Difference between revisions of "AI in education"
KevinYager (talk | contribs) (→Professions) |
KevinYager (talk | contribs) (→Various) |
||
| (18 intermediate revisions by the same user not shown) | |||
| Line 29: | Line 29: | ||
* [https://arxiv.org/abs/2308.02773 EduChat: A Large-Scale Language Model-based Chatbot System for Intelligent Education] | * [https://arxiv.org/abs/2308.02773 EduChat: A Large-Scale Language Model-based Chatbot System for Intelligent Education] | ||
* [https://eurekalabs.ai/ Eureka Labs] (founded by [https://en.wikipedia.org/wiki/Andrej_Karpathy Andrej Karpathy]) aims to create AI-driven courses (first course is [https://github.com/karpathy/LLM101n Intro to LLMs]) | * [https://eurekalabs.ai/ Eureka Labs] (founded by [https://en.wikipedia.org/wiki/Andrej_Karpathy Andrej Karpathy]) aims to create AI-driven courses (first course is [https://github.com/karpathy/LLM101n Intro to LLMs]) | ||
| + | |||
| + | ===LLMs=== | ||
| + | * 2024-12: [https://www.arxiv.org/abs/2412.16429 LearnLM: Improving Gemini for Learning] | ||
| + | |||
===Individual tools=== | ===Individual tools=== | ||
* Chatbot (OpenAI [https://chatgpt.com/ ChatGPT], Anthropic [https://www.anthropic.com/claude Claude], Google [https://gemini.google.com/app Gemini]) | * Chatbot (OpenAI [https://chatgpt.com/ ChatGPT], Anthropic [https://www.anthropic.com/claude Claude], Google [https://gemini.google.com/app Gemini]) | ||
| Line 66: | Line 70: | ||
* 2024-09: [https://docs.iza.org/dp17302.pdf Creative and Strategic Capabilities of Generative AI: Evidence from Large-Scale Experiments] | * 2024-09: [https://docs.iza.org/dp17302.pdf Creative and Strategic Capabilities of Generative AI: Evidence from Large-Scale Experiments] | ||
| − | === | + | ===Art=== |
* 2024-11: [https://doi.org/10.1038/s41598-024-76900-1 AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably] | * 2024-11: [https://doi.org/10.1038/s41598-024-76900-1 AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably] | ||
| + | * 2024-11: [https://www.astralcodexten.com/p/how-did-you-do-on-the-ai-art-turing How Did You Do On The AI Art Turing Test?] | ||
===Professions=== | ===Professions=== | ||
| + | * [https://agi.safe.ai/submit Humanity's Last Exam] | ||
| + | ** [https://x.com/alexandr_wang/status/1835738937719140440 Effort to build] a dataset of challenging (but resolvable) questions in specific domain areas, to act as a benchmark to test whether AIs are improving in these challenging topics. | ||
| + | |||
| + | ====Medical==== | ||
* 2024-03: [https://www.medrxiv.org/content/10.1101/2024.03.12.24303785v1 Influence of a Large Language Model on Diagnostic Reasoning: A Randomized Clinical Vignette Study] | * 2024-03: [https://www.medrxiv.org/content/10.1101/2024.03.12.24303785v1 Influence of a Large Language Model on Diagnostic Reasoning: A Randomized Clinical Vignette Study] | ||
** GPT4 improves medical practitioner work; surprisingly, GPT4 alone scored better than a human with GPT4 as aid (on selected tasks). | ** GPT4 improves medical practitioner work; surprisingly, GPT4 alone scored better than a human with GPT4 as aid (on selected tasks). | ||
* 2024-10: [https://doi.org/10.1001/jamanetworkopen.2024.38535 Perspectives on Artificial Intelligence–Generated Responses to Patient Messages] | * 2024-10: [https://doi.org/10.1001/jamanetworkopen.2024.38535 Perspectives on Artificial Intelligence–Generated Responses to Patient Messages] | ||
* 2024-10: [https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2825395 Large Language Model Influence on Diagnostic Reasoning; A Randomized Clinical Trial] | * 2024-10: [https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2825395 Large Language Model Influence on Diagnostic Reasoning; A Randomized Clinical Trial] | ||
| − | * [https:// | + | ** Use of ChatGPT does not strongly improve medical expert work; but AI alone out-scores human or human+AI |
| − | * | + | * 2024-11: [https://www.nature.com/articles/s41562-024-02046-9 Large language models surpass human experts in predicting neuroscience results] (writeup: [https://medicalxpress.com/news/2024-11-ai-neuroscience-results-human-experts.html AI can predict neuroscience study results better than human experts, study finds]) |
| + | * 2024-12: [https://www.arxiv.org/abs/2412.10849 Superhuman performance of a large language model on the reasoning tasks of a physician] | ||
| + | * 2024-12: [https://arxiv.org/abs/2412.18925 HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMs] | ||
==AI improves human work== | ==AI improves human work== | ||
| Line 82: | Line 93: | ||
* 2023-11: [https://www.nber.org/papers/w31161 Generative AI at Work] (National Bureau of Economic Research) | * 2023-11: [https://www.nber.org/papers/w31161 Generative AI at Work] (National Bureau of Economic Research) | ||
* 2023-12: [https://osf.io/hdjpk The Uneven Impact of Generative AI on Entrepreneurial Performance] ([https://doi.org/10.31219/osf.io/hdjpk doi: 10.31219/osf.io/hdjpk]) | * 2023-12: [https://osf.io/hdjpk The Uneven Impact of Generative AI on Entrepreneurial Performance] ([https://doi.org/10.31219/osf.io/hdjpk doi: 10.31219/osf.io/hdjpk]) | ||
| + | * 2023-12: [https://arxiv.org/abs/2312.05481 Artificial Intelligence in the Knowledge Economy]: Non-autonomous AI (chatbot) benefits least knowledgeable workers; autonomous agents benefit the most knowledgeable workers | ||
* 2024-07: [https://www.microsoft.com/en-us/research/publication/generative-ai-in-real-world-workplaces/ Generative AI in Real-World Workplaces: The Second Microsoft Report on AI and Productivity Research] | * 2024-07: [https://www.microsoft.com/en-us/research/publication/generative-ai-in-real-world-workplaces/ Generative AI in Real-World Workplaces: The Second Microsoft Report on AI and Productivity Research] | ||
| Line 91: | Line 103: | ||
===Forecasting=== | ===Forecasting=== | ||
* 2024-02: [https://arxiv.org/abs/2402.07862 AI-Augmented Predictions: LLM Assistants Improve Human Forecasting Accuracy] | * 2024-02: [https://arxiv.org/abs/2402.07862 AI-Augmented Predictions: LLM Assistants Improve Human Forecasting Accuracy] | ||
| + | |||
| + | ===Finance=== | ||
| + | * 2024-12: [https://dx.doi.org/10.2139/ssrn.5075727 AI, Investment Decisions, and Inequality]: Novices see improvements in investment performance, sophisticated investors see even greater improvements. | ||
| + | |||
===Creativity=== | ===Creativity=== | ||
* 2024-07: [https://www.science.org/doi/10.1126/sciadv.adn5290 Generative AI enhances individual creativity but reduces the collective diversity of novel content] | * 2024-07: [https://www.science.org/doi/10.1126/sciadv.adn5290 Generative AI enhances individual creativity but reduces the collective diversity of novel content] | ||
* 2024-08: [https://www.nature.com/articles/s41562-024-01953-1 An empirical investigation of the impact of ChatGPT on creativity] | * 2024-08: [https://www.nature.com/articles/s41562-024-01953-1 An empirical investigation of the impact of ChatGPT on creativity] | ||
| + | * 2024-10: [https://arxiv.org/abs/2410.03703 Human Creativity in the Age of LLMs] | ||
* 2024-11: [https://conference.nber.org/conf_papers/f210475.pdf Artificial Intelligence, Scientific Discovery, and Product Innovation]: diffusion model increases "innovation" (patents), boosts the best performers, but also removes some enjoyable tasks. | * 2024-11: [https://conference.nber.org/conf_papers/f210475.pdf Artificial Intelligence, Scientific Discovery, and Product Innovation]: diffusion model increases "innovation" (patents), boosts the best performers, but also removes some enjoyable tasks. | ||
| + | * 2024-12: [https://doi.org/10.1080/10400419.2024.2440691 Using AI to Generate Visual Art: Do Individual Differences in Creativity Predict AI-Assisted Art Quality?] ([https://osf.io/preprints/psyarxiv/ygzw6 preprint]): shows that more creative humans produce more creative genAI outputs | ||
===Counter loneliness=== | ===Counter loneliness=== | ||
* 2024-07: [https://arxiv.org/abs/2407.19096 AI Companions Reduce Loneliness] | * 2024-07: [https://arxiv.org/abs/2407.19096 AI Companions Reduce Loneliness] | ||
| − | ==AI passes Turing Test== | + | ==Human Perceptions of AI== |
| + | * 2023-09: [https://www.nature.com/articles/d41586-023-02980-0 AI and science: what 1,600 researchers think. A Nature survey finds that scientists are concerned, as well as excited, by the increasing use of artificial-intelligence tools in research.] | ||
| + | * 2024-11: [https://doi.org/10.1016/S2589-7500(24)00202-4 Attitudes and perceptions of medical researchers towards the use of artificial intelligence chatbots in the scientific process: an international cross-sectional survey] (Nature commentary: [https://www.nature.com/articles/s41592-024-02369-5 Quest for AI literacy]) | ||
| + | |||
| + | ===AI passes Turing Test=== | ||
| + | '''Text Dialog''' | ||
* 2023-05: [https://arxiv.org/abs/2305.20010 Human or Not? A Gamified Approach to the Turing Test] | * 2023-05: [https://arxiv.org/abs/2305.20010 Human or Not? A Gamified Approach to the Turing Test] | ||
* 2023-10: [https://arxiv.org/abs/2310.20216 Does GPT-4 pass the Turing test?] | * 2023-10: [https://arxiv.org/abs/2310.20216 Does GPT-4 pass the Turing test?] | ||
* 2024-05: [https://arxiv.org/abs/2405.08007 People cannot distinguish GPT-4 from a human in a Turing test] | * 2024-05: [https://arxiv.org/abs/2405.08007 People cannot distinguish GPT-4 from a human in a Turing test] | ||
* 2024-07: [https://arxiv.org/abs/2407.08853 GPT-4 is judged more human than humans in displaced and inverted Turing tests] | * 2024-07: [https://arxiv.org/abs/2407.08853 GPT-4 is judged more human than humans in displaced and inverted Turing tests] | ||
| + | '''Art''' | ||
| + | * 2024-11: [https://www.astralcodexten.com/p/how-did-you-do-on-the-ai-art-turing How Did You Do On The AI Art Turing Test?] Differentiation was only slightly above random (60%). AI art was often ranked higher than human-made. | ||
| + | * 2024-11: [https://doi.org/10.1038/s41598-024-76900-1 AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably] | ||
=Uptake= | =Uptake= | ||
| + | * 2023-07: [https://doi.org/10.9734/ajrcos/2023/v16i4392 ChatGPT: Early Adopters, Teething Issues and the Way Forward] | ||
* 2024-03: [https://arxiv.org/abs/2403.07183 Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews] | * 2024-03: [https://arxiv.org/abs/2403.07183 Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews] | ||
* 2024-05: Humlum, Anders and Vestergaard, Emilie, [https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4827166 The Adoption of ChatGPT]. IZA Discussion Paper No. 16992 [http://dx.doi.org/10.2139/ssrn.4827166 doi: 10.2139/ssrn.4827166] | * 2024-05: Humlum, Anders and Vestergaard, Emilie, [https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4827166 The Adoption of ChatGPT]. IZA Discussion Paper No. 16992 [http://dx.doi.org/10.2139/ssrn.4827166 doi: 10.2139/ssrn.4827166] | ||
| Line 114: | Line 141: | ||
** 72% of leaders use genAI at least once a week (c.f. 23% in 2023); 90% agree AI enhances skills (c.f. 80% in 2023) | ** 72% of leaders use genAI at least once a week (c.f. 23% in 2023); 90% agree AI enhances skills (c.f. 80% in 2023) | ||
** Spending on genAI is up 130% (most companies plan to invest going forward) | ** Spending on genAI is up 130% (most companies plan to invest going forward) | ||
| + | |||
| + | ==Usage For== | ||
| + | * 2024-12: [https://assets.anthropic.com/m/7e1ab885d1b24176/original/Clio-Privacy-Preserving-Insights-into-Real-World-AI-Use.pdf Clio: A system for privacy-preserving insights into real-world AI use] (Anthropic [https://www.anthropic.com/research/clio Clio]) | ||
=See Also= | =See Also= | ||
* [https://www.google.com/books/edition/_/cKnYEAAAQBAJ?hl=en&gbpv=1&pg=PA2 UNESCO. Guidance for Generative AI in Education and Research] | * [https://www.google.com/books/edition/_/cKnYEAAAQBAJ?hl=en&gbpv=1&pg=PA2 UNESCO. Guidance for Generative AI in Education and Research] | ||
| + | * [[AI]] | ||
Latest revision as of 12:40, 7 January 2025
Contents
AI in Education
Survey/study of
- 2023-08: Perception, performance, and detectability of conversational artificial intelligence across 32 university courses
- 2023-10: Employees secretly using AI at work.
- 2023-10: Survey shows students using AI more than professors.
- 2023-11: ChatGPT has entered the classroom: how LLMs could transform education
AI improves learning/education
- Mollick, Ethan R. and Mollick, Lilach and Bach, Natalie and Ciccarelli, LJ and Przystanski, Ben and Ravipinto, Daniel, AI Agents and Education: Simulated Practice at Scale (June 17, 2024). The Wharton School Research Paper. doi: 10.2139/ssrn.4871171
- Can enable personalized education.
- Generative AI for Programming Education: Benchmarking ChatGPT, GPT-4, and Human Tutors
- GPT4 can out-perform human tutors.
- Keppler, Samantha and Sinchaisri, Wichinpong and Snyder, Clare, Backwards Planning with Generative AI: Case Study Evidence from US K12 Teachers (August 13, 2024). doi: 10.2139/ssrn.4924786
- Teachers benefit from using AI as a co-pilot to aid in tasks (planning, how to teach topic, explore ideas).
- There is smaller utility in using AI purely as a text-generator (to make quizzes, workbooks, etc.).
- Effective and Scalable Math Support: Evidence on the Impact of an AI- Tutor on Math Achievement in Ghana
- AI Tutoring Outperforms Active Learning
AI harms learning
- A real-world test of artificial intelligence infiltration of a university examinations system: A “Turing Test” case study
- Current grading systems cannot detect AI.
- Bastani, Hamsa and Bastani, Osbert and Sungu, Alp and Ge, Haosen and Kabakcı, Özge and Mariman, Rei, Generative AI Can Harm Learning (July 15, 2024). The Wharton School Research Paper.doi: 10.2139/ssrn.4895486
- Access to ChatGPT harmed math education outcomes.
- 2024-09: AI Meets the Classroom: When Does ChatGPT Harm Learning?
Software/systems
- GPTutor (code)
- EduChat: A Large-Scale Language Model-based Chatbot System for Intelligent Education
- Eureka Labs (founded by Andrej Karpathy) aims to create AI-driven courses (first course is Intro to LLMs)
LLMs
Individual tools
- Chatbot (OpenAI ChatGPT, Anthropic Claude, Google Gemini)
- NotebookLM: Enables one to "chat with documents".
- Google Learn About
AI for grading
- Can Large Language Models Make the Grade? An Empirical Study Evaluating LLMs Ability To Mark Short Answer Questions in K-12 Education (preprint)
Detection
- Do teachers spot AI? Evaluating the detectability of AI-generated texts among student essays
- GenAI can simulate student writing in a way that teachers cannot detect.
- AI essays are assessed more positively than student-written.
- Teachers are overconfident in their source identification.
- Both novice and experienced teachers could not identify texts generated by ChatGPT vs. students
AI Text Detectors Don't Work
- 2024-05: RAID: A Shared Benchmark for Robust Evaluation of Machine-Generated Text Detectors
- 2024-06: Testing of Detection Tools for AI-Generated Text
AI/human
AI out-performs humans
Tests
- 2023-07: SciBench: Evaluating College-Level Scientific Problem-Solving Abilities of Large Language Models
- 2024-06: A real-world test of artificial intelligence infiltration of a university examinations system: A “Turing Test” case study
- AI scores higher than median students.
Creativity
- 2023-09: Best humans still outperform artificial intelligence in a creative divergent thinking task
- Best humans out-perform AI at creativity. (By implication, median humans may not.)
- 2024-02: The current state of artificial intelligence generative language models is more creative than humans on divergent thinking tasks
- 2024-02: Felin, Teppo and Holweg, Matthias, Theory Is All You Need: AI, Human Cognition, and Causal Reasoning (February 24, 2024). doi: 10.2139/ssrn.4737265
- Argues that human "theory-based" creativity is better than AI "data-based".
- 2024-07: Pron vs Prompt: Can Large Language Models already Challenge a World-Class Fiction Author at Creative Text Writing?
- Top human (professional author) out-performs GPT4.
- 2024-09: Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researchers
- LLMs can be creative
- 2024-09: Creative and Strategic Capabilities of Generative AI: Evidence from Large-Scale Experiments
Art
- 2024-11: AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably
- 2024-11: How Did You Do On The AI Art Turing Test?
Professions
- Humanity's Last Exam
- Effort to build a dataset of challenging (but resolvable) questions in specific domain areas, to act as a benchmark to test whether AIs are improving in these challenging topics.
Medical
- 2024-03: Influence of a Large Language Model on Diagnostic Reasoning: A Randomized Clinical Vignette Study
- GPT4 improves medical practitioner work; surprisingly, GPT4 alone scored better than a human with GPT4 as aid (on selected tasks).
- 2024-10: Perspectives on Artificial Intelligence–Generated Responses to Patient Messages
- 2024-10: Large Language Model Influence on Diagnostic Reasoning; A Randomized Clinical Trial
- Use of ChatGPT does not strongly improve medical expert work; but AI alone out-scores human or human+AI
- 2024-11: Large language models surpass human experts in predicting neuroscience results (writeup: AI can predict neuroscience study results better than human experts, study finds)
- 2024-12: Superhuman performance of a large language model on the reasoning tasks of a physician
- 2024-12: HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMs
AI improves human work
- 2023-07: Experimental evidence on the productivity effects of generative artificial intelligence
- 2023-09: Dell'Acqua, Fabrizio and McFowland III, Edward and Mollick, Ethan R. and Lifshitz-Assaf, Hila and Kellogg, Katherine and Rajendran, Saran and Krayer, Lisa and Candelon, François and Lakhani, Karim R., Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality (September 15, 2023). Harvard Business School Technology & Operations Mgt. Unit Working Paper No. 24-013, The Wharton School Research Paper doi: 10.2139/ssrn.4573321
- 2023-11: Generative AI at Work (National Bureau of Economic Research)
- 2023-12: The Uneven Impact of Generative AI on Entrepreneurial Performance (doi: 10.31219/osf.io/hdjpk)
- 2023-12: Artificial Intelligence in the Knowledge Economy: Non-autonomous AI (chatbot) benefits least knowledgeable workers; autonomous agents benefit the most knowledgeable workers
- 2024-07: Generative AI in Real-World Workplaces: The Second Microsoft Report on AI and Productivity Research
Coding
- 2023-02: The Impact of AI on Developer Productivity: Evidence from GitHub Copilot
- 2024-09: Cui, Zheyuan and Demirer, Mert and Jaffe, Sonia and Musolff, Leon and Peng, Sida and Salz, Tobias, The Effects of Generative AI on High Skilled Work: Evidence from Three Field Experiments with Software Developers (September 03, 2024). doi: 10.2139/ssrn.4945566
- 2024-11: Hoffmann, Manuel and Boysel, Sam and Nagle, Frank and Peng, Sida and Xu, Kevin, Generative AI and the Nature of Work (October 27, 2024). Harvard Business School Strategy Unit Working Paper No. 25-021, Harvard Business Working Paper No. No. 25-021, doi: 10.2139/ssrn.5007084
Forecasting
Finance
- 2024-12: AI, Investment Decisions, and Inequality: Novices see improvements in investment performance, sophisticated investors see even greater improvements.
Creativity
- 2024-07: Generative AI enhances individual creativity but reduces the collective diversity of novel content
- 2024-08: An empirical investigation of the impact of ChatGPT on creativity
- 2024-10: Human Creativity in the Age of LLMs
- 2024-11: Artificial Intelligence, Scientific Discovery, and Product Innovation: diffusion model increases "innovation" (patents), boosts the best performers, but also removes some enjoyable tasks.
- 2024-12: Using AI to Generate Visual Art: Do Individual Differences in Creativity Predict AI-Assisted Art Quality? (preprint): shows that more creative humans produce more creative genAI outputs
Counter loneliness
- 2024-07: AI Companions Reduce Loneliness
Human Perceptions of AI
- 2023-09: AI and science: what 1,600 researchers think. A Nature survey finds that scientists are concerned, as well as excited, by the increasing use of artificial-intelligence tools in research.
- 2024-11: Attitudes and perceptions of medical researchers towards the use of artificial intelligence chatbots in the scientific process: an international cross-sectional survey (Nature commentary: Quest for AI literacy)
AI passes Turing Test
Text Dialog
- 2023-05: Human or Not? A Gamified Approach to the Turing Test
- 2023-10: Does GPT-4 pass the Turing test?
- 2024-05: People cannot distinguish GPT-4 from a human in a Turing test
- 2024-07: GPT-4 is judged more human than humans in displaced and inverted Turing tests
Art
- 2024-11: How Did You Do On The AI Art Turing Test? Differentiation was only slightly above random (60%). AI art was often ranked higher than human-made.
- 2024-11: AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably
Uptake
- 2023-07: ChatGPT: Early Adopters, Teething Issues and the Way Forward
- 2024-03: Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews
- 2024-05: Humlum, Anders and Vestergaard, Emilie, The Adoption of ChatGPT. IZA Discussion Paper No. 16992 doi: 10.2139/ssrn.4827166
- 2024-06: Kellogg, Katherine and Lifshitz-Assaf, Hila and Randazzo, Steven and Mollick, Ethan R. and Dell'Acqua, Fabrizio and McFowland III, Edward and Candelon, Francois and Lakhani, Karim R., Don't Expect Juniors to Teach Senior Professionals to Use Generative AI: Emerging Technology Risks and Novice AI Risk Mitigation Tactics (June 03, 2024). Harvard Business School Technology & Operations Mgt. Unit Working Paper 24-074, Harvard Business Working Paper No. 24-074, The Wharton School Research Paper doi: 10.2139/ssrn.4857373
- 2024-06: Delving into ChatGPT usage in academic writing through excess vocabulary
- 2024-09: The Rapid Adoption of Generative AI
- 2024-10: Growing Up: Navigating Generative AI’s Early Years – AI Adoption Report (executive summary, full report)
- 72% of leaders use genAI at least once a week (c.f. 23% in 2023); 90% agree AI enhances skills (c.f. 80% in 2023)
- Spending on genAI is up 130% (most companies plan to invest going forward)
Usage For
- 2024-12: Clio: A system for privacy-preserving insights into real-world AI use (Anthropic Clio)
