OpenAI releases a demo of ChatGPT, a chatbot version of GPT-3 that answers follow-up questions, admits its mistakes, challenges incorrect premises, and more (Will Douglas Heaven/MIT Technology Review)

Will Douglas Heaven / MIT Technology Review:
Meta AI and Papers with Code pull Galactica, a large language model to generate scientific text, days after its launch, amid criticism about incorrect results  —  Galactica was supposed to help scientists.  Instead, it mindlessly spat out biased and incorrect nonsense.

Will Douglas Heaven / MIT Technology Review:
Q&A with OpenAI CEO Sam Altman on why DALL-E 2 made such an impact, the lessons from its success about AI’s next decade, what DALL-E means for society, and more  —  Three things the groundbreaking generative model taught OpenAI’s CEO.  —  Sam Altman, OpenAI’s CEO …

Will Douglas Heaven / MIT Technology Review:
A look at OpenAI’s Minecraft bot, which has learned to complete complex tasks in Minecraft after being trained on 70K hours of video of people playing the game  —  Online videos are a vast and untapped source of training data—and OpenAI says it has a new way to use it.

James Vincent / The Verge:
Stack Overflow temporarily bans users from sharing responses generated by ChatGPT, pending a ruling in the future; most answers are wrong on closer inspection  —  Stack Overflow, the go-to question-and-answer site for coders and programmers, has temporarily banned users from sharing responses generated by AI chatbot ChatGPT.

Melissa Heikkilä / MIT Technology Review:
A look at the privacy concerns raised by OpenAI’s GPT-3, Google’s LaMDA, Meta’s OPT-175B, and other large language models trained on troves of personal data  —  Large language models are trained on troves of personal data hoovered from the internet.  So I wanted to know: What does it have on me?

Melissa Heikkilä / MIT Technology Review:
Alphabet’s DeepMind unveils Sparrow, an AI chatbot that learns from human feedback and searches Google for information to support its claims  —  The lab trained a chatbot to learn from human feedback and search the internet for information to support its claims.