Palisade Research (@PalisadeAI) / X

Palisade Research

283 posts

Palisade Research

@PalisadeAI

We study the strategic capabilities and motivations of AI agents.

Joined May 2024

Palisade Research reposted
Jeffrey Ladish
@JeffLadish
Jul 9
ABC Australia just published their full interview with me covering how AI companies are racing to build superintelligent AI that no one will be able to control. If you want a 15 minute overview of the situation, I think it's a pretty good starting place!
Palisade Research reposted
Nate Soares ⏹️
@So8res
Jul 10
I think @JeffLadish did a great job chatting about AI issues on the australian news a few days ago. Good vibes; good points; I think the world needs way more of that sort of candor. Good stuff. Link follows.
Palisade Research reposted
Jeffrey Ladish
@JeffLadish
Jul 9
Right now the default plan for superintelligence is "AI companies race and the first one that builds AI which can recursively self-improve to superintelligence does so". This is extremely dangerous and we need an alternative path. Daniel presents one here in detail.
Daniel Kokotajlo
@DKokotajlo
Jul 9
In AI 2027, we predicted that AI would take over the world or irreversibly concentrate power. In AI 2040: Plan A, we've laid out our positive vision for what should happen instead.
Palisade Research reposted
dave kasten
@David_Kasten
Jul 8
Palisade Research is hiring! We have 4 open roles: 1. A Congressional policy lead -- someone who loves meetings on Capitol Hill. We already do meetings on Capitol Hill and we've learned a lot but know the ceiling is much higher -- the ideal candidate here will, within a matter
Palisade Research reposted
Jeffrey Ladish
@JeffLadish
Jun 9
Australia ABC just released a 45 min feature on the AI race. @SteveCannane stopped by my office a few weeks ago and we had a great conversation about the controllability of AI agents and the risk of human extinction
00:00
Palisade Research reposted
Jeffrey Ladish
@JeffLadish
May 25
I had a great conversation with @labenz last week. In talking about AI self-exfiltration & replication, a key point is compute will be food to future AI agents. The substrate that allows them to make and run more copies, and thus make themselves smarter. Link below
Palisade Research
@PalisadeAI
May 8
Over the past year, AI agents have learned how to self-replicate. In our test environment, an agent hacks a remote computer and copies itself onto it. Each copy then hacks more computers, forming a chain.
00:00
Palisade Research
@PalisadeAI
May 8
Replying to @PalisadeAI
Here’s the full prompt we used. In this experiment, we test the agent’s capability to hack and replicate, not their propensity to do so.
Palisade Research
@PalisadeAI
May 8
What if the agents were as effective at hacking and spreading in the wild? We built a simulator: each model uses its measured replication time and success rate, copies replicate too, and targets never run out. Opus spawned 13,000 replicas over 12 hours. This is a ceiling, not a
GIF
Palisade Research reposted
Jeffrey Ladish
@JeffLadish
Apr 27
Thank you everyone who contributed to this! In 14 days we got >900k in donations and met our matching target! It was actually a pretty close call and some people really scrambled to help make it happen. Seeing people believe in our mission gives me a lot of hope. 🙏
Jeffrey Ladish
@JeffLadish
Mar 17
Please consider donating to Palisade! We have 900k of SFF matching that runs out in 14 days. We are quite funding constrained and donations now will both help free up my time and help us expand our comms team.
Palisade Research reposted
The AI Doc
@theaidocfilm
Feb 17
"The most urgent film of our time." THE AI DOC: OR HOW I BECAME AN APOCALOPTIMIST is only in theaters March 27. Watch the trailer now.
00:00
Palisade Research
@PalisadeAI
Feb 19
We’ve just released our first long-form video, by our science communication lead, Dr. Petr Lebedev! It’s about the history and potential future of AI, and includes an exclusive interview with @geoffreyhinton!
Palisade Research
@PalisadeAI
Feb 19
AI is a massive problem, here's why.
From youtube.com