51¶¯Âþ

Letter: Learning to think differently about AI

Published 20 May 2026

From Julia Stavenhagen, Hammond, Oregon, US

I was completely prepared to hate Jacob Aron’s column about AI, since I have been dismally disappointed by the motivations of the companies that have trained models like ChatGPT with distressing biases towards “popular” answers that may decrease their accuracy. Aron clearly explained why AI models do this, and how he was able to focus his AI on creating useful tools using simple instructions to the model to return only factual information. Now, if only we could instruct the data centres not to use any water or create heat islands (14 March, p 19).

Issue no. 3596 published 23 May 2026

Sign up to our weekly newsletter

Receive a weekly dose of discovery in your inbox. We'll also keep you up to date with 51¶¯Âþ events and special offers.

Sign up
Piano Exit Overlay Banner Mobile Piano Exit Overlay Banner Desktop