u3a

Macclesfield

Sub-group Meeting 4th November 2025

A number of interesting papers from Communications of the ACM.

A Blog paper on the the effects of generative content on the output of generative models:

How Generative Models are ruining themselvesDownload

A news item about gaps in Large Language Models:

Language models wrestle with gaps in understandingDownload

Language Models can constitute a 'System 0' of thinking:

Can AI expand the human mindDownload

And something challenging from `the Guardian today (28/10//2025):

AI psychosis is a growing danger. ChatGPT is moving in the wrong direction | Amandeep Jutla | The GuardianDownload