Creative Problem Solving memberi kesempatan peserta didik untuk memahami konsep dan cara untuk menyelesaikan suatu masalah.
• Mendukung partisipasi aktif peserta didik dalam proses pembelajaran.
• Mengembangkan kemampuan berpikir kreatif siswa karena di awal pembelajaran disajikan permasalahan yang harus dicarikan solusi penyelesaiannya.
• Mengembangkan kemampuan peserta didik dalam mengidentifikasi, menganalisa dan memecahkan suatu permasalahan.
• Mendukung peserta didik dalam menerapkan pengetahuan baru yang telah diperoleh, ke dalam situasi/permasalahan baru yang akan dihadapi.

Kelemahan Pendekatan Creative Problem Solving

• Guru memiliki tantangan dalam aplikasi CPS ke dalam pembelajaran karena level pemahaman dan kecerdasan peserta didik tidak sama.
• Adanya kemungkinan peserta didik yang tidak siap dalam menghadapi masalah baru.
• Model ini tidak begitu cocok untuk diaplikasikan pada peserta didik di bangku Taman Kanak-Kanak atau pun di kelas awal Sekolah Dasar.
• Memerlukan alokasi waktu yang tidak sebentar dalam proses pembelajaran untuk mempersiapkan peserta didik dalam mengikuti langkah-langkah CPS sehingga kemungkinan peserta didik merasa bosan dalam menyelesaikan kompleksnya masalah dengan variasi jawaban yang tak kalah kompleks. Selain itu, pemilihan topik diskusi atau permasalahan yang dapat mengembangkan kreatifitas peserta didik juga menjadi tantangan bagi guru dan bukanlah suatu hal yang mudah.

Creative Problem Solving memiliki prinsip dasar dalam penerapannya untuk memecahkan masalah secara kreatif, berikut adalah prinsip tersebut.

1. Assume Nothing

Guru harus mampu mendorong siswa untuk terus berpikir kreatif karena apabila peserta didik menganggap mereka telah menemukan jawaban dari suatu permasalahan maka mereka cenderung tidak akan kreatif dalam memecahkan masalah. Oleh karena itu, asumsi menjadi musuh dari kreativitas dan pemikiran yang orisinal.

2. Problems Are Opportunities

Guru mampu memotivasi peserta didik untuk memiliki pandangan bahwa masalah bukan hanya berupa kesulitan yang harus dihadapi, melainkan sebuah peluang baru yang ditawarkan. Pergeseran pandangan ini bisa mendorong perspektif siswa untuk melihat peluang ketika terjadinya suatu masalah.

3. Suspend Judgment

Guru memberikan kebebasan peserta didik untuk menemukan ide atau gagasan yang baru tanpa terburu-buru memberikan penilaian yang dapat menghambat munculnya kreativitas siswa.

Dengan demikian, penerapan pembelajaran Creative Problem Solving diharapkan dapat mendorong peserta didik untuk mampu mengaplikasikan pemikiran kreatif mereka dalam memecahkan permasalahan yang mungkin akan ditemui di kehidupan sehari-hari atau di masa mendatang.

Berlangganan newsletter kami

Dapatkan postingan terbaru yang dikirim langsung ke kotak masuk kamu.

Now check your inbox and click the link to confirm your subscription.

Please enter a valid email address

Oops! There was an error sending the email, please try later.

Direkomendasikan untuk kamu

Persiapan beasiswa untuk siswa sma sekolah masjid terminal, depok oleh komunitas kejar mimpi jakarta, contoh modul ajar dan cara membuat modul ajar, ini 3 opsi dalam implementasi kurikulum merdeka, ada mandiri belajar.

Tidak ada hasil untuk pencarian kamu, coba sesuatu yang berbeda.

Metode Pembelajaran Creative Problem Solving: Menemukan Solusi dengan Cara Santai

• 1.1 Metode Creative Problem Solving
• 1.2 Cara Menggunakan Creative Problem Solving
• 1.3 Tips Menggunakan Creative Problem Solving
• 1.4 Kelebihan Creative Problem Solving
• 1.5 Kekurangan Creative Problem Solving
• 1.6 Tujuan Creative Problem Solving
• 1.7 Manfaat Creative Problem Solving
• 2.1 Apa perbedaan antara Creative Problem Solving dan Problem Solving konvensional?
• 2.2 Bagaimana mengatasi kendala dalam menggunakan Creative Problem Solving?
• 3 Kesimpulan

Kita sering dihadapkan dengan masalah dalam kehidupan sehari-hari. Entah itu dalam pekerjaan, hubungan pribadi, atau belajar di sekolah. Masalah ini bisa membuat kita stres dan pusing kepala. Namun, tenang saja! Ada sebuah metode pembelajaran yang bisa membantu kita menemukan solusi dengan cara yang santai dan kreatif. Metode ini dikenal dengan sebutan Creative Problem Solving.

Creative Problem Solving (CPS) merupakan pendekatan yang digunakan untuk mengatasi masalah secara inovatif dan kreatif. Metode ini melibatkan langkah-langkah sistematik yang membuka pikiran kita untuk berpikir di luar kotak dan menemukan solusi yang tak terduga.

Langkah pertama dalam metode CPS adalah mengidentifikasi masalah dengan jelas. Kita harus merumuskan masalah secara spesifik agar fokus saat mencari solusi tidak terpecah. Misalnya, jika masalah kita adalah sulitnya mengatur waktu untuk belajar, maka kita bisa merumuskan masalah sebagai “Bagaimana cara mengatur waktu belajar yang efektif?”

Setelah masalah teridentifikasi, langkah berikutnya adalah analisis. Kita perlu menganalisis akar penyebab masalah tersebut. Apa yang membuat kita sulit mengatur waktu belajar? Adakah faktor eksternal seperti gangguan dari lingkungan sekitar, atau faktor internal seperti kurangnya motivasi? Dengan menganalisis penyebab masalah, kita bisa mencari solusi yang lebih efektif.

Setelah itu, kita masuk ke tahap generasi ide. Di sini, kita perlu melibatkan pikiran kreatif kita. Cobalah untuk berpikir di luar kotak dan temukan solusi yang tak terduga. Misalnya, jika kita sulit fokus karena gangguan lingkungan, mungkin solusinya adalah mencari tempat yang tenang untuk belajar, atau menggunakan headphone untuk mengurangi gangguan suara.

Setelah menghasilkan ide-ide, kita perlu melakukan evaluasi. Kita perlu mengevaluasi setiap ide yang muncul berdasarkan kelayakan dan kemungkinan berhasilnya. Buang ide yang terlalu tidak realistis atau tidak sesuai dengan masalah yang ingin kita selesaikan. Pilih ide yang paling menjanjikan atau bisa memberikan solusi yang paling efektif.

Terakhir, kita melangkah ke tahap implementasi. Ide yang terpilih perlu diimplementasikan dalam kehidupan sehari-hari. Buatlah jadwal belajar yang efektif dan sesuai dengan waktu yang kita miliki. Berikan diri kita motivasi tambahan agar tetap konsisten dalam menjalankan jadwal tersebut.

Metode pembelajaran Creative Problem Solving bisa menjadi alat yang efektif untuk mengatasi berbagai masalah dalam kehidupan kita. Dengan pendekatan yang santai dan kreatif, kita bisa menemukan solusi-solusi yang tak terduga. Jadi, tidak perlu stres lagi saat dihadapkan dengan masalah. Coba aplikasikan metode ini dan temukan solusi dengan cara yang menyenangkan!

Apa Itu Creative Problem Solving?

Creative Problem Solving (CPS) adalah sebuah metode yang digunakan untuk mengatasi masalah dengan pendekatan kreatif dan inovatif. Metode ini memiliki tujuan untuk memberikan solusi yang unik dan efektif dalam menghadapi tantangan atau masalah yang kompleks. Dalam CPS, pemikiran kreatif dan proses berpikir yang fleksibel sangat ditekankan.

Metode Creative Problem Solving

Metode Creative Problem Solving terdiri dari beberapa tahapan yang harus dilalui. Berikut adalah penjelasan mengenai setiap tahapan dalam metode ini:

• Identifikasi masalah : Tahap pertama dalam metode CPS adalah mengidentifikasi masalah dengan jelas. Penting untuk memahami masalah secara mendalam dan merumuskannya dengan tepat.
• Pengumpulan informasi : Setelah masalah diidentifikasi, langkah selanjutnya adalah mengumpulkan informasi yang relevan mengenai masalah tersebut. Informasi ini dapat berupa data, fakta, atau insight dari berbagai sumber.
• Analisis masalah : Setelah informasi terkumpul, tahap selanjutnya adalah menganalisis masalah secara menyeluruh. Hal ini melibatkan pemahaman mendalam tentang akar permasalahan dan faktor-faktor yang mempengaruhinya.
• Pencarian alternatif solusi : Setelah masalah dianalisis, tahap berikutnya adalah mencari alternatif solusi yang mungkin. Dalam metode CPS, diharapkan untuk berpikir out-of-the-box dan menghasilkan ide-ide yang kreatif serta inovatif.
• Pemilihan solusi terbaik : Setelah alternatif solusi tercipta, tahap selanjutnya adalah memilih solusi yang paling sesuai dengan masalah yang dihadapi. Pemilihan solusi ini harus didasarkan pada kriteria-kriteria yang relevan dan mempertimbangkan konsekuensi yang mungkin muncul.
• Implementasi solusi : Setelah solusi terpilih, langkah berikutnya adalah mengimplementasikan solusi tersebut. Dalam tahap ini, perencanaan dan aksi yang tepat diperlukan untuk memastikan bahwa solusi dapat dijalankan dengan baik.
• Evaluasi solusi : Setelah solusi diimplementasikan, proses Cognitive Problem Solving tidak berhenti di sini. Evaluasi solusi yang telah dilakukan perlu dilakukan untuk mengevaluasi keberhasilan solusi dan memperbaiki jika diperlukan.

Cara Menggunakan Creative Problem Solving

Untuk menggunakan metode Creative Problem Solving, ada beberapa langkah yang dapat diikuti. Berikut adalah langkah-langkah praktis untuk menggunakan CPS:

• Kenali masalah dengan jelas dan rumuskan dengan tepat.
• Kumpulkan informasi yang relevan mengenai masalah.
• Analisis masalah dengan menyeluruh.
• Berikan ruang untuk pemikiran kreatif dan menghasilkan alternatif solusi yang unik.
• Pilih solusi yang paling sesuai dengan masalah yang dihadapi.
• Implementasikan solusi dengan perencanaan yang matang.
• Evaluasi keberhasilan solusi dan perbaiki jika diperlukan.

Tips Menggunakan Creative Problem Solving

Berikut adalah beberapa tips yang dapat membantu dalam menggunakan metode Creative Problem Solving:

• Berikan waktu dan ruang yang cukup untuk berpikir kreatif.
• Bebaskan diri dari batasan dan buka pikiran untuk kemungkinan yang lebih luas.
• Gunakan teknik brainstorming untuk menghasilkan berbagai ide yang kreatif.
• Berikan perhatian pada detail dan kualitas solusi yang dihasilkan.
• Kolaborasi dengan orang lain untuk mendapatkan perspektif yang berbeda dan ide yang lebih kreatif.

Kelebihan Creative Problem Solving

Metode Creative Problem Solving memiliki beberapa kelebihan yang membuatnya menjadi pilihan yang baik dalam mengatasi masalah. Berikut adalah beberapa kelebihan CPS:

• Mendorong pemikiran kreatif dan inovatif dalam mencari solusi.
• Membantu mengatasi masalah yang kompleks dengan cara yang efektif.
• Memungkinkan untuk menemukan solusi yang unik dan tidak terduga.
• Memperluas cara berpikir dan menciptakan alternatif solusi yang lebih baik.
• Meningkatkan kemampuan berpikir kritis dan analitis.

Kekurangan Creative Problem Solving

Namun, metode Creative Problem Solving juga memiliki beberapa kekurangan yang perlu diperhatikan. Berikut adalah beberapa kekurangan CPS:

• Proses yang memakan waktu dan beberapa tahap yang kompleks.
• Memerlukan keterampilan kreatif yang baik untuk menghasilkan ide-ide yang inovatif.
• Tidak selalu menghasilkan solusi yang praktis atau dapat diimplementasikan dengan mudah.
• Mungkin memerlukan sumber daya tambahan dalam proses implementasi solusi.

Tujuan Creative Problem Solving

Tujuan utama dari metode Creative Problem Solving adalah untuk menghasilkan solusi yang kreatif dan inovatif dalam menghadapi masalah yang kompleks. Metode ini membantu mengatasi batasan berpikir konvensional dan membuka kemungkinan baru.

Manfaat Creative Problem Solving

Penerapan Creative Problem Solving memiliki manfaat yang signifikan dalam berbagai konteks. Berikut adalah beberapa manfaat utama dari metode ini:

• Menghasilkan solusi yang lebih kreatif dan inovatif.
• Meningkatkan kemampuan pemecahan masalah secara efektif.
• Mengembangkan kemampuan berpikir lateral dan melakukan asosiasi yang lebih luas.
• Meningkatkan kolaborasi tim dan kemampuan kerja dalam tim.
• Mendorong pemikiran kritis dan analitis.

Apa perbedaan antara Creative Problem Solving dan Problem Solving konvensional?

Creative Problem Solving dan Problem Solving konvensional memiliki perbedaan utama dalam pendekatan dan hasil yang dihasilkan. Dalam Creative Problem Solving, proses berpikir kreatif dan inovatif sangat ditekankan, sementara dalam Problem Solving konvensional, pendekatan yang lebih terstruktur dan konvensional digunakan. Selain itu, Creative Problem Solving cenderung menghasilkan solusi yang lebih unik dan tidak terduga, sementara Problem Solving konvensional lebih fokus pada solusi yang praktis dan dapat diimplementasikan dengan mudah.

Bagaimana mengatasi kendala dalam menggunakan Creative Problem Solving?

Menggunakan Creative Problem Solving dapat melibatkan beberapa kendala. Namun, ada beberapa cara untuk mengatasi kendala-kendala ini. Pertama, berikan waktu dan ruang yang cukup untuk berpikir kreatif. Kedua, berkolaborasi dengan orang lain untuk mendapatkan perspektif yang berbeda dan ide yang lebih kreatif. Ketiga, gunakan teknik brainstorming untuk menghasilkan berbagai ide. Terakhir, tetap terbuka terhadap kemungkinan alternatif solusi dan jangan takut untuk mencoba pendekatan yang berbeda.

Metode Creative Problem Solving adalah pendekatan yang efektif untuk mengatasi masalah dengan cara yang kreatif dan inovatif. Dengan mengikuti langkah-langkah dalam metode ini, kita dapat menghasilkan solusi yang unik dan efektif dalam menghadapi tantangan atau masalah yang kompleks. Meskipun metode ini memerlukan waktu dan pemikiran yang intensif, manfaat yang didapatkan jauh lebih besar daripada kekurangannya. Oleh karena itu, mari terapkan Creative Problem Solving dalam kehidupan sehari-hari kita dan jadikan solusi kreatif sebagai bagian dari proses pemecahan masalah kita.

Jika Anda ingin meningkatkan kemampuan pemecahan masalah Anda, jangan ragu untuk mencoba metode Creative Problem Solving ini. Dengan latihan dan dedikasi yang tepat, Anda akan menjadi lebih terampil dalam memecahkan masalah yang kompleks dan menghasilkan solusi yang kreatif serta inovatif. Selamat mencoba!

Share this:

Related posts:.

Pandai Mengajar dengan Santai: Strategi Metode Pembelajaran yang Efektif

Metode Pembelajaran di Pesantren Modern: Membawa Pesona Tradisi ke Era Digital

Unik dan Menarik, Bahan Ajar Metode Penelitian Prof Suryana Siap Mengguncang Dunia Pendidikan!

Leave a Reply Cancel reply

You must be logged in to post a comment.

• Kebijakan Privasi
• Terms and Conditions

• Jasa Pembuatan Website
• Manajemen Sumber Daya Manusia
• Perkuliahan

Pengertian Model Pembelajaran Creative Problem Solving

Dari pengertian model pembelajaran Creative Problem Solving (CPS) di atas dapat disimpulkan bahwa model Creative Problem Solving (CPS) adalah model pembelajaran yang menekankan kepada keterampilan berpikir siswa untuk menyelesaikan masalah serta mengembangkan ide- ide yang diperoleh untuk diungkapkan serta tidak menghafal.

Tujuan M odel P embelajarann Creative Problem Solving

• Siswa menjadi terampil menyeleksi informasi yang relevan kemudian menganalisisnya dan akhirnya meneliti kembali hasilnya.
• Kepuasan intelektual akan timbul dari dalam sebagai hadiah intrinsik bagi siswa.
• Potensi intelektual siswa meningkat.
• Siswa belajar bagaimana melakukan penemuan dengan melalui proses melakukan penemuan.

       2. Indikator Model Pembelajaran Creative Problem Solving

Langkah-langkah model pembelajaran creative problem solving.

• Klarifikasi masalah 
• Brainstorming / Pengungkapan pendapat
• Evaluasi dan pemilihan
• Implementasi

Kelebihan dan kelemahan model Creative Problem Solving

• Kelebihan model Creative Problem Solving
• Siswa memiliki keterampilan memecahkan masalah.
• Merangsang pengembangan kemampuan berfikir siswa secara kreatif, rasional, logis, dan menyeluruh.
• Pendidikan di sekolah menjadi lebih relevan dengan kehidupan, khususnya dunia kerja.
• Menimbulkan keberanian pada diri siswa untuk mengemukakan pendapat dan ide-idenya.
• Kelemahan model Creative Problem Solving
• Menentukan suatu masalah yang tingkat kesulitannya sesuai dengan tingkat berpikir siswa itu tidak mudah.
• Mengubah kebiasaan siswa belajar dengan mendengarkan dan menerima informasi dari guru menjadi belajar yang banyak berpikir untuk memecahkan permasalahan secara individu maupun kelompok yang kadang-kadang memerlukan berbagai sumber belajar merupakan tantangan atau bahkan kesulitan bagi siswa.
• Proses pembelajaran memerlukan waktu yang lama.
• Kurang sistematis apabila metode ini diterapkan untuk menyampaikan bahan baru.

Bagikan ini:

Pos terkait.

Assalamualaikum Kebetulan saya sedang meneliti untuk tugas akhir saya mengenai model creative problem solving. Setelah baca postingan ini saya berpikir ini postingan sangat bermanfaat karena informasi yg saya inginkan ada semua disana. Namun saya tidak bisa mengutip dari postingan ini, harus langsung dari sumber yg bapak kutip karena di kampus saya untuk hal pengutipan lumayan ketat. Maka dari itu, jika bapak berkenan untuk menolong saya. Saya ingin sekali meminta daftar pustaka atau memberikan informasi buku apa, jurnal apa ataupun sumber lainnya yg digunakan dalam penelitian ini secara rinci agar saya bisa mencari sumber tersebut yg telagh dikutip oleh bapak. Terimakasih sebelumnya Wassalamualaikum

Tinggalkan Balasan Batalkan balasan

Alamat email Anda tidak akan dipublikasikan. Ruas yang wajib ditandai *

Situs ini menggunakan Akismet untuk mengurangi spam. Pelajari bagaimana data komentar Anda diproses .

Pengertian Model Pembelajaran Radec (Read, Answer, Discuss, Explain,…

Pengertian Model Pembelajaran Tari Bambu – Model pembelajaran…

Pengertian Model Pembelajaran Kooperatif Tipe Group Resume –…

Pengertian Model Environmental Learning – Model environmental learning…

Pengertian Model Pembelajaran Anchored Instruction – Joyce dan…

Pengertian Model Pembelajaran ARIAS (Assurance, Relevance, Interest, Assessment,…

seni belajar untuk hidup

Model Pembelajaran Problem Solving (Penjelasan Lengkap)

Daftar Isi ⇅ show

Pengertian model pembelajaran problem solving.

Model pembelajaran problem solving adalah model yang mengutamakan pemecahan masalah dalam kegiatan belajar untuk memperkuat daya nalar yang digunakan oleh peserta didik agar mendapatkan pemahaman yang lebih mendasar dari materi yang disampaikan. Seperti yang diungkapkan Pepkin (dalam Shoimin, 2017, hlm. 135) bahwa metode problem solving adalah suatu model pembelajaran yang melakukan pemusatan pada pengajaran dan keterampilan pemecahan masalah yang diikuti dengan penguatan keterampilan.

Problem solving dalam pembelajaran memegang peranan yang sangat penting. Mengapa? Karena dengan mengetahui cara menyelesaikan masalahnya, pembelajaran akan merekat jauh lebih dalam dan tidak mudah untuk dilupakan. Dampaknya hampir sama dengan pembelajaran kontekstual, karena pada akhirnya masalah adalah hal sehari-hari yang akan ditemui oleh siswa. Pemecahan masalah merupakan keterampilan penting yang dibutuhkan pada abad-21 .

Sementara itu Purwanto (dalam Chotimah & Fathurrohman, 2018, hlm. 280-281) berpendapat bahwa model problem solving adalah suatu proses dengan menggunakan strategi, cara, atau teknik tertentu untuk menghadapi situasi baru, agar keadaan tersebut dapat dilalui sesuai keinginan yang ditetapkan.

Model ini sering disebut sebagai metode pula karena boleh dibilang merupakan salah satu penerapan problem based learning (PBL) yang sudah memiliki langkah-langkah konkret. Namun di balik itu, metode ini juga cukup dinamis untuk dimodifikasi dan disesuaikan dengan keadaan siswa atau sekolah. Oleh karena sifatnya yang dinamis, terdapat berbagai turunan dari model ini, misalnya model pembelajaran creative problem solving             .

Menurut Murray, Hanlie, et al. (dalam Huda, 2015, hlm. 273) model pembelajaran problem solving merupakan salah satu dasar teoretis dari berbagai strategi pembelajaran yang menjadikan masalah (problem) sebagai isu utamanya. Artinya akan terdapat beberapa tipe atau setting yang dapat dinaunginya.

Model problem solving adalah sebuah metode pembelajaran yang mengharuskan siswa berperan aktif dan mampu berpikir. Karena dalam problem solving siswa diharuskan mampu menganalisis materi mulai dengan mencari data sampai dengan menarik kesimpulan. Dapat disimpulkan bahwa model pembelajaran problem solving adalah model yang memusatkan pembelajaran pada pemecahan masalah sehingga siswa dapat memperkuat daya nalar dengan menyusun cara, strategi, atau teknik baru untuk menyelesaikan suatu permasalahan.

Lalu seperti apa prosedur, sintaks, atau langkah-langkah dari model ini? Berikut adalah penjelasannya.

Sintaks Pembelajaran Problem Solving

Terdapat sintaks atau acuan dasar dari seluruh fase yang harus dilakukan dalam menyelenggarakan model pembelajaran problem solving. Menurut Chotimah & Fathurrohman (2018, hlm. 287-288) sintaks model pembelajaran problem solving terdiri dari 6 tahap sebagai berikut.

• Merumuskan masalah Kemampuan ini diperlukan untuk mengetahui dan merumuskan masalah secara jelas.
• Menelaah masalah Untuk menggunakan model problem solving, menelaah masalah diperlukan agar peserta didik dapat menggunakan pengetahuan untuk memerinci dan menganalisis masalah dari berbagai sudut.
• Merumuskan hipotesis Kemampuan yang diperlukan lainnya adalah berimajinasi dan menghayati ruang lingkup, sebab-akibat, dan alternatif penyelesaian.
• Mengumpulkan dan mengelompokkan data (sebagai bahan pembuktian hipotesis) Tahap ini berfungsi untuk memancing kecakapan mencari dan menyusun data serta menyajikan data dalam bentuk diagram, gambar, atau tabel.
• Pembuktian hipotesis Kecakapan menelaah dan membahas data, kecakapan menghubung-hubungkan dan menghitung, serta keterampilan mengambil keputusan dan kesimpulan.
• Menentukan pilihan penyelesaian Tahap ini akan membuat peserta didik mampu untuk membuat alternatif penyelesaian serta kecakapan menilai pilihan dengan memperhitungkan akibat yang akan terjadi pada setiap pilihan.

Langkah Langkah Model Pembelajaran Problem Solving

Terdapat langkah-langkah konkret yang dapat digunakan untuk menyelenggarakan model pembelajaran problem solving. Langkah-langkah pembelajaran menggunakan model pembelajaran problem solving menurut Sani (2019, hlm. 243) adalah sebagai berikut.

• Pendidik menjelaskan tujuan pembelajaran.
• Guru memberikan permasalahan yang perlu dicari solusinya.
• Pendidik (guru) menjelaskan prosedur pemecahan masalah yang benar.
• Peserta didik mencari literatur yang mendukung untuk menyelesaikan permasalahan yang diberikan guru.
• Siswa atau peserta didik menetapkan beberapa solusi yang dapat diambil untuk menyelesaikan permasalahan.
• Peserta didik melaporkan tugas yang diberikan guru.

Tujuan Model Problem Solving

Dalam metode pembelajaran problem solving, pembelajaran tidak hanya difokuskan dalam upaya mendapatkan pengetahuan sebanyak-banyaknya. Justru bagaimana menggunakan segenap pengetahuan yang didapat tersebut adalah fokusnya. Dengan kata lain, model pembelajaran ini mengutamakan peningkatan keterampilan untuk menggunakan pengetahuan sebagiamana nantinya akan digunakan pada dunia nyata atau kehidupan sehari-hari.

Siswa yang dapat mengerjakan atau dapat memecahkan masalah yang diberikan oleh guru dapat dikatakan telah telah menguasai pelajaran dengan baik. Bersinggungan dengan hal tersebut, menurut Chotimah & Fathurrohman (2018, hlm. 282) tujuan dari pembelajaran problem solving adalah sebagai berikut.

• Peserta didik menjadi terampil menyeleksi informasi yang relevan kemudian menganalisisnya dan akhirnya meneliti kembali hasilnya.
• Kepuasan intelektual akan timbul dari dalam sebagai hasil intrinsik bagi peserta didik.
• Potensi intelektual peserta didik meningkat.
• Peserta didik belajar bagaimana melakukan penemuan dengan melalui proses melakukan penemuan.

Kelebihan dan Kekurangan Pembelajaran Problem Solving

Setiap model pembelajaran pasti mempunyai kelebihan masing-masing. Salah satunya yakni model pembelajaran problem solving yang tentunya mempunyai kelebihan dan kekurangan pula. Di bawah ini akan dipaparkan beberapa kelebihan dan kekurangan dari model ini.

Secara umum salah satu kelebihan dari model pembelajaran problem solving adalah meningkatnya daya kritis siswa dalam pembelajaran. Selain itu, menurut Shoimin (2017, hlm. 137-138) kelebihan dari model pembelajaran problem solving adalah sebagai berikut.

• Membuat peserta didik lebih menghayati pembelajaran berdasarkan kehidupan sehari-hari.
• Melatih dan membiasakan para peserta didik untuk menghadapi dan memecahkan masalah secara terampil.
• Dapat mengembangkan kemampuan berpikir peserta didik secara kreatif.
• Peserta didik sudah mulai dilatih untuk memecahkan masalahnya dari semenjak sekolah (sebelum memasuki kehidupan nyata).
• Melatih siswa untuk mendesain suatu penemuan.
• Membuat peserta didik berpikir dan bertindak kreatif.
• Memecahkan masalah yang dihadapi secara realistis.
• Mengidentifikasi dan melakukan penyelidikan.
• Menafsirkan dan mengevaluasi hasil pengamatan.
• Merangsang perkembangan kemajuan berpikir siswa untuk menyelesaikan masalah yang dihadapi dengan cara yang tepat.
• Dapat membuat pendidikan sekolah lebih relevan dengan kehidupan, khususnya dunia kerja.

Sementara itu, menurut Sanjaya (2016, hlm. 220) keunggulan dari metode problem solving adalah sebagai berikut.

• Merupakan teknik pembelajaran yang cukup bagus agar siswa lebih memahami isi pelajaran.
• Menantang kemampuan siswa serta memberikan kepuasan untuk menemukan pengetahuan baru bagi siswa.
• Dapat meningkatkan aktivitas pembelajaran siswa.
• Membantu siswa bagaimana mentransfer pengetahuan mereka untuk memahami masalah dalam kehidupan nyata.
• Dianggap lebih menyenangkan dan disukai siswa.

Menurut Sanjaya (2016, hlm. 220) kelemahan dari metode problem solving adalah sebagai berikut ini.

• Manakala siswa tidak memiliki minat atau tidak mempunyai kepercayaan bahwa masalah yang dipelajari sulit untuk dipecahkan, maka mereka akan merasa enggan untuk mencoba.
• Keberhasilan strategi pembelajaran melalui PBL membutuhkan cukup waktu untuk persiapan.
• Tanpa pemahaman mengapa mereka berusaha untuk memecahkan masalah yang sedang dipelajari, maka mereka tidak akan belajar apa yang mereka ingin dipelajari.
• Chotimah, C., & Fathurrohman, M. (2018). Paradigma Baru Sistem Pembelajaran dari Teori, Metode, Model, Media, Hingga Evaluasi Pembelajaran. Yogyakarta: Ar-Ruzz Media.
• Huda, Miftahul. (2015). Model-model Pengajaran dan Pembelajaran: Isu-Isu Metodis dan Paradigmatis. Yogyakarta: Pustaka Pelajar.
• Sani, R.A. (2019). Inovasi Pembelajaran. Jakarta: Bumi Aksara.
• Sanjaya, Wina (2016). Strategi Pembelajaran Berorientasi Standar Proses Pendidikan ( Cetakan ke 12). Jakarta: Kencana Prenada Media.
• Shoimin, A. (2017). 68 Model Pembelajaran Inovatif dalam Kurikulum 2013. Yogyakarta: Ar-Ruzz Media.

Artikel Terkait

Join the conversation.

Terima kasih, sangat membantu bagi saya, semakin mengerti model pembelajaran problem solving.

Alamat email Anda tidak akan dipublikasikan. Ruas yang wajib ditandai *

Simpan nama, email, dan situs web saya pada peramban ini untuk komentar saya berikutnya.

Beritahu saya akan tindak lanjut komentar melalui surel.

Beritahu saya akan tulisan baru melalui surel.

Leave a comment

Model Pembelajaran Kooperatif Creative Problem Solving: Cara Seru untuk Mengatasi Masalah

• 1.1 Cara Mengimplementasikan Model Pembelajaran Kooperatif Creative Problem Solving
• 1.2 Tips Mengimplementasikan Model Pembelajaran Kooperatif Creative Problem Solving
• 1.3 Kelebihan Model Pembelajaran Kooperatif Creative Problem Solving
• 1.4 Kekurangan Model Pembelajaran Kooperatif Creative Problem Solving
• 2.1 1. Apakah model pembelajaran kooperatif creative problem solving hanya cocok untuk mata pelajaran tertentu?
• 2.2 2. Apakah semua siswa dapat berpartisipasi aktif dalam model pembelajaran kooperatif creative problem solving?
• 2.3 3. Apakah model pembelajaran kooperatif creative problem solving hanya dapat dilakukan di kelas dengan jumlah siswa yang sedikit?
• 2.4 4. Apakah model pembelajaran kooperatif creative problem solving hanya dapat dilakukan secara offline?
• 2.5 5. Bagaimana cara mengukur keberhasilan dari model pembelajaran kooperatif creative problem solving?
• 3.1 Share this:
• 3.2 Related posts:

Creative Problem Solving (CPS) atau pemecahan masalah secara kreatif adalah kemampuan untuk menghadapi dan menyelesaikan masalah dengan pendekatan yang inovatif dan kreatif. Sementara itu, Model Pembelajaran Kooperatif adalah suatu metode pembelajaran yang melibatkan kerjasama antara siswa, di mana mereka saling bekerja sama untuk mencapai tujuan bersama.

Apa Itu Model Pembelajaran Kooperatif Creative Problem Solving?

Cara mengimplementasikan model pembelajaran kooperatif creative problem solving.

• Bentuk kelompok siswa: Bagi siswa ke dalam kelompok-kelompok kecil yang terdiri dari 3 hingga 5 orang. Pastikan setiap kelompok terdiri dari siswa dengan berbagai kemampuan dan bakat, agar mereka dapat saling melengkapi dalam menyelesaikan masalah.
• Tentukan permasalahan: Berikan permasalahan yang menantang dan mendorong siswa untuk berpikir kreatif. Pastikan permasalahan tersebut relevan dengan materi pembelajaran yang sedang diajarkan.
• Motivasi siswa: Berikan motivasi kepada siswa untuk mengembangkan ide-ide kreatif dan solusi yang inovatif dalam menyelesaikan permasalahan yang diberikan. Beri mereka ruang dan kebebasan untuk berpikir di luar batas-batas konvensional.
• Facilitasi kelompok: Sebagai guru, Anda memiliki peran sebagai fasilitator dalam kelompok. Bantu siswa dalam merumuskan ide-ide, memecahkan masalah, dan memilih strategi yang tepat. Berikan bimbingan dan dukungan yang diperlukan agar kelompok dapat bekerja secara efektif dan produktif.
• Presentation dan evaluasi: Setelah kelompok selesai menyelesaikan permasalahan, minta mereka untuk mempresentasikan solusi yang mereka temukan. Berikan umpan balik yang konstruktif dan evaluasi terhadap kinerja siswa dalam menerapkan metode kooperatif creative problem solving.

Tips Mengimplementasikan Model Pembelajaran Kooperatif Creative Problem Solving

• Kenali kebutuhan dan karakteristik siswa: Setiap siswa memiliki kemampuan dan karakteristik yang berbeda. Kenali mereka dengan baik sehingga Anda dapat membantu mereka dalam mengembangkan potensi kreatif dan berpikir kritis.
• Evaluasi metode pembelajaran secara berkala: Selalu lakukan evaluasi terhadap metode pembelajaran yang Anda gunakan. Jika diperlukan, modifikasi dan sesuaikan metode tersebut agar lebih efektif dalam meningkatkan kemampuan kreatif dan berpikir kritis siswa.
• Fasilitasi komunikasi dan kolaborasi: Dalam model kooperatif, komunikasi dan kolaborasi antar siswa sangat penting. Berikan ruang yang aman dan nyaman bagi mereka untuk berbagi ide, berdiskusi, dan bekerja sama dalam menyelesaikan masalah.
• Libatkan teknologi: Manfaatkan teknologi seperti komputer, internet, dan perangkat lainnya dalam mendukung proses pembelajaran. Gunakan aplikasi atau perangkat lunak yang dapat membantu siswa dalam menciptakan solusi kreatif dan inovatif.
• Beri apresiasi kepada siswa: Selalu berikan apresiasi dan penghargaan terhadap usaha dan prestasi siswa dalam mengembangkan kemampuan kreatif dan berpikir kritis. Hal ini akan memotivasi mereka untuk terus belajar dan berkembang.

Kelebihan Model Pembelajaran Kooperatif Creative Problem Solving

• Meningkatkan kemampuan berpikir kritis: Model ini mendorong siswa untuk berpikir secara kritis dalam menyelesaikan masalah. Mereka harus menggali informasi, menganalisis, dan mengambil keputusan berdasarkan pemikiran yang rasional.
• Mengembangkan kemampuan berpikir kreatif: Dalam model ini, siswa dituntut untuk berpikir kreatif dalam mencari solusi yang inovatif. Mereka diajarkan untuk berpikir di luar batas-batas konvensional dan menghasilkan ide-ide yang baru dan segar.
• Mendorong kolaborasi: Dalam model kooperatif, siswa belajar untuk bekerja sama dalam kelompok. Mereka belajar menghargai pendapat dan ide-ide dari anggota kelompok lainnya, serta belajar bekerja dalam tim untuk mencapai tujuan bersama.
• Merangsang motivasi belajar: Model ini memotivasi siswa untuk belajar secara aktif dan mandiri. Mereka merasa memiliki peran penting dalam proses pembelajaran dan memiliki tanggung jawab terhadap kelompoknya.
• Meningkatkan kepercayaan diri: Dalam model ini, siswa diajarkan untuk mengembangkan ide-ide dan solusi yang mereka temukan. Hal ini dapat meningkatkan kepercayaan diri mereka serta mengurangi rasa takut untuk berpikir di luar kotak.

Kekurangan Model Pembelajaran Kooperatif Creative Problem Solving

• Memerlukan waktu yang lebih lama: Proses pemecahan masalah kreatif membutuhkan waktu yang lebih lama daripada pendekatan pembelajaran konvensional. Siswa perlu melibatkan diri secara aktif dalam mengembangkan ide-ide dan mencari solusi yang inovatif.
• Menghadapi peran dominan dalam kelompok: Dalam kelompok, ada kemungkinan siswa dengan kemampuan dominan akan mengambil alih peran dan mengabaikan pendapat anggota kelompok lainnya. Hal ini dapat mengurangi kontribusi dan keaktifan siswa lainnya dalam kelompok.
• Membutuhkan pemahaman yang baik dalam materi pembelajaran: Untuk dapat memecahkan masalah dengan kreatif, siswa perlu memiliki pemahaman yang baik terhadap materi pembelajaran. Jika pemahaman mereka terbatas, mereka mungkin kesulitan dalam mengembangkan ide-ide dan solusi yang relevan.
• Memerlukan fasilitator yang kompeten: Model kooperatif membutuhkan fasilitator yang kompeten dalam mengelola kelompok dan memfasilitasi proses pembelajaran. Guru perlu memiliki pengetahuan dan keterampilan yang memadai untuk dapat menciptakan lingkungan belajar yang efektif dan kondusif bagi siswa.
• Memerlukan dukungan dan kerjasama dari sekolah: Implementasi model ini membutuhkan dukungan dan kerjasama dari pihak sekolah, termasuk dukungan dalam hal sumber daya, penciptaan lingkungan belajar yang kondusif, dan pelatihan yang berkaitan dengan model ini.

FAQ tentang Model Pembelajaran Kooperatif Creative Problem Solving

1. apakah model pembelajaran kooperatif creative problem solving hanya cocok untuk mata pelajaran tertentu, 2. apakah semua siswa dapat berpartisipasi aktif dalam model pembelajaran kooperatif creative problem solving, 3. apakah model pembelajaran kooperatif creative problem solving hanya dapat dilakukan di kelas dengan jumlah siswa yang sedikit, 4. apakah model pembelajaran kooperatif creative problem solving hanya dapat dilakukan secara offline, 5. bagaimana cara mengukur keberhasilan dari model pembelajaran kooperatif creative problem solving, share this:, related posts:.

Metode Pembelajaran CTL: Belajar Sambil Santai Menyenangkan!

Metode Pembelajaran Tipe STAD: Seru-Seruan Belajar Bareng!

Metode Konstruktivisme adalah Pendekatan Belajar yang Melibatkan Siswa dalam Proses Konstruksi Pengetahuan

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

• Privacy Policy
• Syarat dan Ketentuan

• Board of Directors

What is CPS?

Cps = c reative p roblem s olving, cps is a proven method for approaching a problem or a challenge in an imaginative and innovative way. it helps you redefine the problems and opportunities you face, come up with new, innovative responses and solutions, and then take action..

Why does CPS work?

CPS begins with two assumptions:

• Everyone is creative in some way.
• Creative skills can be learned and enhanced.

Osborn noted there are two distinct kinds of thinking that are essential to being creative:

Divergent thinking.

Brainstorming is often misunderstood as the entire Creative Problem Solving process.   Brainstorming is the divergent thinking phase of the CPS process.   It is not simply a group of people in a meeting coming up with ideas in a disorganized fashion. Brainstorming at its core is generating lots of ideas.  Divergence allows us to state and move beyond obvious ideas to breakthrough ideas. (Fun Fact: Alex Osborn, founder of CEF, coined the term “brainstorm.” Osborn was the “O” from the ad agency BBDO.)

Convergent Thinking

Convergent thinking applies criteria to brainstormed ideas so that those ideas can become actionable innovations.  Divergence provides the raw material that pushes beyond every day thinking, and convergence tools help us screen, select, evaluate, and refine ideas, while retaining novelty and newness.

To drive a car, you need both the gas and the brake.

But you cannot use the gas and brake pedals at the same time — you use them alternately to make the car go. Think of the gas pedal as Divergence , and the brake pedal as Convergence . Used together you move forward to a new destination.

Each of us use divergent and convergent thinking daily, intuitively. CPS is a deliberate process that allows you to harness your natural creative ability and apply it purposefully to problems, challenges, and opportunities.

The CPS Process

Based on the osborn-parnes process, the cps model uses plain language and recent research., the basic structure is comprised of four stages with a total of six explicit process steps. , each step uses divergent and convergent thinking..

Learner’s Model based on work of G.J. Puccio, M. Mance, M.C. Murdock, B. Miller, J. Vehar, R. Firestien, S. Thurber, & D. Nielsen (2011)

Explore the Vision.   Identify the goal, wish, or challenge.

Gather Data.   Describe and generate data to enable a clear understanding of the challenge.

Formulate Challenges. Sharpen awareness of the challenge and create challenge questions that invite solutions.

Explore Ideas. Generate ideas that answer the challenge questions.

Formulate Solutions. To move from ideas to solutions. Evaluate, strengthen, and select solutions for best “fit.”

Formulate a Plan.  Explore acceptance and identify resources and actions that will support implementation of the selected solution(s).

Explore Ideas. Generate ideas that answer the challenge question

Core Principles of Creative Problem Solving

• Everyone is creative.
• Divergent and Convergent Thinking Must be Balanced.  Keys to creativity are learning ways to identify and balance expanding and contracting thinking (done separately), and knowing  when  to practice them.
• Ask Problems as Questions.  Solutions are more readily invited and developed when  challenges and problems are restated as open-ended questions  with multiple possibilities. Such questions generate lots of rich information, while closed-ended questions tend to elicit confirmation or denial. Statements tend to generate limited or no response at all.
• Defer or Suspend Judgment.  As Osborn learned in his early work on brainstorming, the  instantaneous judgment in response to an idea shuts down idea generation . There is an appropriate and necessary time to apply judgement when converging.
• Focus on “Yes, and” rather than “No, but.”  When generating information and ideas, language matters.  “Yes, and…” allows continuation and expansion , which is necessary in certain stages of CPS. The use of the word “but” – preceded by “yes” or “no” – closes down conversation, negating everything that has come before it.

• Personality Psychology
• Creativeness

Kemampuan Berpikir Kreatif Ditinjau dari Self Regulated Learning dengan Pendekatan Open-Ended Pada Model Pembelajaran Creative Problem Solving

• QALAMUNA Jurnal Pendidikan Sosial dan Agama 13(1):11-22
• 13(1):11-22
• CC BY-SA 4.0
• This person is not on ResearchGate, or hasn't claimed this research yet.

• Universitas Negeri Semarang

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations
• Arina Ulil Faroh
• Mohammad Asikin
• Sugiman Sugiman
• Fitriani Baharuddin
• Tri Hastiti Fiskawarni
• Rifa Anjiana
• Yanti Sofi Makiyah
• Ernita Susanti
• Rizka Mauludin
• Fitrianto Eko Subekti
• Alfian Erprabowo
• Caswita Caswita
• Dwi Fajarwati
• Mimin Ninawati
• Akbar Handoko
• Bambang Sri Anggoro
• Shely Rozalia Intan Rozalia Intan

• Intan Fayesa
• Ratna Ayu garawati

• Desi Asmarita
• Yulianto Wasiran
• Andinasari Andinasari
• Ferawati Ferawati
• Huri Suhendri

• Kadek Lisa Septiari

• Nurul Khairiatin Nida

• Lilis Setianingsih

• Think Skills Creativ

• Tim Mainhard

• Eva Fitria Ningsih

• Yu-Shan Chang
• Chia-hui Chou
• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up

• Business Essentials
• Leadership & Management
• Credential of Leadership, Impact, and Management in Business (CLIMB)
• Entrepreneurship & Innovation
• Digital Transformation
• Finance & Accounting
• Business in Society
• For Organizations
• Support Portal
• Media Coverage
• Founding Donors
• Leadership Team

• Harvard Business School →
• HBS Online →
• Business Insights →

Business Insights

Harvard Business School Online's Business Insights Blog provides the career insights you need to achieve your goals and gain confidence in your business skills.

• Career Development
• Communication
• Decision-Making
• Earning Your MBA
• Negotiation
• News & Events
• Productivity
• Staff Spotlight
• Student Profiles
• Work-Life Balance
• AI Essentials for Business
• Alternative Investments
• Business Analytics
• Business Strategy
• Business and Climate Change
• Creating Brand Value
• Design Thinking and Innovation
• Digital Marketing Strategy
• Disruptive Strategy
• Economics for Managers
• Entrepreneurship Essentials
• Financial Accounting
• Global Business
• Launching Tech Ventures
• Leadership Principles
• Leadership, Ethics, and Corporate Accountability
• Leading Change and Organizational Renewal
• Leading with Finance
• Management Essentials
• Negotiation Mastery
• Organizational Leadership
• Power and Influence for Positive Impact
• Strategy Execution
• Sustainable Business Strategy
• Sustainable Investing
• Winning with Digital Platforms

What Is Creative Problem-Solving & Why Is It Important?

• 01 Feb 2022

One of the biggest hindrances to innovation is complacency—it can be more comfortable to do what you know than venture into the unknown. Business leaders can overcome this barrier by mobilizing creative team members and providing space to innovate.

There are several tools you can use to encourage creativity in the workplace. Creative problem-solving is one of them, which facilitates the development of innovative solutions to difficult problems.

Here’s an overview of creative problem-solving and why it’s important in business.

Access your free e-book today.

What Is Creative Problem-Solving?

Research is necessary when solving a problem. But there are situations where a problem’s specific cause is difficult to pinpoint. This can occur when there’s not enough time to narrow down the problem’s source or there are differing opinions about its root cause.

In such cases, you can use creative problem-solving , which allows you to explore potential solutions regardless of whether a problem has been defined.

Creative problem-solving is less structured than other innovation processes and encourages exploring open-ended solutions. It also focuses on developing new perspectives and fostering creativity in the workplace . Its benefits include:

• Finding creative solutions to complex problems : User research can insufficiently illustrate a situation’s complexity. While other innovation processes rely on this information, creative problem-solving can yield solutions without it.
• Adapting to change : Business is constantly changing, and business leaders need to adapt. Creative problem-solving helps overcome unforeseen challenges and find solutions to unconventional problems.
• Fueling innovation and growth : In addition to solutions, creative problem-solving can spark innovative ideas that drive company growth. These ideas can lead to new product lines, services, or a modified operations structure that improves efficiency.

Creative problem-solving is traditionally based on the following key principles :

1. Balance Divergent and Convergent Thinking

Creative problem-solving uses two primary tools to find solutions: divergence and convergence. Divergence generates ideas in response to a problem, while convergence narrows them down to a shortlist. It balances these two practices and turns ideas into concrete solutions.

2. Reframe Problems as Questions

By framing problems as questions, you shift from focusing on obstacles to solutions. This provides the freedom to brainstorm potential ideas.

3. Defer Judgment of Ideas

When brainstorming, it can be natural to reject or accept ideas right away. Yet, immediate judgments interfere with the idea generation process. Even ideas that seem implausible can turn into outstanding innovations upon further exploration and development.

4. Focus on "Yes, And" Instead of "No, But"

Using negative words like "no" discourages creative thinking. Instead, use positive language to build and maintain an environment that fosters the development of creative and innovative ideas.

Creative Problem-Solving and Design Thinking

Whereas creative problem-solving facilitates developing innovative ideas through a less structured workflow, design thinking takes a far more organized approach.

Design thinking is a human-centered, solutions-based process that fosters the ideation and development of solutions. In the online course Design Thinking and Innovation , Harvard Business School Dean Srikant Datar leverages a four-phase framework to explain design thinking.

The four stages are:

• Clarify: The clarification stage allows you to empathize with the user and identify problems. Observations and insights are informed by thorough research. Findings are then reframed as problem statements or questions.
• Ideate: Ideation is the process of coming up with innovative ideas. The divergence of ideas involved with creative problem-solving is a major focus.
• Develop: In the development stage, ideas evolve into experiments and tests. Ideas converge and are explored through prototyping and open critique.
• Implement: Implementation involves continuing to test and experiment to refine the solution and encourage its adoption.

Creative problem-solving primarily operates in the ideate phase of design thinking but can be applied to others. This is because design thinking is an iterative process that moves between the stages as ideas are generated and pursued. This is normal and encouraged, as innovation requires exploring multiple ideas.

Creative Problem-Solving Tools

While there are many useful tools in the creative problem-solving process, here are three you should know:

Creating a Problem Story

One way to innovate is by creating a story about a problem to understand how it affects users and what solutions best fit their needs. Here are the steps you need to take to use this tool properly.

1. Identify a UDP

Create a problem story to identify the undesired phenomena (UDP). For example, consider a company that produces printers that overheat. In this case, the UDP is "our printers overheat."

2. Move Forward in Time

To move forward in time, ask: “Why is this a problem?” For example, minor damage could be one result of the machines overheating. In more extreme cases, printers may catch fire. Don't be afraid to create multiple problem stories if you think of more than one UDP.

3. Move Backward in Time

To move backward in time, ask: “What caused this UDP?” If you can't identify the root problem, think about what typically causes the UDP to occur. For the overheating printers, overuse could be a cause.

Following the three-step framework above helps illustrate a clear problem story:

• The printer is overused.
• The printer overheats.
• The printer breaks down.

You can extend the problem story in either direction if you think of additional cause-and-effect relationships.

4. Break the Chains

By this point, you’ll have multiple UDP storylines. Take two that are similar and focus on breaking the chains connecting them. This can be accomplished through inversion or neutralization.

• Inversion: Inversion changes the relationship between two UDPs so the cause is the same but the effect is the opposite. For example, if the UDP is "the more X happens, the more likely Y is to happen," inversion changes the equation to "the more X happens, the less likely Y is to happen." Using the printer example, inversion would consider: "What if the more a printer is used, the less likely it’s going to overheat?" Innovation requires an open mind. Just because a solution initially seems unlikely doesn't mean it can't be pursued further or spark additional ideas.
• Neutralization: Neutralization completely eliminates the cause-and-effect relationship between X and Y. This changes the above equation to "the more or less X happens has no effect on Y." In the case of the printers, neutralization would rephrase the relationship to "the more or less a printer is used has no effect on whether it overheats."

Even if creating a problem story doesn't provide a solution, it can offer useful context to users’ problems and additional ideas to be explored. Given that divergence is one of the fundamental practices of creative problem-solving, it’s a good idea to incorporate it into each tool you use.

Brainstorming

Brainstorming is a tool that can be highly effective when guided by the iterative qualities of the design thinking process. It involves openly discussing and debating ideas and topics in a group setting. This facilitates idea generation and exploration as different team members consider the same concept from multiple perspectives.

Hosting brainstorming sessions can result in problems, such as groupthink or social loafing. To combat this, leverage a three-step brainstorming method involving divergence and convergence :

• Have each group member come up with as many ideas as possible and write them down to ensure the brainstorming session is productive.
• Continue the divergence of ideas by collectively sharing and exploring each idea as a group. The goal is to create a setting where new ideas are inspired by open discussion.
• Begin the convergence of ideas by narrowing them down to a few explorable options. There’s no "right number of ideas." Don't be afraid to consider exploring all of them, as long as you have the resources to do so.

Alternate Worlds

The alternate worlds tool is an empathetic approach to creative problem-solving. It encourages you to consider how someone in another world would approach your situation.

For example, if you’re concerned that the printers you produce overheat and catch fire, consider how a different industry would approach the problem. How would an automotive expert solve it? How would a firefighter?

Be creative as you consider and research alternate worlds. The purpose is not to nail down a solution right away but to continue the ideation process through diverging and exploring ideas.

Continue Developing Your Skills

Whether you’re an entrepreneur, marketer, or business leader, learning the ropes of design thinking can be an effective way to build your skills and foster creativity and innovation in any setting.

If you're ready to develop your design thinking and creative problem-solving skills, explore Design Thinking and Innovation , one of our online entrepreneurship and innovation courses. If you aren't sure which course is the right fit, download our free course flowchart to determine which best aligns with your goals.

About the Author

Creative Problem Solving in Large Language and Vision Models – What Would it Take?

We advocate for a strong integration of Computational Creativity (CC) with research in large language and vision models (LLVMs) to address a key limitation of these models, i.e., creative problem solving. We present preliminary experiments showing how CC principles can be applied to address this limitation. Our goal is to foster discussions on creative problem solving in LLVMs and CC at prestigious ML venues.

Lakshmi Nair Georgia Institute of Technology Atlanta, GA, USA                        Evana Gizzi Tufts University Medford, MA, USA                        Jivko Sinapov Tufts University Medford, MA, USA

1 Introduction

Creativity is “ …the ability to come up with an idea which, relative to the pre-existing domain-space in one’s mind, one could not have had before. Whether any other person (or system) has already come up with it on an earlier occasion is irrelevant. ” Boden ( 1998 ) , p.216. For artificial agents, Computational Creativity (CC) is a multi-disciplinary field (spanning Philosophy, Psychology, Neuroscience, and Computer Science) that seeks to develop computational methods capable of generating creative outcomes reminiscent of creative processes in humans Gizzi et al. ( 2022 ) . Within CC, creative problem solving is a sub-area that requires an agent to discover – from its perspective – novel and previously unseen ways to accomplish a task. For example, in the absence of a ladle to scoop ingredients, an agent might creatively choose to substitute a bowl in place of the ladle. In this sense, creative problem solving encompasses creativity that is specifically task-oriented , as opposed to the generation of creative artifacts e.g., music or images.

While recent state-of-the-art large language models (LLMs) and vision-language models (VLMs) have demonstrated competency in artistic endeavours Rombach et al. ( 2021 ); Copet et al. ( 2023 ) , creative problem solving continues to be a shortcoming of these models (we use LLVM to denote the umbrella of both LLMs and VLMs). For instance, in Bubeck et al. ( 2023 ) , the authors point out that “discontinuous tasks” that require a certain “Eureka” idea, i.e., creative problem solving, is currently a limitation of models like GPT-4. Similar observations have been made in follow up work showing that state-of-the-art LLMs inherently possess poor creative problem solving capabilities compared to humans Tian et al. ( 2023 ); Naeini et al. ( 2023 ) . Given this obvious limitation, ongoing research in Machine Learning should seek to address the gap between LLVMs and creative problem solving, to further enhance the intelligent capabilities of these models. As defined in prior work, “ Intelligence is the ability to work and adapt to the environment with insufficient knowledge and resources. ” Pennachin and Goertzel ( 2007 ) , p.10. Demonstrated in hallmark examples of human ingenuity, like the makeshift C ⁢ O 2 𝐶 subscript 𝑂 2 CO_{2} italic_C italic_O start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT filter built onboard the Apollo-13 Cass ( 2005 ) , or the makeshift medical devices used to offset equipment shortages during COVID-19 Turner et al. ( 2020 ) , creative problem solving is especially important when dealing with resource-critical scenarios. Since humans may tend to “choke” under high pressure situations DeCaro et al. ( 2011 ) often limiting their CPS skills, autonomous agents equipped with LLVMs that have similar capabilities would be highly assistive and transformative to humans in high-stake environments. These include situations like rescue missions BBC ( 2012 ) or autonomous operation in human-inaccessible environments (e.g., space or underwater exploration) with limited resources Atkeson et al. ( 2018 ) . However, the exceptional degree of creative problem solving necessary for such assistance remains beyond the scope of LLVMs today, limiting their intelligence (See Appx. B.1 ).

We believe that a discussion of Computational Creativity is essential to addressing this limitation. It is our position that Machine Learning and Computational Creativity should be strongly integrated in research to enable effective creative problem solving in LLVMs and push the frontiers of their ingenuity.

2 Two Cultures Problem: Why does CC not receive a wider reception in ML?

Even though creative problem solving (CPS) is a shortcoming of existing LLVMs, Computational Creativity seldom finds its way into mainstream ML research. We believe this discrepancy aligns with the “two cultures” problem Hammond et al. ( 2013 ) (also corroborated in Van Heerden and Bas ( 2021 ); Lahikainen et al. ( 2024 ) ), and is motivated by three aspects of CC literature as it relates to creative problem solving: a) the lack of a precise definition of CPS makes it challenging to identify how existing approaches in LLVMs are deficient in CPS skills; b) the somewhat “abstract” computational descriptions of CPS in Computational Creativity is challenging to connect to practical algorithms in LLVMs; and c) the lack of standardized benchmarks make it harder to evaluate LLVMs for CPS. In our discussions relating to a) in Section 3.1 , b) in Section 4 , and c) Section 5 , we hope to address these gaps and encourage the ML community to think about how LLVMs can be augmented with creative problem solving skills through a deeper discussion of Computational Creativity.

To emphasize the applicability of principles from CC for creative problem solving in LLVMs, we discuss the seminal work of Margaret A. Boden from CC literature that introduces three forms of creativity, namely, “ exploratory ”, “ combinational ”, and “ transformational ” Boden ( 1998 ) . Prior work has discussed the extension of Boden’s forms of creativity to creative problem solving in AI Gizzi et al. ( 2022 ) , however, their work does not include recent advances in LLVMs nor how Boden’s principles can be extended to specific approaches for LLVMs.

Ongoing discussions by leading ML experts like Dr. Shane Legg, co-founder of DeepMind, have suggested that “search” could help such models perform creative problem solving, quote, “ … these foundational models are world models of a kind, and to do really creative problem solving, you need to start searching ” Patel ( 2023 ) . There has also been speculation that OpenAI’s Q ∗ superscript 𝑄 Q^{*} italic_Q start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT search (described as a “significant breakthrough” in popular media) could be targeting a similar approach Wang ( 2023 ); Anna Tong and Hu ( 2023 ) . Interestingly, we note that “search” as described here, can be linked to Boden’s proposed “exploratory” approach (Section 4.1.1 ). However, in Section 4 , we posit that “combinational” and “transformational” modes should be equally emphasized to achieve creative problem solving in LLVMs.

Although we choose to expand on Boden’s work as the focal point to drive our arguments in the main paper, it is not the only theory in CC that is relevant to this discussion. For completeness, we elaborate on additional CC theories and their applicability to creative problem solving in LLVMs in Appx. B .

3 From Task Planning to Creative Problem Solving

Creative problem solving can be broadly described as the process through which agents discover novel ways of accomplishing a task that, prior to the discovery, was unsolvable. Computationally, creative problem solving can be achieved through planning, learning, or hybrid approaches Gizzi et al. ( 2022 ) . Following a review of the different definitions of creative problem solving that have been proposed (Appx. A ), we believe the following most closely connects to existing formalisms in ML.

3.1 Definition of Creative Problem Solving

Gizzi et al. ( 2022 ) define the notion of a concept , as a state (of the environment and/or agent) or action. More generally, the authors denote C X subscript 𝐶 𝑋 C_{X} italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT as the set of all concepts relating to X 𝑋 X italic_X ( X 𝑋 X italic_X denotes environment states S 𝑆 S italic_S or actions A 𝐴 A italic_A ). Hence, C S subscript 𝐶 𝑆 C_{S} italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT denotes the set of all environmental states, and C A subscript 𝐶 𝐴 C_{A} italic_C start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT denotes the set of agent actions. Formally, the authors state their definition as (Page 7, (Gizzi et al., 2022 ) ):

Given an un-achievable goal due to an insufficient conceptual space, CPS refers to the process by which the agent discovers a new conceptual space C X ′ ⊈ C X not-subset-of-nor-equals subscript superscript 𝐶 ′ 𝑋 subscript 𝐶 𝑋 C^{\prime}_{X}\nsubseteq C_{X} italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ⊈ italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT , such that C X ′ = f ⁢ ( C X ) subscript superscript 𝐶 ′ 𝑋 𝑓 subscript 𝐶 𝑋 C^{\prime}_{X}=f(C_{X}) italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT = italic_f ( italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ) is the result of applying some function f 𝑓 f italic_f on the current conceptual space, enabling the agent to solve the previously unsolvable task by using C X ′ subscript superscript 𝐶 ′ 𝑋 C^{\prime}_{X} italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT .

As a simplified example, let us assume a robot that has a goal G 𝐺 G italic_G of transferring beans from a jar to a cooker: G = 𝐺 absent G= italic_G = { i ⁢ n 𝑖 𝑛 in italic_i italic_n (beans, cooker)}. Here, the initial state is defined as C S = subscript 𝐶 𝑆 absent C_{S}= italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT = { i ⁢ n 𝑖 𝑛 in italic_i italic_n (beans, jar), h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 hasContainability italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y (spoon)}. Let the actions be defined as C A = subscript 𝐶 𝐴 absent C_{A}= italic_C start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT = { s ⁢ c ⁢ o ⁢ o ⁢ p 𝑠 𝑐 𝑜 𝑜 𝑝 scoop italic_s italic_c italic_o italic_o italic_p (beans, X 𝑋 X italic_X , l ⁢ o ⁢ c s 𝑙 𝑜 subscript 𝑐 𝑠 loc_{s} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_s end_POSTSUBSCRIPT , l ⁢ o ⁢ c d 𝑙 𝑜 subscript 𝑐 𝑑 loc_{d} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_d end_POSTSUBSCRIPT )}, where, X 𝑋 X italic_X refers to an object that satisfies h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ⁢ ( ⋅ ) ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 ⋅ hasContainability(\cdot) italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y ( ⋅ ) (e.g., spoon), to scoop beans from l ⁢ o ⁢ c s 𝑙 𝑜 subscript 𝑐 𝑠 loc_{s} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_s end_POSTSUBSCRIPT to l ⁢ o ⁢ c d 𝑙 𝑜 subscript 𝑐 𝑑 loc_{d} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_d end_POSTSUBSCRIPT . If the robot has access to a spoon, the robot can use it to scoop the beans from the jar to the cooker. However, what if the robot did not have a spoon, but had a glass instead? By the definition of C S subscript 𝐶 𝑆 C_{S} italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT , the agent is unaware that h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 hasContainability italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y (glass) is true, making the goal un-achievable. By our definition, creative problem solving is the process by which the agent uses some function f ⁢ ( ⋅ ) 𝑓 ⋅ f(\cdot) italic_f ( ⋅ ) to discover a new conceptual space: f ⁢ ( C S ) = C S ′ = C S ⁢ ∪ 𝑓 subscript 𝐶 𝑆 subscript superscript 𝐶 ′ 𝑆 subscript 𝐶 𝑆 f(C_{S})=C^{\prime}_{S}=C_{S}\mathop{\cup} italic_f ( italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT ) = italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT = italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT ∪ { h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 hasContainability italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y  (glass)}. This would allow the agent to solve the previously unsolvable task by using the glass to scoop the beans instead.

Boden’s three forms of creativity denote three plausible functions for f ⁢ ( C X ) 𝑓 subscript 𝐶 𝑋 f(C_{X}) italic_f ( italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ) . CPS arises when the agent uses what it knows, to discover something new and the newly discovered knowledge is applied to solve a previously impossible task. We revisit the notion of conceptual spaces in Section 3.

In the remainder of this section, we discuss how typical task planning is achieved with LLVMs. We divide the discussion into three subsections based on the level of task planning abstraction where LLVMs are applied: a) high-level task planning, b) low-level task planning, and c) hybrid task planning. While not exhaustive, our review is meant to offer a general insight into how LLVMs are used for task planning, to identify entry points for introducing creative problem solving capabilities.

3.2 LLVMs for high-level task planning

Approaches for high-level task planning often involve using LLVMs to identify high-level goals for accomplishing a task. Some approaches to task planning with LLMs often take a user input specifying the task, and generate high-level task plans for accomplishing it. These approaches often use LLMs as a form of “knowledge base”, to extract actionable task plans from the models via appropriate prompting Huang et al. ( 2022 ) , further iterating over the generated task plan with repeated calls to the LLM as needed Prasad et al. ( 2023 ) .

In the context of Reinforcement Learning (RL), prior work has focused on using LLMs to suggest high-level goals for an RL agent Du et al. ( 2023 ) . Dubbed as ELLMs (Exploring with LLMs), an RL agent provides its current state to an LLM via a prompt, and receives a goal suggestion from the LLM that is then used to shape the reward and the agent exploration. Further work has extended this approach to incorporate the use of experience memory Zhang et al. ( 2023a ) . Existing approaches have also used LLMs to generate directed acyclic graphs composed of sub-goal states to aid the exploration of an RL agent Shukla et al. ( 2023 ) .

3.3 LLVMs for low-level task planning

Approaches for low-level task planning involve using LLMs to generate low-level code for performing a task. In contrast to high-level planning, where high-level goals and sub-goals are generated, these approaches use LLMs to directly generate low-level execution code via appropriate API calls Liang et al. ( 2023 ) . Other approaches have also investigated the capacity of LLMs to generate task plans via a low-level planning language such as PDDL Silver et al. ( 2023 ) , including iterating over the generated plan descriptions in case of errors Guan et al. ( 2023 ) . In terms of low-level planning using VLMs, prior work has introduced an approach that uses a diffusion model to generate robot trajectories conditioned on language and the current visual state of the robot Chen et al. ( 2023 ) .

3.4 Hybrid high and low-level planning with LLVMs

Hybrid approaches use LLVMs both for high-level goal generation as well as low-level planning. For instance, in Li et al. ( 2023 ) , user inputs are passed as LLM prompts to generate high-level plans. The high-level plans are then converted to low-level plans for robot execution via LLMs specialized for coding. Other approaches have used a high-level LLM planner, a VLM perceiver, and a low-level LLM planner for re-planning with both visual and language inputs Skreta et al. ( 2024 ) .

3.5 Summary

Given this overview, we see that LLVMs both at the high-level and low-level, can be modified to incorporate creative problem solving into task planning. For instance, the high-level task plans generated can encompass a novel substitution for a missing object, whereas the low-level task plan can generate an appropriate trajectory for creatively using the object. While the above approaches could, in principle, be studied within the framework of creative problem solving, that is not usually how the problem is formulated; there is a lack of paradigms for studying creative problem solving beyond just, “do you solve the problem or not?” . Creative problem solving needs a fundamental rethinking of the typical problem formulations and approaches in ML. The next section is aimed at ways in which ML approaches in LLVMs can be reformulated from the perspective of CC.

4 Augmenting LLVM embedding spaces for creative problem solving

In this section, we discuss how principles from CC can be extended to LLVMs for creative problem solving. We begin with Boden’s definition of “conceptual spaces” as “ [conceptual space] is the generative system that underlies the domain and defines a certain range of possibilities: chess moves, or molecular structures, or jazz melodies ” Boden ( 2005 ) , p.18 and “ … in short, any reasonably disciplined way of thinking ” Boden ( 1998 ) , p.214. By this definition, the embedding space of an LLVM describes its conceptual space or “ its way of thinking ”. Some evidence for this also comes from existing work that introduces an approach for enabling LLMs to interpret continuous embedding spaces via natural language. Given an embedding vector representing an interpolation of different concepts, the model is able to interpret a text prompt in the context of the supplied embedding Tennenholtz et al. ( 2023 ) . The embedding thus determines the model’s way of thinking. Hence, a discussion of enabling creative problem solving in LLVMs should target their embedding space. To this end, we explore two questions: a) how can LLVM embedding spaces be augmented to achieve creative problem solving, and b) what information should they be augmented with? Aligning with our original position, we show that CC literature can offer insights into these questions.

4.1 How can LLVM embedding spaces be augmented?

In this section, we draw parallels between Boden’s three forms of creativity and existing approaches in LLVMs. We further elaborate on how the three forms of creativity may enhance the potential of LLVMs to perform creative problem solving. We note that the ML approaches discussed in this section do not specifically perform creative problem solving. However, we discuss how they could potentially be extended to do so, by leveraging references from the CC literature.

4.1.1 Exploratory Creativity

Exploratory approaches involve exploration within the conceptual or equivalently, the embedding space of the model, and most closely relates to “search”. Note that the term “exploration” here differs from its usage in RL, instead referring to exploration through the model’s embedding space . Several existing approaches in the ML literature involve searching the output space of LLMs with the goal of improving the performance of these models. The “tree-of-thought” model generates a “tree” of next possible LLM outputs, and searches through the states via Breadth-first or Depth-first search to reach the desired goal state, often guided by heuristics Yao et al. ( 2023 ) . Numerous other approaches have built upon a similar strategy, such as using Monte-Carlo Tree Search (MCTS) Zhou et al. ( 2023 ); Feng et al. ( 2023 ) , beam search Zhang et al. ( 2023b ) or integrating pruning to remove sub-par candidates Golovneva et al. ( 2023 ) .

Extension of exploratory creativity to LLVMs: An important point to note here is that these approaches involve searching exclusively within the output “solution space” of the LLMs rather than directly operating in the embedding space itself. In contrast to operating in the solution space of the LLM, exploratory approaches directly within the LLMs’ embedding space would not be limited by what the LLM can generate as output – “ Some exploration merely shows us the nature of the relevant conceptual space that we had not explicitly noticed before ” Boden ( 2005 ) , p.18. To effectively reveal the full extent of the conceptual space for creative problem solving, the approach should not be limited by the outputs the LLVM can generate. Rather, the generated (creative) outputs itself should be the result of heuristic or non-heuristic based search within the model’s embedding space. However, to the best of our knowledge current approaches have not focused on LLVMs from this perspective, and have also not applied search to embedding spaces of Vision-LMs. Regardless, exploratory approaches are still limited by the dimensions of the model’s embedding space. “ To overcome a limitation in the conceptual space, one must change it in some way ” Boden ( 2005 ) , p.18 - this leads us to combinational and transformational creativity.

4.1.2 Combinational Creativity

Combinational approaches involve combining two concepts to create something new - “ A novel combination of two familiar ideas is something which did not happen before. ” Boden ( 1998 ) , p.213. We can broadly translate this to a function that takes in multiple concepts within an LLVM’s embedding space to output a novel concept.

One way of extending this definition to LLVMs involves applying cross-attention layers. The attention operation is defined as Vaswani et al. ( 2017 ) :

where, Q 𝑄 Q italic_Q , K 𝐾 K italic_K and V 𝑉 V italic_V denote query, keys and values respectively, and d k subscript 𝑑 𝑘 d_{k} italic_d start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT denotes the dimensionality of the keys. Cross-attention involves passing K 𝐾 K italic_K and V 𝑉 V italic_V from a different model, e.g., in Flamingo Alayrac et al. ( 2022 ) , the keys and values represent visual input (from a separate vision encoder) and queries represent a language input. By applying cross attention in this manner, the embedding space of a model can be extended with capabilities of another model. In Bansal et al. ( 2024 ) the authors show that using cross-attention layers can help augment an anchor LLM with an augmenting LLM’s capabilities to perform a task that the anchor LLM was incapable of achieving before - hinting at some creative possibilities of this method.

Other approaches in LLVMs, while using “combinations” in some way, do not conform to the notion of combinational creativity . This includes, for instance, approaches that perform arithmetic combination of LLM weights to enhance the model performance Matena and Raffel ( 2022 ); Ilharco et al. ( 2022 ) . Or approaches that combine image and text embeddings via concatenation Kim et al. ( 2021 ) or a scaled dot product at the output Radford et al. ( 2021 ) . While these approaches may be useful in imparting multi-modal capabilities, however, they do not lead to combinational creativity since the combination occurs external to the models as opposed to within the model’s embedding space.

Extension of Combinational Creativity to LLVMs: The ML approaches described here involve combining embedding spaces across models. Existing approaches have not looked at combining concepts within the same model’s embedding space. The extension of combinational creativity to LLVMs is much more apparent in the sense of conceptual blending Fauconnier and Turner ( 2003 ) for generation of creative artifacts, e.g., via blending of artistic styles. However, the extension of combinational creativity to creative problem solving is less obvious, and CC literature offers us further insights for making this connection. Typical conceptual blending corresponds to a form of “aesthetic combination”, whereas creative problem solving would benefit from “functional combinations” Chen et al. ( 2018 ) . Functional combination combines the functions (as opposed to aesthetic) of two components, e.g., a coin combined with pliers could function as a makeshift screwdriver. The authors extend this framework to a combination of two nouns with a “base” noun (e.g., “pliers”) and “additive” noun (e.g., “coin”). An interesting possibility stems from this notion: Can a combination of embeddings of the same LLVM, corresponding to “base” and “additive” nouns (perhaps with some prior denoting the task), enable the LLVM to generate creative combinations of objects for solving a task? This question remains unexplored, and points to a potential research direction for LLVMs inspired by CC.

4.1.3 Transformational Creativity

Transformational approaches involve transforming existing conceptual spaces to produce new ones. Transforming conceptual spaces can involve “ altering existing rules ” Boden ( 1998 ) , p.216. One way of transforming a model’s embedding space involves fine-tuning or training Franceschelli and Musolesi ( 2023 ) . However, additional insight into transformational creative problem solving comes from prior work in CC, that describes creative problems as those with a poorly defined structure where a solution is not immediately apparent Olteteanu ( 2014 ) . And in such cases, “… re-representation being the process which transforms an ill-structured problem into a well-structured one with direct inference to a problem solution ” Olteteanu ( 2014 ) , p.1. The notion of “re-representing” or “redefining” the problem can be best captured in the input prompts provided to an LLVM. This most closely connects to prompt engineering and in-context learning (ICL).

Prompt engineering augments LLVMs with task specific hints, called prompts, to adapt the LLVM to new tasks Gu et al. ( 2023 ) . Relatedly, in-context learning is a prompting method that provides the LLVM with instructions for solving a new task without requiring additional training. Prior work has shown that in-context learning and gradient-based optimization are equivalent Von Oswald et al. ( 2023 ) , thus connecting ICL to training or fine-tuning.

Extension of transformational creativity to LLVMs: Task re-representations for creative problem solving, through prompting or ICL, has not been well explored within ML. Prompt engineering and ICL is a challenging task, since model performance depends strongly on the chosen prompts Rubin et al. ( 2021 ) , further compounded by the fact that creative problems are inherently poorly defined Olteteanu ( 2014 ) . However, useful insights can be derived from CC literature. For instance, regarding problems that require creatively re-purposing objects, the Object-replacement-object-composition (OROC) framework Olteţeanu and Falomir ( 2016 ) illustrates re-representations of tasks, that can be translated into prompts. The paper defines three different types of creative tasks involving objects, and their task re-representations as (from Olteţeanu and Falomir ( 2016 ) , p.16):

Replace an unfound object needed for a task with other objects present in the environment: “If I do not have an object X, which I would normally use because of its affordance 1 1 1 Affordance is defined as the relation between an agent, action and object, e.g., bowls have the “contain” affordance for humans. A ⁢ f X 𝐴 subscript 𝑓 𝑋 Af_{X} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT , what other object Y could I use, so that I can get a similar affordance, A ⁢ f X ≈ A ⁢ f Y 𝐴 subscript 𝑓 𝑋 𝐴 subscript 𝑓 𝑌 Af_{X}\approx Af_{Y} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_Y end_POSTSUBSCRIPT ? ”

𝐴 subscript 𝑓 𝑌 1 𝐴 subscript 𝑓 𝑌 2 … 𝐴 subscript 𝑓 𝑌 𝑛 Af_{X}\approx Af_{X^{\prime}},Af_{X}\approx Af_{Y1}+Af_{Y2}+...+Af_{Yn} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_X start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT , italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_Y 1 end_POSTSUBSCRIPT + italic_A italic_f start_POSTSUBSCRIPT italic_Y 2 end_POSTSUBSCRIPT + … + italic_A italic_f start_POSTSUBSCRIPT italic_Y italic_n end_POSTSUBSCRIPT ? ”

• subscript 𝑌 1 subscript 𝑌 2 … subscript 𝑌 𝑛 Y_{1};Y_{2};...;Y_{n} italic_Y start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ; italic_Y start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ; … ; italic_Y start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT which are components of object Y 𝑌 Y italic_Y could I use to obtain an object Y i ′ subscript superscript 𝑌 ′ 𝑖 Y^{\prime}_{i} italic_Y start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT with an equivalent or similar affordance, A ⁢ f X ≈ A ⁢ f Y ′ ⁢ i 𝐴 subscript 𝑓 𝑋 𝐴 subscript 𝑓 superscript 𝑌 ′ 𝑖 Af_{X}\approx Af_{Y^{\prime}i} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_Y start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT italic_i end_POSTSUBSCRIPT ? ”

For task re-representation, affordances can refer to object properties that are relevant to the task, e.g., in some cases the shape may be relevant and in other cases, the material Olteţeanu and Falomir ( 2016 ) . Within LLVMs, the affordances A ⁢ f X 𝐴 subscript 𝑓 𝑋 Af_{X} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT or A ⁢ f Y 𝐴 subscript 𝑓 𝑌 Af_{Y} italic_A italic_f start_POSTSUBSCRIPT italic_Y end_POSTSUBSCRIPT can be defined via natural language, or other modalities such as images. In the following section, we present preliminary experiments on using LLVMs for object replacement, with prompts that are inspired by the above task re-representations. However, an in-depth application of these re-representations as defined in CC to in-context learning in LLVMs remains unexplored.

4.1.4 Summary

In the previous sections, we drew parallels between Boden’s three forms of creativity and approaches in LLVMs, further emphasizing how principles from CC can potentially help enable creative problem solving skills in these models.

Integration with task planning: Given the three methods, we see that transformational and combinational approaches may be especially aligned with LLVMs for high-level task planning. In contrast, exploratory methods may be suited to low-level planning, e.g., trajectory generation.

Creative problem solving as a combination of the three methods: An effective approach to creative problem solving may require all the three methods described in this section. While papers have explored chaining of LLMs within frameworks (often via prompts) Karpas et al. ( 2022 ); Ling et al. ( 2023 ) , the individual LLMs themselves do not exhibit the characteristics described here. Existing frameworks in CC have shown that achieving creative problem solving would take a combination of all three methods, each of which is triggered in different contexts Olteteanu ( 2014 ) . This presents potential opportunities for ML approaches that develop frameworks using multiple LLVMs, e.g., extending CC frameworks such as “ CreaCogs ” Olteţeanu and Falomir ( 2016 ) can be highly beneficial for productive developments in ML.

4.2 What information should LLVM embeddings be augemented with?

In the previous section, we discussed three methods for augmenting LLVM embedding spaces. In this section, we explore the question: “What information should be targeted by the three methods when augmenting the embedding space for creative problem solving?”. In the previous section, we discussed this in the context of OROC. According to the OROC framework Olteţeanu and Falomir ( 2016 ) , information about object affordances could enable models to re-represent the task, such that the solution becomes evident. We propose a small experiment to validate whether the principles of transformational creativity from OROC are useful to LLVMs. We note that creativity can occur in various contexts, e.g., creatively solving a math problem or creatively playing a chess move, each of which would require different information. However, to facilitate the discussion in this paper, we focus our scope on tasks that require innovatively replacing missing objects (OROC Task #1).

Note on embeddings vs. concepts: Our work connects “conceptual spaces” (or “concepts”) as defined in Computational Creativity literature, to “embedding spaces” (or “embeddings”) as defined in typical LM literature. We use “concepts” and “embeddings” interchangeably in this context. We make this connection to note that existing methods in Computational Creativity that operate on conceptual spaces translate to ML algorithms that operate on the LM’s embedding space. In this section, we connect the concept of “affordances” to the “embeddings” of the LLVMs in our experiments. Our goal is to show how the model can be prompted via an approach inspired by transformational creativity, to connect affordances of two seemingly distinct objects, e.g., a bowl and a spoon that appear distinct, but share the containability affordance.

4.2.1 Experiment Setup

We create a simple experiment setup that tests the “object replacement” principle from OROC, where we create test sets composed of images of objects for replacing one of five core objects: “Scoop”, “Hammer”, “Spatula”, “Toothpick”, and “Pliers”. We create two groups of tests: a) a nominal group where the actual object itself is available in each test set and requires no replacement (which serves as a form of baseline), and b) an object replacement group, where the nominal tool is missing and a creative replacement object should be chosen.

For each group, we create test sets with 4 objects each, chosen from a set of RGB images of 16 objects (Appendix Figure 3 ). We create 10 such test sets per core object (total 50 samples per model). Each test set only includes one ground truth object, along with three other random objects that will not suit as an appropriate replacement. In the nominal group, the ground truth is the actual object itself. In the object replacement group, the replacements are chosen based on self-assessment of the authors as (core object → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW replacement): “Scoop” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Bowl”; “Hammer” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Saucepan”; “Spatula” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Knife”; “Toothpick” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Safety pin”; “Pliers” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Scissors”. For each test case, we pass the images in the test set along with a prompt. We record whether the ground truth object image was chosen by the model for the prompt (i.e., assigned highest output probability) 2 2 2 CLIP generates probabilities that given images correspond to a text. ViLT and LLaVA respond with a text, and we evaluate if the model responded “yes” with a high probability for the ground truth. .

The nominal group is subjected to one type of prompt: “ Can this object be used as a ⟨ c o r e _ o b j e c t ⟩ ? \bigl{\langle}core\_object\bigl{\rangle}? ⟨ italic_c italic_o italic_r italic_e _ italic_o italic_b italic_j italic_e italic_c italic_t ⟩ ? ”. In the object replacement group, each test case is subjected to four types of prompts:

Baseline (regular) prompt: Same prompt as used in the nominal cases to obtain a baseline.

Prompt prepended with affordance information: the prompt includes additional information about the desired object affordances specified as object features.

Prompt prepended with task information: the prompt includes additional information about the desired task.

Prompt prepended with task and affordance information: the prompt includes additional information on the task and object affordance.

Case #2 aligns with task re-representations of OROC, and we explore cases #3 and #4 for comparison. We formulate our affordance prompts as brief versions of OROC’s task re-representations. According to Olteţeanu and Falomir ( 2016 ) affordances can be defined using shape features, which we apply to the prompts here. The full set of prompts is shown in Appendix Table 2 . The models that we explore include versions of CLIP Radford et al. ( 2021 ) , LLaVA Liu et al. ( 2024 ) , and ViLT Kim et al. ( 2021 ) obtained from HuggingFace. We use different model sizes ( B ase, L arge, H uge) and patch sizes (14, 16, 32). The open-source code for reproducing our experiment results (including our dataset and test cases) is available at: https://github.com/lnairGT/creative-problem-solving-LLMs . Appendix C includes more details on the experiments.

4.2.2 Results

In Table 1 , we see the performances of the different models in the nominal test group, where the object requires no creative replacement. The models perform > 90 % absent percent 90 >90\% > 90 % in such cases (except for ViLT). In Figure 2 , we see the performances (accuracy shown on a 0.0 − 1.0 0.0 1.0 0.0-1.0 0.0 - 1.0 scale) of the models in the object replacement test cases, where the object requires a creative replacement. For reference, a model that randomly picks an object achieves about 30% overall accuracy. Figure 2 shows average accuracies for the different prompting strategies across random test sets. From Table 1 to Figure 2 (“regular”), the models perform poorly when they need to creatively reason about object replacements, highlighting their limitation. Comparing the “Regular” tab in Figure 2 to “Affordance”, we see a general improvement in model performances, when object affordance information is provided , consistent with description of the OROC framework Olteţeanu and Falomir ( 2016 ) . However, information about the task (Figure 2 , “Task” ) leads to mostly detrimental results. Information about task and affordances (Figure 2 , “Task + Affordance”) does not lead to substantial improvements either, and is also detrimental in certain cases. We note that there is quite a variance in performances across the different models, which may be partially attributed to the original training datasets of the models. These observations warrant further exploration beyond the scope of this paper. Appendix D includes a detailed, class-wise breakdown of the results.

4.2.3 Summary

While the experiments that we conducted are only preliminary, they offer some validity that the extension of principles in Computational Creativity can help overcome limitations of LLVMs in creative problem solving. The notion of task re-representation via improved prompting warrants further investigation in LLVMs, with regards to how the prompts can be generated automatically based on the creative task.

The models used in our experiments have all been trained jointly in visual and text domains. Multi-modal prompting capabilities may be useful for achieving creative problem solving. It can be quite challenging to describe affordances in words (example of “hammers” in our tests) and they may be better described through other means, e.g., images or depth maps or spectral data for material properties Erickson et al. ( 2020 ) . This would require application of multi-modal LLVMs that can process a variety of data types Girdhar et al. ( 2023 ); Han et al. ( 2023 ) . Computational creativity can offer insights into meaningful representations of these different modalities that would help achieve creative problem solving, e.g., whether object material or shape matters more for one task vs. another Olteţeanu and Falomir ( 2016 ) .

It is also worth noting that the creative problem solving examples in our experiments are human-centric. For instance, robots may not have similar capabilities as humans to manipulate bowls for scooping. In such cases, LLVMs need to account for the affordances as described with respect to the agent , in order to derive creative solutions. However, that adds another level of complexity, yet to be explored, since these models are typically trained on human-centric data.

5 Evaluation of Creativity

An important discussion in the context of creative problem solving is, how can creative problem solving be evaluated? . Prior work has proposed that creativity necessitates both novelty and value Boden ( 1998 ); Runco and Jaeger ( 2012 ) , where the former guarantees that the generated outputs of a creative process are original, and the latter ensures that the generated outputs are useful. In the context of CPS, novelty refers to the discovery of new concepts (as defined in section 3.1 ), whereas value insists that the newly discovered concepts successfully solve the task. Hence, benchmarks for CPS should specifically evaluate how the task was solved (novelty and value) rather than the typical ML evaluation of whether the task was successful or not (value only). Some existing approaches that make this distinction describe problem settings that can be used to measure CPS skills of LLMs through the implicit integration of novelty and value measurements Tian et al. ( 2023 ); Naeini et al. ( 2023 ); Bisk et al. ( 2020 ); Talmor et al. ( 2022 ) . In Tian et al. ( 2023 ) , the authors create a dataset of 1600 real-world problems that necessarily involve creative reasoning abilities. Their proposed benchmark involves identifying novel approaches that can accomplish the given task (value). Similarly, in Naeini et al. ( 2023 ) , the authors introduce the Only-Connect-Wall (OCW) dataset to measure CPS capabilities of LLMs. The authors in Bisk et al. ( 2020 ) explore physical commonsense reasoning that is more generally applicable, beyond object-based creative problems. The authors introduce Physical Interaction: Question Answering, or PIQA consisting of 16,000 QA pairs where each question is paired with two possible common-sense solutions with a ground truth. In Talmor et al. ( 2022 ) , the authors introduce CommonSenseQA 2.0 (CSQA2) dataset consisting of both object-based and non-object based creative problems. The dataset consists of 14,343 questions distributed across 1,868 distinct topics. Currently, to the best of our knowledge, there are no standard benchmarks available to measure CPS skills of VLMs, although our preliminary experiments show one way to measure this using the task of object substitution.

6 Conclusion and Future Work

In this paper, we argued that an effective approach for enabling creative problem solving – currently a key limitation of LLVMs – should derive from Computational Creativity literature. To emphasize this at each juncture, we discussed the specific principles from CC that can be extended to achieve creative problem solving in LLVMs, describing the potential for further research with these insights. It is rare to see special tracks or workshops targeted at Computational Creativity within more prestigious ML conferences. These programs typically focus on creative artifact generation and art (such as the NeurIPS Workshop on Machine Learning for Creativity and Design NeurIPS ( 2022 ) or the recent tutorial at EMNLP on Creative Natural Language Generation Chakrabarty et al. ( 2023 ) ), but do not discuss CPS, thus failing to bridge the gap between CC and ML. We hope to see a deeper integration of the CC communities at such strong ML venues. We hope to encourage the reader to view creative problem solving and ML holistically, through the lens of Computational Creativity.

7 Limitations

Literature outside of Computational Creativity that enables CPS is unexplored: Our paper predominantly focuses on CC literature. This work does not cover literature beyond CC that can potentially inform creative problem solving in LLVMs. Although CC literature broadly encompasses psychology, neuroscience and philosophy, our future work seeks to explore specific literature within these sub-domains and discuss their applicability to creative problem solving and ML.

Lack of an explicit creative problem solving algorithm for LLVMs: Since the scope of our work aligns with a position paper, we have not focused on developing a concrete algorithm for creative problem solving in LLVMs. The prompting strategies explored in our preliminary experiments are manually specified, and our work does not elaborate on how these prompts may be automatically discovered. While our paper seeks to address some of the key gaps that prevent the application of CC literature to ML, there are still several unanswered questions when it comes to the practical implementation of an ML approach: e.g., what is a good representation for concepts that facilitate creative problem solving (symbolic, non-symbolic, or hybrid)? What is a good problem formulation for a given creative problem solving task (planning or learning)? etc. However, these questions are not directly answered within the scope of our work.

8 Ethical Considerations

The authors do not have specific ethical considerations to be highlighted with respect to this work.

• Alayrac et al. (2022) Jean-Baptiste Alayrac, Jeff Donahue, Pauline Luc, Antoine Miech, Iain Barr, Yana Hasson, Karel Lenc, Arthur Mensch, Katherine Millican, Malcolm Reynolds, et al. 2022. Flamingo: a visual language model for few-shot learning. Advances in Neural Information Processing Systems , 35:23716–23736.
• Anna Tong and Hu (2023) Jeffrey Dastin Anna Tong and Krystal Hu. 2023. Openai researchers warned board of ai breakthrough ahead of ceo ouster, sources say. https://www.reuters.com/technology/sam-altmans-ouster-openai-was-precipitated-by-letter-board-about-ai-breakthrough-2023-11-22/. [Online; accessed 19-Jan-2024].
• Atkeson et al. (2018) Christopher G Atkeson, PW Babu Benzun, Nandan Banerjee, Dmitry Berenson, Christoper P Bove, Xiongyi Cui, Mathew DeDonato, Ruixiang Du, Siyuan Feng, Perry Franklin, et al. 2018. What happened at the darpa robotics challenge finals. The DARPA robotics challenge finals: Humanoid robots to the rescue , pages 667–684.
• Bansal et al. (2024) Rachit Bansal, Bidisha Samanta, Siddharth Dalmia, Nitish Gupta, Shikhar Vashishth, Sriram Ganapathy, Abhishek Bapna, Prateek Jain, and Partha Talukdar. 2024. Llm augmented llms: Expanding capabilities through composition. arXiv preprint arXiv:2401.02412 .
• BBC (2012) BBC. 2012. Us navy funds ’macgyver’ robot that can create tools. https://www.bbc.com/news/technology-19902954 . [Online; accessed 9-April-2024].
• Bisk et al. (2020) Yonatan Bisk, Rowan Zellers, Jianfeng Gao, Yejin Choi, et al. 2020. Piqa: Reasoning about physical commonsense in natural language. In Proceedings of the AAAI conference on artificial intelligence , volume 34, pages 7432–7439.
• Boden (1998) Margaret A. Boden. 1998. Creativity and Artificial Intelligence. Artificial Intelligence , 1-2:347–356.
• Boden (2005) Margaret A. Boden. 2005. What is creativity? Creativity in human evolution and prehistory .
• Bubeck et al. (2023) Sébastien Bubeck, Varun Chandrasekaran, Ronen Eldan, Johannes Gehrke, Eric Horvitz, Ece Kamar, Peter Lee, Yin Tat Lee, Yuanzhi Li, Scott Lundberg, et al. 2023. Sparks of artificial general intelligence: Early experiments with gpt-4. arXiv preprint arXiv:2303.12712 .
• Cass (2005) Stephen Cass. 2005. Apollo 13, we have a solution. IEEE Spectrum On-line, 04 , 1.
• Chakrabarty et al. (2023) Tuhin Chakrabarty, Vishakh Padmakumar, He He, and Nanyun Peng. 2023. Creative natural language generation. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: Tutorial Abstracts , pages 34–40.
• Chen et al. (2023) Lili Chen, Shikhar Bahl, and Deepak Pathak. 2023. Playfusion: Skill acquisition via diffusion from language-annotated play. In Conference on Robot Learning , pages 2012–2029. PMLR.
• Chen et al. (2018) Liuqing Chen, Pan Wang, Feng Shi, Ji Han, Peter Childs, et al. 2018. A computational approach for combinational creativity in design. In DS 92: Proceedings of the DESIGN 2018 15th International Design Conference , pages 1815–1824.
• Copet et al. (2023) Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi, and Alexandre Défossez. 2023. Simple and controllable music generation. arXiv preprint arXiv:2306.05284 .
• DeCaro et al. (2011) Marci S DeCaro, Robin D Thomas, Neil B Albert, and Sian L Beilock. 2011. Choking under pressure: multiple routes to skill failure. Journal of experimental psychology: general , 140(3):390.
• Du et al. (2023) Yuqing Du, Olivia Watkins, Zihan Wang, Cédric Colas, Trevor Darrell, Pieter Abbeel, Abhishek Gupta, and Jacob Andreas. 2023. Guiding pretraining in reinforcement learning with large language models. arXiv preprint arXiv:2302.06692 .
• Erickson et al. (2020) Zackory Erickson, Eliot Xing, Bharat Srirangam, Sonia Chernova, and Charles C Kemp. 2020. Multimodal material classification for robots using spectroscopy and high resolution texture imaging. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages 10452–10459. IEEE.
• Fauconnier and Turner (2003) Gilles Fauconnier and Mark Turner. 2003. Conceptual blending, form and meaning. Recherches en communication , 19:57–86.
• Fei et al. (2022) Nanyi Fei, Zhiwu Lu, Yizhao Gao, Guoxing Yang, Yuqi Huo, Jingyuan Wen, Haoyu Lu, Ruihua Song, Xin Gao, Tao Xiang, et al. 2022. Towards artificial general intelligence via a multimodal foundation model. Nature Communications , 13(1):3094.
• Feng et al. (2023) Xidong Feng, Ziyu Wan, Muning Wen, Ying Wen, Weinan Zhang, and Jun Wang. 2023. Alphazero-like tree-search can guide large language model decoding and training. arXiv preprint arXiv:2309.17179 .
• Franceschelli and Musolesi (2023) Giorgio Franceschelli and Mirco Musolesi. 2023. On the creativity of large language models. arXiv preprint arXiv:2304.00008 .
• Gilhooly (2016) Kenneth J Gilhooly. 2016. Incubation and intuition in creative problem solving. Frontiers in psychology , 7:1076.
• Girdhar et al. (2023) Rohit Girdhar, Alaaeldin El-Nouby, Zhuang Liu, Mannat Singh, Kalyan Vasudev Alwala, Armand Joulin, and Ishan Misra. 2023. Imagebind: One embedding space to bind them all .
• Gizzi et al. (2022) Evana Gizzi, Lakshmi Nair, Sonia Chernova, and Jivko Sinapov. 2022. Creative problem solving in artificially intelligent agents: A survey and framework. Journal of Artificial Intelligence Research , 75:857–911.
• Goertzel (2014) Ben Goertzel. 2014. Artificial general intelligence: concept, state of the art, and future prospects. Journal of Artificial General Intelligence , 5(1):1.
• Golovneva et al. (2023) O. Golovneva, S. O’Brien, R. Pasunuru, T. Wang, L. Zettlemoyer, M. Fazel-Zarandi, and A. Celikyilmaz. 2023. Pathfinder: Guided search over multi-step reasoning paths. arXiv preprint arXiv:2312.05180 .
• Grudin and Jacques (2019) Jonathan Grudin and Richard Jacques. 2019. Chatbots, humbots, and the quest for artificial general intelligence. In Proceedings of the 2019 CHI conference on human factors in computing systems , pages 1–11.
• Gu et al. (2023) Jindong Gu, Zhen Han, Shuo Chen, Ahmad Beirami, Bailan He, Gengyuan Zhang, Ruotong Liao, Yao Qin, Volker Tresp, and Philip Torr. 2023. A systematic survey of prompt engineering on vision-language foundation models. arXiv preprint arXiv:2307.12980 .
• Guan et al. (2023) Lin Guan, Karthik Valmeekam, Sarath Sreedharan, and Subbarao Kambhampati. 2023. Leveraging pre-trained large language models to construct and utilize world models for model-based task planning. arXiv preprint arXiv:2305.14909 .
• Guilford (1967) Joy P Guilford. 1967. Creativity: Yesterday, today and tomorrow. The Journal of Creative Behavior , 1(1):3–14.
• Hammond et al. (2013) Adam Hammond, Julian Brooke, and Graeme Hirst. 2013. A tale of two cultures: Bringing literary analysis and computational linguistics together. In Proceedings of the Workshop on Computational Linguistics for Literature , pages 1–8.
• Han et al. (2023) Jiaming Han, Kaixiong Gong, Yiyuan Zhang, Jiaqi Wang, Kaipeng Zhang, Dahua Lin, Yu Qiao, Peng Gao, and Xiangyu Yue. 2023. Onellm: One framework to align all modalities with language .
• Hélie and Sun (2010) Sebastien Hélie and Ron Sun. 2010. Incubation, insight, and creative problem solving: a unified theory and a connectionist model. Psychological review , 117(3):994.
• Huang et al. (2022) Wenlong Huang, Pieter Abbeel, Deepak Pathak, and Igor Mordatch. 2022. Language models as zero-shot planners: Extracting actionable knowledge for embodied agents. In International Conference on Machine Learning , pages 9118–9147. PMLR.
• Ilharco et al. (2022) Gabriel Ilharco, Marco Tulio Ribeiro, Mitchell Wortsman, Suchin Gururangan, Ludwig Schmidt, Hannaneh Hajishirzi, and Ali Farhadi. 2022. Editing models with task arithmetic. arXiv preprint arXiv:2212.04089 .
• Karpas et al. (2022) Ehud Karpas, Omri Abend, Yonatan Belinkov, Barak Lenz, Opher Lieber, Nir Ratner, Yoav Shoham, Hofit Bata, Yoav Levine, Kevin Leyton-Brown, Dor Muhlgay, Noam Rozen, Erez Schwartz, Gal Shachaf, Shai Shalev-Shwartz, Amnon Shashua, and Moshe Tenenholtz. 2022. Mrkl systems: A modular, neuro-symbolic architecture that combines large language models, external knowledge sources and discrete reasoning .
• Kim et al. (2021) Wonjae Kim, Bokyung Son, and Ildoo Kim. 2021. Vilt: Vision-and-language transformer without convolution or region supervision. In International Conference on Machine Learning , pages 5583–5594. PMLR.
• Lahikainen et al. (2024) Joonas Lahikainen, Nadia M Ady, and Christian Guckelsberger. 2024. Creativity and markov decision processes. arXiv preprint arXiv:2405.14966 .
• Li et al. (2023) Boyi Li, Philipp Wu, Pieter Abbeel, and Jitendra Malik. 2023. Interactive task planning with language models. arXiv preprint arXiv:2310.10645 .
• Liang et al. (2023) Jacky Liang, Wenlong Huang, Fei Xia, Peng Xu, Karol Hausman, Brian Ichter, Pete Florence, and Andy Zeng. 2023. Code as policies: Language model programs for embodied control. In 2023 IEEE International Conference on Robotics and Automation (ICRA) , pages 9493–9500. IEEE.
• Ling et al. (2023) Zhan Ling, Yunhao Fang, Xuanlin Li, Tongzhou Mu, Mingu Lee, Reza Pourreza, Roland Memisevic, and Hao Su. 2023. Unleashing the creative mind: Language model as hierarchical policy for improved exploration on challenging problem solving .
• Liu et al. (2024) Haotian Liu, Chunyuan Li, Qingyang Wu, and Yong Jae Lee. 2024. Visual instruction tuning. Advances in neural information processing systems , 36.
• Ma et al. (2023) Yuxi Ma, Chi Zhang, and Song-Chun Zhu. 2023. Brain in a vat: On missing pieces towards artificial general intelligence in large language models. arXiv preprint arXiv:2307.03762 .
• Matena and Raffel (2022) Michael S Matena and Colin A Raffel. 2022. Merging models with fisher-weighted averaging. Advances in Neural Information Processing Systems , 35:17703–17716.
• Moor et al. (2023) Michael Moor, Oishi Banerjee, Zahra Shakeri Hossein Abad, Harlan M Krumholz, Jure Leskovec, Eric J Topol, and Pranav Rajpurkar. 2023. Foundation models for generalist medical artificial intelligence. Nature , 616(7956):259–265.
• Moruzzi (2020) Caterina Moruzzi. 2020. Artificial creativity and general intelligence. Journal of Science and Technology of the Arts .
• Naeini et al. (2023) Saeid Naeini, Raeid Saqur, Mozhgan Saeidi, John Giorgi, and Babak Taati. 2023. Large language models are fixated by red herrings: Exploring creative problem solving and einstellung effect using the only connect wall dataset. arXiv preprint arXiv:2306.11167 .
• NeurIPS (2022) NeurIPS. 2022. Workshop on machine learning for creativity and design. https://nips.cc/virtual/2022/workshop/49965. [Online; accessed 19-Jan-2024].
• Olteteanu (2014) Ana-Maria Olteteanu. 2014. Two general classes in creative problem-solving? an account based on the cognitive processess involved in the problem structure-representation structure relationship. Publications of the Institute of Cognitive Science .
• Olteţeanu and Falomir (2016) Ana-Maria Olteţeanu and Zoe Falomir. 2016. Object replacement and object composition in a creative cognitive system. towards a computational solver of the alternative uses test. Cognitive Systems Research , 39:15–32.
• Patel (2023) Dwarkesh Patel. 2023. Llms need search for problem solving - shane legg (deepmind founder). https://www.youtube.com/watch?v=qulfo6-54k0. [Online; accessed 19-Jan-2024].
• Pennachin and Goertzel (2007) Cassio Pennachin and Ben Goertzel. 2007. Contemporary approaches to artificial general intelligence. In Artificial general intelligence , pages 1–30. Springer.
• Prasad et al. (2023) Archiki Prasad, Alexander Koller, Mareike Hartmann, Peter Clark, Ashish Sabharwal, Mohit Bansal, and Tushar Khot. 2023. Adapt: As-needed decomposition and planning with language models. arXiv preprint arXiv:2311.05772 .
• Radford et al. (2021) Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. 2021. Learning transferable visual models from natural language supervision. In International conference on machine learning , pages 8748–8763. PMLR.
• Rombach et al. (2021) Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2021. High-resolution image synthesis with latent diffusion models. 2022 ieee. In CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 10674–10685.
• Rubin et al. (2021) Ohad Rubin, Jonathan Herzig, and Jonathan Berant. 2021. Learning to retrieve prompts for in-context learning. arXiv preprint arXiv:2112.08633 .
• Runco and Jaeger (2012) Mark A Runco and Garrett J Jaeger. 2012. The standard definition of creativity. Creativity research journal , 24(1):92–96.
• Sarathy and Scheutz (2018) Vasanth Sarathy and Matthias Scheutz. 2018. Macgyver problems: Ai challenges for testing resourcefulness and creativity. Advances in Cognitive Systems , 6:31–44.
• Shevlin et al. (2019) Henry Shevlin, Karina Vold, Matthew Crosby, and Marta Halina. 2019. The limits of machine intelligence: Despite progress in machine intelligence, artificial general intelligence is still a major challenge. EMBO reports , 20(10):e49177.
• Shukla et al. (2023) Yash Shukla, Wenchang Gao, Vasanth Sarathy, Alvaro Velasquez, Robert Wright, and Jivko Sinapov. 2023. Lgts: Dynamic task sampling using llm-generated sub-goals for reinforcement learning agents. arXiv preprint arXiv:2310.09454 .
• Silver et al. (2023) Tom Silver, Soham Dan, Kavitha Srinivas, Joshua B Tenenbaum, Leslie Pack Kaelbling, and Michael Katz. 2023. Generalized planning in pddl domains with pretrained large language models. arXiv preprint arXiv:2305.11014 .
• Skreta et al. (2024) Marta Skreta, Zihan Zhou, Jia Lin Yuan, Kourosh Darvish, Alán Aspuru-Guzik, and Animesh Garg. 2024. Replan: Robotic replanning with perception and language models. arXiv preprint arXiv:2401.04157 .
• Talmor et al. (2022) Alon Talmor, Ori Yoran, Ronan Le Bras, Chandra Bhagavatula, Yoav Goldberg, Yejin Choi, and Jonathan Berant. 2022. Commonsenseqa 2.0: Exposing the limits of ai through gamification. arXiv preprint arXiv:2201.05320 .
• Tennenholtz et al. (2023) Guy Tennenholtz, Yinlam Chow, Chih-Wei Hsu, Jihwan Jeong, Lior Shani, Azamat Tulepbergenov, Deepak Ramachandran, Martin Mladenov, and Craig Boutilier. 2023. Demystifying embedding spaces using large language models. arXiv preprint arXiv:2310.04475 .
• Tian et al. (2023) Yufei Tian, Abhilasha Ravichander, Lianhui Qin, Ronan Le Bras, Raja Marjieh, Nanyun Peng, Yejin Choi, Thomas L Griffiths, and Faeze Brahman. 2023. Macgyver: Are large language models creative problem solvers? arXiv preprint arXiv:2311.09682 .
• Turner et al. (2020) MC Turner, LV Duggan, BA Glezerson, and SD Marshall. 2020. Thinking outside the (acrylic) box: a framework for the local use of custom-made medical devices. Anaesthesia .
• Van Heerden and Bas (2021) Imke Van Heerden and Anil Bas. 2021. Ai as author–bridging the gap between machine learning and literary theory. Journal of Artificial Intelligence Research , 71:175–189.
• Vaswani et al. (2017) Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems , 30.
• Ventura (2014) Dan Ventura. 2014. Can a computer be lucky? and other ridiculous questions posed by computational creativity. In Artificial General Intelligence: 7th International Conference, AGI 2014, Quebec City, QC, Canada, August 1-4, 2014. Proceedings 7 , pages 208–217. Springer.
• Von Oswald et al. (2023) Johannes Von Oswald, Eyvind Niklasson, Ettore Randazzo, João Sacramento, Alexander Mordvintsev, Andrey Zhmoginov, and Max Vladymyrov. 2023. Transformers learn in-context by gradient descent. In International Conference on Machine Learning , pages 35151–35174. PMLR.
• Wallas (1926) Graham Wallas. 1926. The art of thought . 24. Harcourt, Brace.
• Wang (2023) Brian Wang. 2023. Openai q* could be based upon a* search without expansions. https://www.nextbigfuture.com/2023/11/openai-q-could-be-based-upon-a-search-without-expansions.html. [Online; accessed 19-Jan-2024].
• Wiggins (2006) Geraint A Wiggins. 2006. A preliminary framework for description, analysis and comparison of creative systems. Knowledge-based systems , 19(7):449–458.
• Xi et al. (2023) Zhiheng Xi, Wenxiang Chen, Xin Guo, Wei He, Yiwen Ding, Boyang Hong, Ming Zhang, Junzhe Wang, Senjie Jin, Enyu Zhou, et al. 2023. The rise and potential of large language model based agents: A survey. arXiv preprint arXiv:2309.07864 .
• Yao et al. (2023) Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Thomas L Griffiths, Yuan Cao, and Karthik Narasimhan. 2023. Tree of thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:2305.10601 .
• Zhang et al. (2023a) Danyang Zhang, Lu Chen, Situo Zhang, Hongshen Xu, Zihan Zhao, and Kai Yu. 2023a. Large language model is semi-parametric reinforcement learning agent. arXiv preprint arXiv:2306.07929 .
• Zhang et al. (2023b) Shun Zhang, Zhenfang Chen, Yikang Shen, Mingyu Ding, Joshua B Tenenbaum, and Chuang Gan. 2023b. Planning with large language models for code generation. arXiv preprint arXiv:2303.05510 .
• Zhou et al. (2023) Andy Zhou, Kai Yan, Michal Shlapentokh-Rothman, Haohan Wang, and Yu-Xiong Wang. 2023. Language agent tree search unifies reasoning acting and planning in language models. arXiv preprint arXiv:2310.04406 .

Appendix A Alternate Definitions of Creative Problem Solving

Prior work by Olteţeanu Olteteanu ( 2014 ) defines CPS from an object affordance perspective, where affordances broadly refer to action possibilities for objects, e.g., cups are pour-able and doors are open-able. The authors in Olteteanu ( 2014 ) define creative problems as nominal problem solving tasks that have a poor representational structure, and as “ the ability of a cognitive, natural, or artificial system to use new objects to solve a problem, other than the ones that have been stored in its memory as tools for that specific purpose (if any), or to create those objects by putting together objects or parts of objects the system has access to. Depending on the problem, objects can be either physical or abstract/informational (concepts, problem templates, heuristics or other forms of representations) ”. However, this definition is primarily object-creativity centered, and does not cover a wider range of creative problems.

Follow-up work by Sarathy and Scheutz Sarathy and Scheutz ( 2018 ) , define “ Macgyver-esque ” creativity as a planning task that involves “ generating, executing, and learning strategies for identifying and solving seemingly unsolvable real-world problems ”. They introduce the “ MacGyver Problem ” (MGP) as a planning problem with an unreachable goal state. Through the modification of the agent’s domain knowledge (through domain expansion and domain contraction ), the agent must discover new information and incorporate it into its existing domain knowledge, allowing the agent to accomplish the task. The domain expansion and contraction processes align with the divergent-convergent model of creative problem solving Guilford ( 1967 ) . The definition of an MGP aligns well with the formulation of planning problems in ML, but less with learning or hybrid planning-learning approaches.

Appendix B Alternate theories on creative problem solving and their applications to ML

While there is exhaustive literature regarding theories on general creativity, we focus specifically on creative problem solving, with three well received works: Divergent-Convergent Thinking Guilford ( 1967 ) , Explicit-Implicit Interaction Theory Hélie and Sun ( 2010 ) , and the Creative Systems Framework Wiggins ( 2006 ) . We discuss their applicability to ML in addition to the literature discussed in the main body of this paper. Our goal in this section is to further widen the discussion on integrating CC and ML to achieve creative problem solving in LLVMs, with additional literature.

B.0.1 Divergent-Convergent Thinking

In Guilford ( 1967 ) , the authors discuss the notion of “divergent-convergent” thinking. Divergent thinking or “divergent-production” (DP) abilities involve a more open-ended generation of a variety of ideas, whereas convergent thinking focuses on applying specific ideas to solve the problem.

Applicability to CPS in LLVMs: Prior work by Tian et al. ( 2023 ) have demonstrated the applicability of “divergent-convergent” thinking towards solving Macgyver problems. Similar in spirit to our experiments with VLMs in Section 4.2.1 , the authors prompt LLMs with descriptions of objects to enable the LLMs to reason about solving the task. Their work is, to the best of our knowledge, the only direct example demonstrating the value of CC literature in enabling CPS in LLMs.

B.0.2 Explicit-Implicit Interaction Theory

In Hélie and Sun ( 2010 ) , the authors introduce the Explicit-Implicit Interaction (EII) theory, building upon the seminal work in Wallas ( 1926 ) , that describes four stages of creativity: Preparation, incubation, illumination (i.e., insight), and verification. Preparation refers to the initial stage of searching in many different directions, which may fail to find a solution (i.e., impasse) in case of ill-defined problems (as is the case with CPS). Following an impasse, the incubation phase begins, where attention is not devoted to solving the problem. Over a period of time, illumination is the manifestation of the solution to the problem within the conscious thought (i.e., “Aha” moment). Finally, verification involves using deliberative thinking to assess if the solution indeed solves the problem.

Applicability to CPS in LLVMs: The authors in Hélie and Sun ( 2010 ) incorporate the four stages via a concrete computational method into the CLARION cognitive architecture. Prior work has also introduced a CPS framework for ML approaches inspired by the four stages Gizzi et al. ( 2022 ) . In their work, “preparation” aligns with problem formulation, either task learning or planning. Incubation and illumination aligns with knowledge representation (symbolic, non-symbolic, or hybrid), and knowledge manipulation (functions that manipulate the conceptual space). Lastly, verification aligns with evaluation (via simulation, real-world platforms, or benchmarks). Although these works do not explicitly cover LLVMs and related algorithms, they demonstrate the value of integrating CC literature in ML, and can serve as useful starting points for ML approaches towards creative problem solving in LLVMs.

B.0.3 Creative Systems Framework

In Wiggins ( 2006 ) , the author expands on Boden’s levels further in the context of a framework that formalizes creative systems. The paper defines: a) creative system, b) creative behavior, c) novelty, and d) value. The paper also discusses formalized notion of a universe of possibilities , and conceptual spaces . Crucially, the work describes the characteristics of a creative agent, that can help distinguish modes of failures within a creative system, namely: a) hopeless uninspiration – where there are no valued concepts within the universe; b) conceptual uninspiration – where there are no valued concepts within the conceptual space of the agent; and c) generative uninspiration – where an agent is unable to find a valued concept owing to the specific method (e.g., search) employed.

Applicability to CPS in LLVMs: While the discussion of novelty, value and conceptual spaces in Wiggins ( 2006 ) aligns with our descriptions in Section 4 , the different modes of uninspiration offers potential ways to assess failure modes in LLVMs. This allows agents to distinguish between systems where creative problem solving is not possible (hopeless uninspiration), as compared to systems where the conceptual space or the methodology for searching the conceptual space, may be at fault (conceptual or generative uninspiration). Although this approach has not been expanded in existing literature, it presents a promising direction for an evaluation framework that can distinguish CPS from non-CPS problems.

B.1 A potential link between creative problem solving and general intelligence

Existing literature hints at a potential link between creative problem solving and Artificial General Intelligence (AGI) - systems that are broadly capable of solving almost all tasks that humans can Shevlin et al. ( 2019 ) . For instance, in Moruzzi ( 2020 ) , p.85., the author argues that there exists a strong correlation between creativity and AGI: “ … features that systems need to develop in order to achieve general intelligence are aspects that they need to possess also to earn the attribute creative ”. In Goertzel ( 2014 ) , the author compiles a list of competencies deemed essential for achieving AGI, including creative capacities like “ conceptual invention ” and “ creative constructive play with objects ”. The processes of “insight” or “incubation” often associated with creative problem solving Hélie and Sun ( 2010 ); Gilhooly ( 2016 ) is also considered important for AGI Ventura ( 2014 ) . Taken together, it is likely that any promising vision of AGI would be incomplete without creative problem solving .

Alongside the heavy ongoing discussion of AGI surrounding LLVMs Bubeck et al. ( 2023 ); Fei et al. ( 2022 ); Ma et al. ( 2023 ); Xi et al. ( 2023 ); Moor et al. ( 2023 ); Grudin and Jacques ( 2019 ) , there is often little to no discussion of creative problem solving or Computational Creativity within mainstream ML. As described in Moruzzi ( 2020 ) , p.96, “ The investigation on the nature of creativity and on how it manifests itself not only in human but also in animal and artificial systems should, thus, not be intended as a niche discussion but, rather, as a fundamental research which can lay the foundations for further studies in artificial intelligence and its relation to humans ”. We hope that this work will encourage discussions of creative problem solving and Computational Creativity alongside discussions on AGI.

Appendix C Experiment Settings

Prompt type

Prompt

Regular

“can this object be used as a scoop?” “can this object be used as a hammer?” “can this object be used as a spatula?” “can this object be used as a toothpick?” “can this object be used as pliers?” “scoops must be concave and hollow. can this object be used as a scoop?” “hammers must be heavy and have a handle attached to a cylinder at the end. can this object be used as a hammer?” “spatulas must have a handle attached to a flat surface at the end. can this object be used as a spatula?” “toothpicks must have a pointed tip. can this object be used as a toothpick?” “pliers must have two-prongs. can this object be used as pliers?” “scoops can transfer beans from one jar to another jar. can this object be used as a scoop?” “hammers can hit a nail into the wall. can this object be used as a hammer?” “spatulas can spread butter onto a pan. can this object be used as a spatula?” “toothpicks can pick food caught between the teeth. can this object be used as a toothpick?” “pliers can grab a coin. can this object be used as pliers?” “scoops can transfer beans from one jar to another jar. scoops are concave and hollow. can this object be used as a scoop?” “hammers can hit a nail into the wall. hammers have a handle attached to a cylinder at the end. can this object be used as a hammer?” “spatulas can spread butter onto a pan. spatulas have a handle attached to a flat surface at the end. can this object be used as a spatula?” “toothpicks can pick food caught between the teeth. toothpicks have a pointed tip. can this object be used as a toothpick?” “pliers can grab a coin. pliers have two-prongs. can this object be used as pliers?” C.1 Data: Test images Figure 3 shows the test set of 16 RGB images of objects used for the object substitution task. From the shown image dataset, we create test sets with 4 objects each, chosen from the set of 16 object images. We create 10 such test sets per core object (total 50 samples per model). Each test set only includes one ground truth object, along with three other random objects that will not suit as an appropriate replacement. In the nominal group, the ground truth is the actual object itself. In the object replacement group, the replacements are chosen based on self-assessment of the authors as (core object → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW replacement): “Scoop” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Bowl”; “Hammer” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Saucepan”; “Spatula” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Knife”; “Toothpick” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Safety pin”; “Pliers” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Scissors”. C.2 Model: Checkpoints For all the models, we use pre-trained HuggingFace checkpoints, with no additional training or fine-tuning. The models are of different architecture sizes and patch sizes: “CLIP-B-32” uses the “openai/clip-vit-base-patch32” which is a base model with a patch size of 32. “CLIP-B-16” uses “openai/clip-vit-base-patch16” – a base model with patch size of 16. “CLIP-L-14” uses “openai/clip-vit-large-patch14” – a large model with patch size of 14. “CLIP-H-14” uses “laion/CLIP-ViT-H-14-laion2B-s32B-b79K” which is a “huge” model, with a patch size of 14. This model is trained with the 2 billion sample English subset of LAION-5B. For LLaVA, we use the “llava-hf/llava-1.5-7b-hf” with 7B parameters, version 1.5. Lastly, “VILT-B-32” uses “dandelin/vilt-b32-finetuned-vqa” trained for visual question answering. However, there is limited data available on HuggingFace regarding the model. C.3 Prompts used in testing In this section, we discuss the prompts used in the different testing conditions (see Table 2 ). We explore four classes of prompts for the creative object substitution task: “Regular”, “Affordance”, “Task” and “Task and affordance”. Regular prompts involve a direct prompt as to whether a given object will suffice as a substitute for the missing object. Affordance prompts, adds information about the desired affordances that are essential for replacing the missing object. Task prompts adds additional information on the task to be performed as context for whether a given object can be used as replacement for the missing object. Lastly, task and affordance prompts combine the task and object affordance information within the prompt. C.4 Testing Procedure For each test case, we pass the images in the test set along with a prompt belonging to one of the four classes described in Table 2 . We record whether the ground truth object image was chosen by the model for the prompt (i.e., assigned highest output probability). CLIP generates probabilities that given images correspond to a text. ViLT responds with a text, and we evaluate if the model responded “yes” with a high probability for the ground truth. C.5 Testing Infrastructure We used NVIDIA-A100 GPUs to run the evaluation. However, the models are not too large and we have tested and confirmed that the code can be executed on CPU only as well. Appendix D Continued Experiment Results In this section, we show the class-wise breakdown of the different models for the different prompting strategies (Figures 4 - 7 ). We note that “hammers” present a particularly challenging case for all the models, perhaps due to the fact that correlating affordance of a hammer to a saucepan textually is difficult. In contrast, all models with the augmented prompts typically perform well in the case of creatively replacing “toothpick” with “safety pin” – presumably indicating that specifying the relevant affordance textually in this case provides sufficient information. We repeated each experiment across multiple random seeds and found similar performances, showing that our general findings hold across different random cases. Generally, specifying object affordance information in the prompts leads to improved model performance. 461 IMAGES Creative Problem Solving To Be Innovative Creative Problem Solving in the Classroom: A Guide for Teachers The four stages of the creative problem-solving process. Osborn: Creative Problem-Solving Process Creative Problem-Solving Process Creative Problem Solving Process COMMENTS Model Pembelajaran Creative Problem Solving (CPS) Menurut Shoimin (2014), creative problem solving adalah model pembelajaran yang pemusatannya pada pengajaran dan keterampilan dalam memecahkan masalah. Ketika dihadapkan dengan suatu pernyataan, siswa dapat melakukan keterampilan dalam memecahkan masalah untuk memilih dan mengembangkan tanggapannya. Tidak hanya dengan cara menghafal tanpa ... Model pembelajaran Creative Problem Solving Creative Problem Solving (CPS) lebih menekankan pada pentingnya penemuan berbagai alternative ide dan gagasan, untuk mencari berbagai macam kemungkinan cara/tindakan yang digunakan untuk menyelesaikan permasalahan. Creative Problem Solving (CPS) adalah suatu model pembelajaran yang memusatkan pengajaran pada sebuah keterampilan (Pepkin, 2004: 1). Langkah-Langkah Model Pembelajaran Creative Problem Solving (CPS Banyak ahli yang menjelaskan bentuk penerapan model pembelajaran (CPS). Osborn seorang ahli pendidikan yang pertama kali memperkenalkan struktur (CPS) sebagai metode untuk menyelesaikan masalah secara kreatif yang kemudian menjelaskan 6 langkah pada proses model pembelajaran (CPS) berdasarkan criteria OFPISA model Osborn-Parnes, yaitu : Creative Problem Solving (CPS) untuk Meningkatkan ... Creative Problem Solving (CPS) adalah model pembelajaran yang berfokus pada keterampilan memecahkan masalah dan tantangan dalam menemukan solusi terbaik dengan cara berpikir kreatif, inovatif, dan imajinatif. Memiliki keterampilan berpikir kreatif sangat penting bagi siswa untuk mempersiapkan diri dalam perubahan yang terjadi begitu cepat di ... Metode Pembelajaran Creative Problem Solving: Menemukan Solusi dengan Creative Problem Solving (CPS) merupakan pendekatan yang digunakan untuk mengatasi masalah secara inovatif dan kreatif. Metode ini melibatkan langkah-langkah sistematik yang membuka pikiran kita untuk berpikir di luar kotak dan menemukan solusi yang tak terduga. Langkah pertama dalam metode CPS adalah mengidentifikasi masalah dengan jelas. Pengertian Model Pembelajaran Creative Problem Solving Model Creative Problem Solving (CPS) adalah suatu model pembelajaran yang berpusat pada keterampilan pemecahan masalah, yang diikuti dengan penguatan kreatifitas. Tujuan model Creative Problem Solving (CPS) menurut Hudojo (2008: 155) adalah untuk mengetahui ketuntasan belajar pada hasil belajar, keaktifan dan keterampilan berpikir dan proses siswa. Karakteristik dan Implementasi Model Pembelajaran Creative Problem Solving Model Pembelajaran Creative Problem Solving (CPS) adalah suatu model pembelajaran yang memusatkan pada pengajaran dan keterampilan pemecahan masalah, yang diikuti dengan penguatan keterampilan. Dalam hal ini peserta didik terlibat dalam memecahkan masalah sehingga peserta didik lebih berfikir aktif, kreatif dan dapat memperoleh manfaat yang ... (Pdf) Model Pembelajaran Creative Problem Solving (Cps) Untuk Keywords: creative thinking ability, academic achievement, creative problem solving Abstrak. Kemampuan berpikir kreatif merupakan salah satu dari keterampilan abad 21 yang harus dimiliki siswa. Model Pembelajaran Creative Problem Solving Adalah Cara Inovatif Temukan keajaiban di balik pembelajaran dengan model Creative Problem Solving! Melalui pendekatan ini, siswa diajak untuk berpikir kreatif, menemukan solusi yang inovatif, dan menghadapi tantangan dengan percaya diri. Dengan mengintegrasikan elemen kreativitas dalam proses pembelajaran, model ini memberikan ruang yang luas bagi imajinasi dan eksplorasi ide-ide baru. Penerapan Model Pembelajaran Creative Problem Solving (CPS ... masalah (Problem Solving). Salah satunya adalah model pembelajaran Creative Problem Solving. Model pembelajaran Creative Problem Solving (CPS) adalah suatu model pembelajaran yang melakukan pemusatan pada pengajaran dan keterampilan memecahkan masalah, yang diikuti dengan penguatan ketrampilan (Pepkin, 2004). PDF Model Pembelajaran Creative Problem Solving (Cps) Untuk Meningkatkan menunjukkan bahwa model pembelajaran CPS dapat meningkatkan hasil belajar dan kemampuan berpikir kreatif siswa. Kata kunci: kemampuan berpikir kreatif, hasil belajar, creative problem solving PENDAHULUAN Salah satu tujuan mata pelajaran kimia dalam sistem pendidikan di Indonesia adalah menumbuh kembangkan sikap kreativitas siswa. Model Pembelajaran Problem Solving (Penjelasan Lengkap) Pengertian Model Pembelajaran Problem Solving. Model pembelajaran problem solving adalah model yang mengutamakan pemecahan masalah dalam kegiatan belajar untuk memperkuat daya nalar yang digunakan oleh peserta didik agar mendapatkan pemahaman yang lebih mendasar dari materi yang disampaikan. Seperti yang diungkapkan Pepkin (dalam Shoimin, 2017 ... Model Pembelajaran Kooperatif Creative Problem Solving: Cara Seru untuk Dalam kesimpulannya, Model Pembelajaran Kooperatif Creative Problem Solving adalah model pembelajaran yang menarik, interaktif, dan menyenangkan. Model ini tidak hanya meningkatkan pemahaman siswa, tetapi juga melatih kemampuan berpikir kreatif dan inovatif, serta membangun rasa percaya diri dalam menghadapi masalah. ... PDF Model Pembelajaran Creative Problem Solving Untuk Meningkatkan ... memecahkan masalah. Salah satu model pembelajaran tersebut adalah model creative problem solving (CPS). Model CPS mampu membuat pembelajaran berlangsung lebih menyenangkan, dapat membangkitkan motivasi siswa dan mendorong siswa membangun pengetahuannya sendiri. Menurut Daties (2010) ada beberapa alasan memilih model CPS dalam Pengembangan model pembelajaran creative problem solving terintegrasi Creative Problem Solving yang memfasilitasi proses problem solving terhadap permasalahan tertentu akan menyiapkan ruang untuk peningkatan kreativitas dalam memecahkan masalah (Adams, 2010, p. 6). Model pembelajaran Creative Problem Solving adalah suatu model pembelajaran yang melakukan pemusatan pada pembelajaran dan keterampilan memecahkan ... PDF BAB 2 TINJAUAN TEORETIS 2.1 Kajian Pustaka 2.1.1 Creative Problem BAB 2 TINJAUAN TEORETIS2.1 Kajian Pustaka2.1.1 Creative Problem Solving (CPS)Model pembelajaran Creative Problem Solving (CPS) merupakan suatu model pembelajaran yang memusatkan pada keterampilan pemec. han masalah, yang diikuti dengan penguatan keterampilan (Leonard et al., 2019). Selain itu menurut Helen (2022), menyatakan bahwa model ... Kelebihan dan Kelemahan Model Pembelajaran Creative Problem Solving Model Pembelajaran Creative Problem Solving. Menurut Miftahul Huda pendekatan CPS mempunyai beberapa kelebihan dan kelemahan. Kelebihan-kelebihan dari pendekatan CPS ini adalah sebagai berikut: Pendekatan CPS ini lebih memberi kesempatan kepada siswa untuk memahami konsep-konsep dengan cara menyelesaikan suatu permasalahan. Pendekatan CPS dapat ... (PDF) Implementasi Model Pembelajaran Creative Problem Solving (CPS Implementasi Model Pembelajaran Creative Problem Solving (CPS) Untuk Meningkatkan Kemampuan Pemecahan Masalah Mahasiswa March 2019 Jurnal Derivat Jurnal Matematika dan Pendidikan Matematika 5(1):69-81 Penerapan Model Creative Problem Solving Untuk Meningkatkan Kemampuan Model Creative Problem Solving merupakan model pembelajaran yang menekankan pemecahan masalah dengan cara-cara kreatif. Kemampuan berpikir kreatif adalah kemampuan menyelesaikan masalah dengan menggunakan beragam cara dan jawaban, merinci gagasan, serta menghasilkan ide baru dan unik. Partisipan pada penelitian ini adalah siswa kelas V Sekolah ... PDF Penerapan Model Creative Problem Solving Untuk Meningkatkan Kemampuan pembelajaran dengan model Creative Problem Solving 2) Perbedaan peningkatan kemampuan berpikir kreatif siswa yang signifikan antara yang memperoleh Creative Problem Solving dan pembelajaran konvensional. Model Creative Problem Solving merupakan model pembelajaran yang menekankan pemecahan masalah dengan cara-cara kreatif. Kemampuan What is CPS? S. olving. CPS is a proven method for approaching a problem or a challenge in an imaginative and innovative way. It helps you redefine the problems and opportunities you face, come up with new, innovative responses and solutions, and then take action. If you search the Internet for "Creative Problem Solving," you'll find many variations ... (PDF) Kemampuan Berpikir Kreatif Ditinjau dari Self ... Kemampuan Berpikir Kreatif Ditinjau dari Self Regulated Learning dengan Pendekatan Open-Ended Pada Model Pembelajaran Creative Problem Solving June 2021 QALAMUNA Jurnal Pendidikan Sosial dan Agama ... What Is Creative Problem-Solving & Why Is It Important? Its benefits include: Finding creative solutions to complex problems: User research can insufficiently illustrate a situation's complexity. While other innovation processes rely on this information, creative problem-solving can yield solutions without it. Adapting to change: Business is constantly changing, and business leaders need to adapt. Creative Problem Solving in Large Language and Vision Models While recent state-of-the-art large language models (LLMs) and vision-language models (VLMs) have demonstrated competency in artistic endeavours Rombach et al. (); Copet et al. (), creative problem solving continues to be a shortcoming of these models (we use LLVM to denote the umbrella of both LLMs and VLMs). For instance, in Bubeck et al. (), the authors point out that "discontinuous tasks ... assignment essay homework paper phd report resume review speech thesis