The State of the Art in ML - Aug 29th, 2021

[The Science Wiki]

OpenAI Codex AI21 Labs trains a massive language model to rival OpenAI’s GPT-3 Deleting unethical data sets isn’t good enough AI system may diagnose dementia in a day - BBC...

The Machine Learning Wiki Unsubscribe from all emails Openai.com OpenAI Codex We’ve created an improved version of OpenAI Codex, our AI system that translates natural language to code, and we are releasing it through our... Also by Vice.com and Venturebeat.com Sam Altman @sama Today OpenAI launched Codex, which is an API that uses AI to write code from natural language. There's a demo video here: https://t.co/hOOfo3MZlu OpenAI @OpenAI OpenAI Codex has a qualitatively new capability — it can write code with sufficient accuracy that users can direct their computer in natural language. Codex is now in OpenAI API, for people to build new businesses or integrate with existing applications: https://t.co/n3U4DJ7fqC https://t.co/3ETsyXMQ7T Venturebeat.com AI21 Labs trains a massive language model to rival OpenAI’s GPT-3 AI21 Labs, a Tel Aviv, Israel-based startup, has trained a massive English-language model it claims is superior to GPT-3 in some ways. Also by Venturebeat.com and Ai21.com Carl Carrie (@🏠) @carlcarrie The startup AI21 Labs trains a massive language model supporting multi-words / n-grams - Jurassic-1 Jumbo — contains 178 billion parameters, or 3 billion more than OpenSIs GPT-3 (but less than than PanGu-Alpha, HyperCLOVA, and Wu Dao 2.0). #NLP https://t.co/3MI59GEiu9 Technologyreview.com Deleting unethical data sets isn’t good enough This is a really important paper, because people s eyes have not generally been open to the complexities, and potential harms and risks, of... Also by Venturebeat.com, Skynettoday.com, Theregister.com, Medium.com, and Fastcompany.com Karen Hao @_KarenHao As norms have changed in the AI community, data sets once considered inventive and foundational are now seen as unethical breaches of privacy. But researchers who've attempted to delete the data have faced a problem: the internet is forever. 😬https://t.co/rYQt5hBeWE Bbc.com AI system may diagnose dementia in a day - BBC News The system being tested should spot the disease long before symptoms or obvious brain signs develop. Also by Theguardian.com, Thehill.com, Indiatoday.in, and Newatlas.com Techxplore.com An AI expert explains why it's hard to give computers common sense Imagine you're having friends over for lunch and plan to order a pepperoni pizza. You recall Amy mentioning that Susie had stopped eating meat.... Also by Theconversation.com and Eetimes.com Quantamagazine.org Computer Scientists Discover Limits of Major Research Algorithm | Quanta Magazine The most widely used technique for finding the largest or smallest values of a math function turns out to be a fundamentally difficult computational... hardmaru @hardmaru Computer Scientists Discover Limits of Major Research Algorithm: A new result in complexity theory establishes why the gradient descent algorithm cannot solve some kinds of problems quickly. (@QuantaMagazine) https://t.co/yXgoNp5TdZ https://t.co/8RqT8RSKWS More  → This email was sent to hul-27...@googlegroups.com. Our mailing address is: The Science Wiki 237 Kearny St 426 San Francisco, CA 94108 Terms  ·  Privacy  ·  Change Email Address  ·  Unsubscribe  ·  Get in touch with us

[The Science Wiki]