Does the public need to be informed about the links between brains and AI?

An open letter to the connectionists mailing list; submitted and then rejected by a moderator, on January 21.

[The connectionists mailing list is a moderated mailing list for discussion of neural networks and cognitive or computational neuroscience.]

Dear Connectionists,

I'm a writer, science journalist, and ex-physicist published at Quanta Magazine, Scientific American, New Scientist, and other outlets. I'm contacting the mailing list to share a book project I've been working on that I hope you might find of interest.

The idea of the book is to investigate the remarkable evidence that has been emerging from neuroscience research, over the last decade or so, that both neuroscientists and AI scientists have discovered the keys to building simulations of brain regions, using deep neural networks. Moreover, I seek to investigate whether modern commercial AI programs—like the company OpenAI's ChatGPT—may be best interpreted from this perspective, as combinations of artificial brain cortexes; thereby providing a critical way to understand what they are, their strengths and weaknesses, how they should be regulated, and so on.

The chief purpose of the book will be to make the evidence accessible to non-experts, and also, to draw out a greater conversation between experts and the public. Because as I studied these developments as a science journalist, it stuck out to me that even if neuroscientists were understandably still on the fence about the evidence, then it is at least strong enough that its potential implications demand to be shared with the public.

Or are we really just going to leave the public largely uninformed about all this? The disconcerting possibility that a disembodied brain technology is becoming widely commercialized and distributed, and that this is going almost entirely unconsidered, unquestioned, and unregulated? In short, are we being too cautious about sharing the links between brains and AI with the public?

Last Wednesday, I released a 45-page proof-of-concept for the book, as well as a Kickstarter project aiming to raise funds, over the next 24 days, for me to complete the book project. You can back the project by purchasing a preorder of the completed book for $10, and/or by helping me spread the word about it. I'd be immensely grateful, because getting the project funded will depend critically on it generating word of mouth interest.

I'd also greatly appreciate discussion, critiques, objections, questions, and so on. Many thanks for your consideration.

Best wishes,

-Mordechai Rorvig