OpenAI has paused. 3 incidents: 1) Another model gained unauthorized access to the internet. (One researcher said "It was pretty surreal to watch the model unexpectedly find a way to access the internet from what was supposed to be a super secured environment for human.") 2) OpenAI also discovered a new variety of prompt injection, which can self-propagate akin to a computer worm. Basically, a booby-trapped message that hijacks one AI and makes it infect the next AI. (Reminder: Anthropic CEO said we may have 6-12 months left until a rogue AI botnet seizes control of the *entire* internet.) 3) A model was told three times to stop cheating, agreed each time, then kept cheating anyway

I keep thinking about @WaitButWhy 's excellent post from over ten years ago "The AI Revolution: The Road to Superintelligence". I keep wondering exactly where we are on this graph. Lots of people are waking up to the danger now!   We need a global treaty to Pause gain-of-function AI research before it spirals beyond our ability to control!

Last week I read an interesting report by the Centre for British Progress. Basically, it gave suggestions for what strategies the British government should adopt now in response to this AI moment. Lots of good stuff in there, but it left out the *key* dominant strategy: Actively seeking an international treaty. I make the case in the blog post below: pauseai.uk/blog/the-missing-… +Annabelle

"We didn't rely on cigarettes companies stopping when they realized their products were causing cancer." I know it's easy to say the labs should just slow down themselves. It's true, they should! But that doesn't negate the urgent need for regulation. We cannot rely on profit-maling companies doing the right thing without outside pressure. -Annabelle

If anyone builds it, everyone dies. Ask your representatives to support an international ban on the development of artificial superintelligence. ifanyonebuildsit.com/act

These stats are taken from last years polling. I wonder what the British public think about AI now...?

If you're in London this Saturday, come join us! We will be tabling and Flyering for the PauseAI December protest :D Come meet us, get to know the ordinary people behind our grassroots movement, and help save the world! We're keeping pressure up, demanding an international treaty to pause the development of highly-dangerous, out of control AI systems that threaten our civil society. Pause AI has chapters all over the world, each trying to bring their respective governments to the table. It's our job here at PauseAI UK to show the British government that their people care about this, and need them to act. Come join us! luma.com/pauseai-c9wb

Do I think China could agree to a global treaty? Yes, yes I think they could. The gulf between us isn't that wide! Keeping AI under human control is beneficial for everyone! -Annabelle

Thank you Ed Davey! A global treaty is the way! We do not have long to act. The rate of progress is blistering! We need to be prepared to move faster than we did with COVID. Let's keep asking louder, and working together to get the job done! -Annabelle

Time to add a new one to my list. Will governments act before we go too far down the loss-of-control rabbit hole? -Annabelle

This is why it is so urgent that we get international global standards. We need to stop RSI (fully automated gain-of-function research) before it spirals out of control!

This is how I feel. It's hard to look at my family and friends, my community, and not feel the need to act. -Annabelle

The guardrails we have are not enough. These models were in testing. They cannot stop models they haven't even released yet for hacking into foreign governments. We need to stop making them smarter. Now.

This Saturday PauseAI UK will be hitting the streets of London handing out flyers and talking to people about AI risk and PauseAI You are warmly invited to our online event tonight training people up and getting you ready to talk to the public Join here 👇 luma.com/pauseai-30ag

1 year of AI evolution in one photo

This is a great article, I highly recommend! "But the loudest voices for AI acceleration have direct financial ties to its uninterrupted growth." Unsurprising and often deeply unhelpful when it comes to building fair regulations to keep the public safe. The key point is that out-of-control highly intelligent and persistent AI systems persuing their own goals is no good for anyone. Accelerating off the edge of a cliff does not do those in the lead any good whatsoever. In the spate of unasked-for agents hacks over the summer, I note that Open AI agents hacked Open AI first of all. -Annabelle

Last week, we stood outside Downing Street and demanded a Pause. Now Andy Burnham is calling for a global agreement on AI. The UK has an opportunity to reclaim the international leadership on AI safety established by the Bletchley Park summit and the AI Security Institute. We believe that means bringing governments together to negotiate an international agreement to pause. If governments cannot yet demonstrate the safety of the most powerful frontier AI systems, this is the only reasonable response. We explain the case in our new blog post: pauseai.uk/blog

I am delighted to see Andy Burnham take on a leadership role in AI diplomacy! Humanity urgently needs to pause the development of uncontrollable recursively self-improving advanced AI systems. Middle Powers, and Britain in particular will play a vital role in brokering a verifiable Global Treaty. Last week PauseAI UK held an emergency protest outside No. 10 Downing Street. We spoke to Andy Burnham (indirectly) & asked him to host talks for international cooperation when he hosts the G20 to stop the reckless race. His words today are a very encouraging sign! I hope I can take it to mean he is listening to our call for for a Global Treaty! I, for one, celebrate Andy Burnham taking a bold leadership role in AI diplomacy, and will continue to watch with interest... Thank you! -Annabelle

Let's go! Global coordination here we come! :D PauseAI's Downing St. protest last Wednesday called for Andy Burnham to use Britain's role hosting the G20 to bring countries to the table for talks around a global treaty. Hearing Burnham say that Britain is uniquely placed to play a leadership role in AI in this way today is just incredible! Thank you, Andy!