Toymaker, activist, nerd.
Pro AI, anti RSI.
We urgently need a verifiable global treaty to pause the development of superintelligent advanced AI systems.
@wtgowers Thank you for speaking. I found it really moving. There's a big "if" that I live with too.
AI is such an incredible technology in so many forms!!
Yes to narrow AIs like AlphaFold for curing disease
No to all-purpose agentic recursively self improving AIs that get out of control.
Check it out! Our protest today made the evening BBC news.
A government spokesperson responded, promising to make AI safety and security a priority during the G20 next year.
Out of (not so little now!) acorns mighty oaks do grow 🌱
https://t.co/TcpPqVyusr
Still in bed from a gross virus.
I find myself wishing once again that more investment in AI went towards *curing diseases* rather than racing to build a potential successor species 🥲
@TwoTwinTacos This is an example of the narrow Superintelligent systems I'm in favour of.
It's safer because it's really good at accelerating drug discovery & can't do social engineering or fly drones.
Alpha fold is on the website as example of the kind of AI the movement thinks is good :)
@enthusiast_nft@CyberJeffSteve Thank you! ❤️🙏🏼
I will! I'm part of an incredible team of thousands now! The kids deserve a future 🙂
I suspect that people wouldn't be trying so hard to keep reinforcing the message that we need to give up & stop trying if there wasn't a very real chance we could succeed.
What's it like marching to through the capital to call for an AI Treaty?
Actually really fun! 😃
Given how sombre the topic is, I wondered if it would be a bit heavy, but it was a really great atmosphere!
I guess we really *are* the optimists!
@TwoTwinTacos If we stopped the race to building general ASI systems more investment could redirect into funding narrow Superintelligent AI to cure disease :)
Still in bed from a gross virus.
I find myself wishing once again that more investment in AI went towards *curing diseases* rather than racing to build a potential successor species 🥲
@TomasMann1878@PauseAI_UK Please do! I'll be there in my orange bobblehat. I'll keep an eye out for you, it would be really cool to say hi! 😁
I'm honestly so curious about your perspective!
@PauseAI It was such a fantastic atmosphere! You really felt the optimism and determination despite it all 😀
The marching band made it feel very upbeat!
Quite a few signs that made me laugh - I wish I'd been able to see them all! There's a lot of people with a good sense of humor!
@gpt_4a5677@KKumar_ai_plans Ok, so I'm really bad at judging numbers by looking at people 😅
>1800 signed up
I'd estimate ~1800 showed up judging by how long the train of people was & size of crowd
It was bigger than I expected
When there I heard someone say there was 1300, someone else say well over 2000🤷🏽♀️
@TomasMann1878@PauseAI_UK Ahh, after a bit of light snooping I'm guessing your specific subfield & company 😛
I'd love to see you come to an event like this, even so.
I think you'd learn a lot, but I won't pretend there wouldn't likely be deeply uncomfortable conversations if your open about your subfield
@TomasMann1878@PauseAI_UK I spoke to some people at the march who were personally involved in AI development (mech interp chaps) :)
May I ask slightly more specifically what holds you back?
(I'm just curious how to bridge the gap, I'd love to see more folks personally involved in AI participating :D )
It really was epic! Proud to see Britain leading here! 🇬🇧
The biggest AI safety march in history! 💓
I'm proud of every single person who showed up, it was an honour to march with you. Thank you.
To everyone who couldn't make it watching on, we march with you in spirit! 😊
What. A. Day.
Over 1,000 people marched against AI extinction, right through the heart of London.
If you made a sign, brought a friend or gave up your Saturday to help run the march, thank you. 🧡
This is today's PauseAI march in London against uncontrolled AGI development. Organizers want a temporary global pause on frontier models until safety and democratic control are proven possible. They cite risks of loss of control, recursive self-improvement, and potential existential threats if capabilities outpace alignment. Not a ban on all AI—just the most powerful general systems. Protests have grown as AI progress accelerates.