Introducing ChatGPT search: OpenAI releases its search engine.

Google CEO says over 25% of new Google code is generated by AI: Their computer programmers are turning into code reviewers.

The AI-nuclear weapons analogy, explained : Vox catalogues some similarities and differences between the two technologies.

The AI-Generated Product Reviews Choking the Internet Are Now Illegal: The FTC makes certain types of fake reviews illegal, whether human or robots wrote them, in the US.

Ban LLMs Using First-Person Pronouns: Maybe we shouldn’t allow AI chatbots to refer to themselves as ‘I’.

Ed Newton-Rex, who organised the recent anti AI ingesting everyone’s work for free statement that artists of all sorts of fame levels are signing makes a good point.

There are three key resources that generative AI companies need to build AI models: people, compute, and data. They spend vast sums on the first two - sometimes a million dollars per engineer, and up to a billion dollars per model. But they expect to take the third - training data - for free.

Big tech AI companies aren’t stingy with their money for everything. That’s one reason why they’re so unprofitable. It’s just that magic third ingredient that often attracts the $0 compensation rate.

Thom Yorke and Julianne Moore join thousands of creatives in AI warning: Even superstars are concerned about AI companies shovelling up the artistic output of humanity without constraint or compensation.

All What Is Delicious to Man: Today’s implausible vision of superabundance is a lot grimmer than the one that they were promised in the 1830s.

The Subprime AI Crisis: Ed Zitron worries about what happens when the unprofitability of the current generative AI business becomes unsustainable.

A Personal Take on Using LLMs: Eleanor Konik’s view on when she finds it reasonable to use large language models.

Introducing canvas: A new interface for OpenAI’s ChatGPT that’s designed for better “collaboration” when coding or writing.

Using LLMs to Simplify your Writing: Using LLMs as a personal writing editor seems like one of their more benign uses if you’re happy to fact-check the results - and this idea is one my colleagues probably wish I’d take up.

Ban warnings fly as users dare to probe the “thoughts” of OpenAI’s latest model: OpenAI’s typical levels of actual openness surfacing again as they threaten to ban anyone asking their new model why it says the things it does.