Andrew Hider

It's really important that clinicians are involved as generative #AI inevitably gets deployed in mental health services. It's important that we have a frame to think about safety as this starts to happen. Only clinicians can build this frame. psyarxiv.com/wpe2c/?s=09

1. 🚨 Finally I get to share my new paper: "A Disproof of LLM Consciousness." I show that *no* falsifiable and non-trivial theories of consciousness could ever work for LLMs. Intriguingly, turns out that cracking continual learning might change this web3.arxiv.org/pdf/2512.12802

New paper!! Systematic review of prediction models for self-harm and suicide. Largest of its kind. Provides a comprehensive overview of 196 models and external validations.

@emollick Thanks. We reference that Wolfram text in this recent paper re use of AI in mental health settings, precisely because of the points you make. cambridge.org/core/journals/…

Yes. yes LLMs are not just matrix multiplication but adding that there are non-linear functions as well doesn't really do anything to resolve the central mystery of why these models can do what they do. And here is the source of Wolfram's paragraph: writings.stephenwolfram.com/2023/02/what-i…

We really have not made a lot of progress on explaining the deep mystery of LLMs: How does a model using matrix multiplication to predict the next word manage to simulate human thought well enough to do all the very human-like things it does? And what does that mean about us?

Claim: gpt-5-pro can prove new interesting mathematics. Proof: I took a convex optimization paper with a clean open problem in it and asked gpt-5-pro to work on it. It proved a better bound than what is in the paper, and I checked the proof it's correct. Details below.

@DrJohnABaker @DoSho Cheers John for the feedback. Yes - huge potential to improve safety if used responsibly. Risks both ways (AI and human) need to be managed.

@ahider @DoSho Really interesting paper, there are probably huge gains for mental health organisations which sensibly use LLM to enhance quality and safety of services.

On the same day that Illinois steps in to regulate mental health AI, a complete coincidence that @DoSho and I have a paper on mental health AI safety published in BJPsych Bulletin Open Access. #AI #alignment cambridge.org/core/journals/…

@MarkLRuffalo Kramer’s “Listening to Prozac” and Gilbert’s “ Depression - the evolution of powerlessness” came out at almost the same time in the early 90s. The primary clinical praxis of the former prevailed, although both integrate multiple perspectives on aetiology.

Thoughts about Christmas stories from Glasgow SW Foodbank. And a plea for help in tough times. facebook.com/share/p/15XUfX… Please share @kevinbridges86 (you get a mention) @andrewrobertso5 @jk_rowling @LewisCapaldi @PaoloNutini @SFoodbanks @FoundationCFC @RFC_Charity

All of us at Iris Care Group would like to say a huge congratulations to Andrew Debnam, who won the Carer in the Home Award at the 2024 South Wales Health and Care Awards last night! 🏆🎉 #SWHealthAwards #SocialCare

Helpful new @OECD report on AI and the clinical workforce. oecd.org/en/publication…

The policy question in health and social care: the balance between centralised and decentralised systems. researchgate.net/publication/33…

@MarkLRuffalo I'm looking for peer reviewers for a aper I've written about the computational instantiation of that exact problem, if you're interested? x.com/jmirpub/status…

Those who attended our workshops at #RRNConf24 will be getting a thematic summary of the interactive collective intelligence exercise so that your great ideas don't get lost in the wind! Great to be around people who get that people are not numbers and organisations are not data.

Collective stupidity sounds like a negative concept, but it's actually helpful to think about how humans in groups foul things up, to plan how to engineer out system risk. geoffmulgan.com/post/a-theory-…

Logic, maths, AI and common sense.... writings.stephenwolfram.com/2023/09/rememb…

We're delighted to be joined by @LizDurrant19, speaking about the importance of creating a culture in which people feel they belong and an environment for the people we support and our staff to flourish. #RRNConf24

Next to take the stage is Alexis Quinn, RRN Manager, speaking about cultural restraint and asking the question, “What is a culture of care?” #RRNConf24

One of my heroes @AQuinnUnbroken @theRRNetwork "Why do seemingly caring staff behave in ways that cause harm?" #RRNConf24

@DrWinarick @euthymiatech Yep. The safety and effectiveness of both being constantly critically appraised.

@ahider @euthymiatech It’s two different services - ai therapy and real therapy for the exact reasons you specify.

Good to see conversations happening about the nuts and bolts of AI disruption in clinical psychology and psychiatry. Patient safety has to be prime. Our minds are incredibly robust but also at times painfully fragile. Machines are incredibly helpful but currently, they don’t care