A list of resources, articles, and opinions related to big language models & robotics

Teresa Berndson / Better AI Drawing / Letter Word Text Taxonomy / Licensed under CC-BY 4.0.

We’ve put together several articles, opinions, videos, and resources related to the big language model. Some of these links also cover other generative models. We will update this list regularly to add more interesting resources.

How do they work

• What is a Generative AI model?, Kate Soulevideo from IBM Technologies.
• What is GPT-4 and how is it different from ChatGPT?, Alex HernGuard.
• What Does ChatGPT Do… and Why Does It Work?, Stephen Wolfram.
• Understanding the Big Language Model — Transformative Reading Lists, Sebastian Raska.
• How ChatGPT is Trainedvideos by Ari Seff.
• ChatGPT – what is it? How does it work? Should we be excited? Or scared?Deep DhillonRadical AI Podcast.

Journal articles, conferences and arXiv

• Scientist’s Perspective on the Potential of Generative AI in Their Field, Meredith Ringel MorrisarXiv.
• LaMDA: A Language Model for Dialog Applications, Romal Thopilan et alarXiv.
• What Language Model to Train if You Had One Million GPU Hours?, Teven Le Scao et al.arXiv.
• Alpaca: A Strong, Followable Instruction Model, Rohan Taori et al.
• Process for Adapting Language to Society Models (PALMS) to Value Targeted Data Sets, Irene Solaiman, Christy DennisonNeuroIPS 2021.
• About the Dangers of Stochastic Parrots: Can Language Models Get Too Big? , Emily Bender, Timnit Gebru, Angelina McMillan-Major, Shmargaret Shmitchell2021 FACTS.
• Big Language Model Survey, Wayne Xin Zhao et al.arXiv.
• Watermark for Large Language Model, John Kirchenbauer, Jonas Geiping, Yuxin Wen, Jonathan Katz, Ian Miers, Tom GoldsteinarXiv.
• Between Subjectivity and Encumbrance: Power Dynamics in Data Annotation for Computer Vision, Milagros Miceli, Martin Schuessler, Tianling YangACM Proceedings on Human-Computer Interaction.
• AI classifier to show AI written text, OpenAI.
• Pythia: Suite for Analyzing Large Language Models Across Training and Scaling, Stella Biderman et alarXiv.
• GPT-4 Technical Report, OpenAIarXiv.
• GPT-4 System Card, OpenAI.
• BloombergGPT: The Big Language Model for Finance, Shijie Wu et alarXiv.

Newspapers, magazines, University websites and blogpost articles

• Why an exam aimed at humans may not be a good benchmark for an LLM like GPT-4, Ben DicksonBeat Effort.
• Does GPT-4 Really Understand What We Say?, David KrakauerNautilus.
• The large language model is biased. Can logic help save them?, Rachel GordonMIT News.
• Ecosystem graphs for ML models and relationships, researcher at Stanford University.
• ChatGPT struggles with the Wordle conundrum, which shows a lot about how it works, Michael G MaddenConversation.
• AIhub coffee corner: The big language model for scholarly writingAIhub.
• ChatGPT Is an Opaque Web JPEG, Ted ChiangNew Yorkers.
• ChatGPT, Galactica, and Progress Trap, Abeba Birhane and Deborah RajiWired.
• ChatGPT can’t lie to you, but you still shouldn’t trust it, Mackenzie GrahamConversation.
• AI information retrieval: A search engine researcher explains the promise and dangers of letting ChatGPT and its cousins ​​search the web for you, Chirag ShahConversation.
• A small step for research but a big leap for usabilityInterview with Fredrick HeintzLinköping University.
• ChatGPT threatens language diversity. More needs to be done to protect our differences in the age of AI, Collin BjorkConversation.
• Column: Afraid of AI? The startup selling it wants you, Trader BrianLos Angeles Times.
• The three ways an AI chatbot is a security disaster, Melissa HeikkiläMIT Technology Overview.
• Time: OpenAI Uses Kenyan Workers for Less Than $2 Per Hour to Make ChatGPT Less Toxic, Billy PerigoTIME.
• Misplaced fear of ‘evil’ ChatGPT obscures the real harm being done, John NaughtonGuard.
• Darktrace warns of an increase in AI-enhanced fraud since the release of ChatGPT, Mark SweneyGuard.
• Lawmakers struggle to differentiate between AI and human email, Kate BlackwoodKornell Chronicle.
• The Colombian judge said he used ChatGPT in the verdict, Luke TaylorGuard.
• Bhashini: At your service, an Indian speaking chatbot powered by ChatGPTvideos from The Economic Times.

Podcasts and video discussions

• Limitations of ChatGPT with Emily M. Bender and Casey FieslerRadical AI Podcast.
• CLAIRE AQuA: “ChatGPT and the Big Language Model”CLAIRE.
• Su Lin Blodgett on Making Language Only TechnologiesThe Good Robots Podcast.

Focus on LLM and robotics

• ChatGPT for Robotics: Design Principles and Model CapabilitiesMicrosoft.
• OpenAI and Figure join the race to become humanoid robot workers, Loz BlaineNew Atlas.
• Inner Monologue: Reasoning Embodied through Planning with Language Models, Wenlong Huang et al.arXiv.
• PaLM-E: An embodied multimodal language model, Danny DriessGoogle.
• Consciousness, Embodiment, Language Models (with Professor Murray Shanahan)YouTube videos from Machine Learning Street Talk.

Focus on LLM and education

• Opinion: ChatGPT – what does it mean for academic integrity?, Giselle ByrnesMassey University.
• Debate: ChatGPT offers an invisible opportunity to sharpen students’ critical skills, Erika Darics, Lotte van PoppelConversation.
• ChatGPT and cheating: 5 ways to change the way students score, Louis Volante, Christopher DeLuca Don A. KlingerConversation.
• ChatGPT: students may use AI to cheat, but it’s an opportunity to rethink grading as a whole, Sam IllingworthConversation.
• Guru Quick Guide to ChatGPT, @herftedukator.

Relating to art and other creative processes

• ‘ChatGPT says I don’t exist’: how artists and writers fight AI, Vanessa ThorpeGuard.
• AI and the future of work: 5 experts on what ChatGPT, DALL-E and other AI tools mean for artists and knowledge workers, Lynne Parker, Casey Greene, Daniel Coin, Kentaro Toyama, Mark FinlaysonConversation.
• Is there a way to pay creators whose work is used to train the AI? Yes, but it’s not easy, Brendan Paul MurphyConversation.
• ChatGPT is a push for higher education to rethink assessment, Sioux McKenna, Dan Dixon, Daniel Oppenheimer, Margaret Blackie, Sam IllingworthConversation.
• AI Art: How artists use and deal with machine learningYouTube videos from the Museum of Modern Art.

Misinformation, fake news and their impact on journalism

• Misinformation Monitor: March 2023, focus on GPT-4News Keeper.
• Fake news craze: why ChatGPT could be disastrous for truth in journalism, Emily BelGuard.
• Defending Against Neural Fake News, Rowan Zellers et alarXiv.

Regulations and policies

• ‘Political propaganda’: China restricts access to ChatGPT, Helen DavidsonGuard.
• Chatbots, deepfakes and voice clones: AI scams for saleUS Federal Trade Commission blog post.

Source link