iiinova — where your digital self keeps evolving

Reposted by Andrew West

bryan newbold @bnewbold.net · 07/10/2026

ah damn, RIP Margaret Hamilton. she took safety-critical software engineering seriously, and got people to the moon and back. news.mit.edu/2026/margare...

old photos of a woman (Margaret Hamilton) standing next to a tall stack of books (printouts of Apollo space mission software source code)

102144004735

Reposted by Andrew West

Ethan Mollick @emollick.bsky.social · 07/10/2026

Some early first-hand accounts of the experience of encountering a narrow superhuman intelligence as mathematicians grapple with the hundreds of big AI proofs released by OpenAI. Problems solved in inhuman ways that make us wonder what it means to actually know things... scottaaronson.blog?p=10169

410115

Andrew West @awest.uk · 07/10/2026

It was such a surprise, given how much they bigged up the script. Especially when the source material was so good.

010

Andrew West @awest.uk · 30/09/2026

Very important to watch yesterday’s big speech so you know what’s going to happen to the UK’s productivity figures in the next year. Nice of OpenAI not to clash with whatever Andy Burnham was saying too.

000

Reposted by Andrew West

conputer dipshit @davidcrespo.bsky.social · 30/09/2026

good to hear. I think it’s clear that if you know how to learn and want to learn, AI will not impede you and will most likely help. what I don’t know is what proportion of people are like that. people talk about college kids like they have to be forced to learn, or tricked into it. isn’t that weird?

2323

Andrew West @awest.uk · 29/09/2026

Fwiw, the opposite argument is: —Anthropic were founded on worrying about the risks —Anthropic think that someone else will do it worse —Anthropic have called for regulation to slow them and others Obv there are financial incentives and I understand why people focus on those. But the tech exists

010

Andrew West @awest.uk · 29/09/2026

Yeah they’ve really botched it :-(

000

Reposted by Andrew West

Zach Weinersmith @zachweinersmith.bsky.social · 28/09/2026

I don't understand the need to say dickwad CEOs can't make good stuff. This is the world we live in. A lot of my favorite things in the arts and sciences come from absolute bastards. The real argument should be about how to channel dickwaddery into good social outcomes, not whether the stuff works.

3533839

Andrew West @awest.uk · 19/09/2026

If you want to get angry, get angry that some org said ‘here’s a sexy AI actress you can put on tv and ask questions’ and uk media fell for it instantly

000

Andrew West @awest.uk · 19/09/2026

There is no AI council directing all use of AI. Someone using AI to make an AI actress does not tell you anything useful.

110

Andrew West @awest.uk · 19/09/2026

The attention being paid to Tilly Norwood is like me turning up to your economics class and saying ‘you guys assume everyone is a rational actor and that’s just not true lol’

100

Andrew West @awest.uk · 16/09/2026

HPMOR seems to have been pretty effective in getting people interested in AI safety early. Scott Alexander:

210

Andrew West @awest.uk · 15/09/2026

It’s because Google make the very weird decision to put a weak AI at the top of Google searches. The latest versions of ChatGPT and Claude are incomparably more intelligent, which is where the worry is coming from.

100

Reposted by Andrew West

Alex Hern @hern.bsky.social · 13/09/2026

So many people with aviation psychosis claiming that the Wrights’ invention is going to reshape global geography when it can’t even go more than 25 yards

1125331

Andrew West @awest.uk · 13/09/2026

I don’t think asking a question on a public thread is sea lioning…? Even if I have misunderstood what you were saying. But I am sorry to have made your day worse.

100

Andrew West @awest.uk · 13/09/2026

Ok got it. I think we have come to different conclusions from the available evidence, but I understand where you’re coming from. Thanks for clarifying.

000

Andrew West @awest.uk · 13/09/2026

Ok that’s not a very nice thing to say and I don’t think the conversation justifies it. She replied to me first! But ok I will make an exit.

220

Andrew West @awest.uk · 13/09/2026

Ok got it - thanks for clarifying!

000

Andrew West @awest.uk · 13/09/2026

Ok got it. So I don’t think I was saying anything about underlying intelligence. Is that what the original request was for? Empiral evidence of intelligence as opposed to the ability to generate proofs?

200

Andrew West @awest.uk · 13/09/2026

I am genuinely trying to talk like a human being here, I promise. The answer would be that no, your printer has not generated new mathematical proofs verified in lean. The LLM, as far as I can tell, has. So what am I getting wrong?

220

Andrew West @awest.uk · 13/09/2026

I assume we have different understandings of ‘verifiable’ and ‘empirical’ then…? How are these terms being used here? I mean by it ‘generating a proof that works in Lean is good evidence that you have made an achievement in maths’, I think. What am I misunderstanding?

200

Andrew West @awest.uk · 13/09/2026

And the fuck-ai people are making the information environment really unpleasant.

000

Andrew West @awest.uk · 13/09/2026

People are going to start losing their jobs to AI. We urgently need ways to deal with this, but the policy wonks still think it’s sci-fi and the academic commentators who they lean on are still telling everyone the tech is all hype, for some goddamn reason. It’s a bad situation!

220

Andrew West @awest.uk · 13/09/2026

They absolutely should, yes. Every decent person in the middle should be working together to find ways to deal with this - with one eye on how we have failed massively in the past. The tech is here and this needs urgent attention. But the extreme viewpoints suck all the air out of the room.

030

Andrew West @awest.uk · 13/09/2026

Muting the word ‘billionaire’ is very effective, I recommend

000

Andrew West @awest.uk · 13/09/2026

Verifiable maths results are kinda the definition of empirical evidence, no?

100

Andrew West @awest.uk · 13/09/2026

I think it’s very easy to slip into the comfort zone of lefty critique of big business. Everyone knows what to say and think and who the teams are. It’s not that there’s no value there, but meanwhile there is a novel technology developing rapidly, with ramifications that need serious attention.

000

Reposted by Andrew West

James O'Malley @jamesomalley.co.uk · 10/09/2026

The Guardian’s data centre reporting is needlessly hysterical (FREE TO READ!) takes.jamesomalley.co.uk/p/the-guardi...

takes.jamesomalley.co.uk

The Guardian’s data centre reporting is needlessly hysterical

Less talk, more context.

83715

Andrew West @awest.uk · 08/09/2026

Ok cool. I have no problem with worrying about corporate behaviour. I just wanted to push back against the idea that it’s not real (which wasn’t you). AI capabilities are a massive opportunity+concern and we need all hands on deck to deal with the reality. The bluesky set would be a boon.

110

Andrew West @awest.uk · 08/09/2026

So is your serious contention that this was all human work that OpenAI is trying to claim credit for? Is that what you genuinely think is happening here?

101

Andrew West @awest.uk · 08/09/2026

They are very strongly denying this fwiw, and say their solution is different

110

Andrew West @awest.uk · 04/09/2026

There was a time I thought viewing everything through the lens of power dynamics was insightful. Now I think it’s blinkered, ignores the hard work, and makes all coverage the same. It’s a brilliant meme because it makes everyone involved feel clever, and all analysis hard enough to feel like work.

100

Andrew West @awest.uk · 02/09/2026

Hard to see what BBC breaking news notifications are achieving at this point

000

Andrew West @awest.uk · 31/08/2026

Neal Stephenson is a good exception - he seems outright intrigued

000

Reposted by Andrew West

James O'Malley @jamesomalley.co.uk · 28/08/2026

Any MP who signs on to Burgon's motion is not acting within the national interest. We need more data centres – and it comes down to a question of security: takes.jamesomalley.co.uk/p/the-case-f...

takes.jamesomalley.co.uk

The case for Nightingale Datacentres

Britain isn’t prepared for the looming AI national security crisis

104517

Reposted by Andrew West

Sarah Jones @sarahjones.bsky.social · 25/08/2026

The Imagination Library is a truly wonderful thing, a beautiful idea that meant so much to so many people at home and all over the world

121237

Reposted by Andrew West

FT @fucktheory.bsky.social · 25/08/2026

Imagine being talented and beautiful and rich and famous and still the most noteworthy thing about you is how good a person you are.

99268016331

Reposted by Andrew West

James Ball @jamesrball.com · 22/08/2026

Gonna briefly do an AI take: it used to be possible to have interesting discussions about AI here, and it's got progressively less so. You can't just think AI is bad, or overhyped, you've also got to subscribe to a very particular worst-case scenario about *how* it's bad and how bad it is.

14146041

Andrew West @awest.uk · 14/08/2026

Last August was dry, and I had to stay up late one night monitoring my parents’ neighbour burning his scrub in an impromptu bonfire. Nice line of trees between me and him. Little jolt of adrenaline every time a spark went their way. Some people are just not that bright and don’t watch the news.

000

Andrew West @awest.uk · 14/08/2026

So that really triggered the mad libertarians huh

000

Andrew West @awest.uk · 14/08/2026

Everyone knows that one day something awful will happen and they’ll be banned from public sale. Dunno why we have to wait till then.

000

Reposted by Andrew West

Victor Geislinger @victorsvector.com · 13/08/2026

Been testing this 'sign-language-to-text' (SL2T) for almost a year & it's so awesome to see this out! Looked at this almost a decade ago & knew we'd get the tech & ML models there eventually (but super challenging). Talking w/ some Deaf folks, an equivalent of 'voice-to-text' was huge for them

deepmind.google

Putting sign language AI into users’ hands

Introducing sign-language-to-text (SL2T), our breakthrough model powering new sign language features for Deaf and hard of hearing users.

210533

Andrew West @awest.uk · 12/08/2026

Just walked to a more westerly window in order to be closer to the sun

000

Reposted by Andrew West

Saloni @scientificdiscovery.dev · 11/08/2026

BIG NEW piece by me. A few years ago, if you asked me to summarize my entire worldview in one article, I would’ve said it was too broad and too difficult. But this year, I decided to do it.

worksinprogress.co

Every disease is a policy failure

Few people know how much medicine has progressed in their lifetimes. Fewer know how much faster it could be progressing.

1327189

Andrew West @awest.uk · 10/08/2026

I get that the sirens’ song is inherently not composable. But Christopher Nolan has infinite money and any artist in the world would take his phone call. Maybe try *something*. Hell, get AI to invent a musical psychology and compose something that induces a compulsion oh no wait

020

Reposted by Andrew West

Zach Weinersmith @zachweinersmith.bsky.social · 01/08/2026

There's lots of stuff LLMs can't do, but if you want to disbelieve all the recent math, you have to believe a bunch of Fields Medalists with long track records are just straightforwardly lying about the thing they care most about. Here's a recent blog by Gowers: gowers.wordpress.com/2026/05/08/a...

gowers.wordpress.com

A recent experience with ChatGPT 5.5 Pro

We are all having to keep revising upwards our assessments of the mathematical capabilities of large language models. I have just made a fairly large revision as a result of ChatGPT 5.5 Pro, to whi…

2348591

Andrew West @awest.uk · 01/08/2026

A lot of people understandably see it as a threat, and I think they gish gallop themselves as a coping mechanism. It doesn’t work and billionaires and techbros and copyright and it’ll ruin the children and stochastic parrot and Gary Marcus and it’s a bubble and the economics don’t work and and and

1131

Reposted by Andrew West

Simon Willison @simonwillison.net · 23/07/2026

Tucked away in this article is an appeal to the AI skeptics to PLEASE stop writing off stories like this OpenAI accidental exploit of Hugging Face as a dishonest marketing trick Frontier models can find and exploit vulnerabilities now, it helps nobody to pretend that they can't!

Resist the temptation to write this off as a stunt #

There will inevitably be some people who dismiss this story as a dishonest marketing trick by OpenAI to make their models sound terrifyingly effective. I found 81 instances of the term “marketing” in the Hacker News discussion of the incident.

To those people I say pull your heads out of the sand—you’re now including Hugging Face in your conspiracy theories, just so you can deny the crescendo of evidence here!

The best models we have today have the ability to both find and exploit new vulnerabilities. The ExploitGym paper itself concludes that “autonomous exploit development by frontier AI agents is no longer a hypothetical capability”, and this incident is a perfect example of exactly that.

108910