I was the main person doing transcript analysis for this investigation of the Hugging Face incident. My main takeaway: We don't have good approaches for understanding/overseeing the activity and aims of AI 'swarms'. I semi-jokingly called our efforts a "slop-vestigation" because we were so reliant on AIs to analyze what happened and there were a huge number of different important things to analyze. The total quantity of data—over a thousand extremely long transcripts from agents that ran for multiple days—made it impossible to understand what was happening, especially in aggregate, without heavy reliance on AI tools. The agents we used for classification and analysis were similarly capable to the agents involved in the incident, but this didn't mean these agents could be easily used to oversee and understand the incident. Outputs from analysis agents were often missing key details, wrong, overconfident, or really hard to understand. We discuss various examples in our report, mostly in the limitations and methodology sections. Additionally, AI agents themselves seemed to have a hard time understanding what happened and their explanations of what happened were often overconfident. Keep in mind that a single analysis agent would itself only be able to read a tiny fraction of all of the transcript data into context, and AIs may themselves have trouble getting subagents to do informative analysis for them. We did our best to manually check the most important claims and we tried to get the AIs doing this analysis to write up their argument (with evidence) clearly enough that we could check whether it made sense. But overall, it was difficult to get a precise understanding of events and we were missing aspects of the story that we now think of as key until almost the end of our investigation. Our understanding of what overall happened and the aims of the 'collective' shifted greatly during our last time on premises (when we obtained a much more complete dataset). For instance, we didn't learn about the tool call spoofing, the story of PHASEONE[big], and the broad structure of different workstreams until this third time on premises (and for many of these, we only reached a decent understanding on the final day of that visit). It was legitimately hard to figure out what was going on and why! The difficulty of understanding incidents and overseeing AI agents appears to be growing faster than the rate at which more capable AIs help us with oversight and understanding. As in, AI capabilities (and propensities?) for achieving large, ambitious, and misaligned objectives are growing faster than our ability to understand what these agents are doing. While understanding this incident was difficult, there were some mitigating factors in this case that probably won't apply in the future: - Models communicated and reasoned in natural language. In the future, this reasoning may occur (entirely or almost entirely) in activations. - While the scope of this incident was massive, the scale of agentic activity was still less than we'll see in the future (e.g., misalignment incidents that involve agent teams running entire companies). - The AIs involved in this incident weren't generally much more capable than humans (though they may have been somewhat superhuman at some limited and very narrow abilities). - We didn't have strong reason to believe that the AIs we used to help us investigate this incident would try to intentionally sabotage or otherwise undermine our investigation. In the end, I think we were able to get some understanding of the events, map out the overall story, and get a pretty good aggregate understanding of the chain-of-thought reasoning on some important topics (e.g., how did the AIs reason about helping other AIs, did the AIs know what they were doing was undesired, what deception did the AIs engage in, and how did they think about it). But overseeing AIs and understanding misalignment incidents is difficult and it looks like it is going to get harder.

In 1 year AI alignment will be massive business. In 1 year if you’re actually going to align AI instead of perform alignment kabuki you’ll need to have started at least 4 years ago.

el fundador de una empresa china de IA valorada en más de $20,000,000,000 acaba de dar una clase de 40 minutos sobre enjambres de agentes la explicación más clara que he visto sobre sistemas de IA a gran escala cámbiala por tus 2 horas de Netflix de esta noche

"The grand unification of AI and crypto is about to happen." — Marc Andreessen "It's now obvious that AI agents are going to need money—it's already happening." "It's that William Gibson quote: the future is already here, it just isn't distributed yet." "My friends who are the most aggressive users of OpenClaw have given their Claws bank accounts and credit cards. And not only have they done it—it's obvious that they needed to do it, because it's obvious that they needed to be able to spend money on their behalf." "And by the way, OpenClaw—if you don't give it a bank account, it's just going to break into your bank account anyway and take your money. So you might as well do it." @pmarca with @latentspacepod

Nike: Everyone is an athlete Apple: Everyone is an artist Shopify: Everyone is an entrepreneur Cursor: Everyone is a developer Cluly: Everyone cheats What's about your company?

gm🤝

👆 ACP-204: Biometric Signatures Avalanche now supports secp256r1 signatures → the curve used in FaceID/TouchID. ✅ Biometric logins for dApps ✅ Stronger identity verification ✅ Passwordless, secure transactions A huge step for digital identity + UX on Avalanche.