It seems from your link like CFAR has taken responsibility, taken corrective action, and states how they’ll do everything in their power to avoid a similar abuse incident in the future.
I think in general the way to deal with abuse situations within an organization is to identify which authority should be taking appropriate disciplinary action regarding the abuser’s role and privileges. A failure to act there, like CFAR’s admitted process failure that they later corrected, would be concerning if we thought it was still happening.
If every abuse is being properly disciplined by the relevant organization, and the rate of abuse isn’t high compared to the base rate in the non-rationalist population, then the current situation isn’t a crisis - even if some instances of abuse unfortunately involve the perpetrator referencing rationality or EA concepts.
Great post! I agree with this analogy.
I think the fire stands for value creation. My Lean MVP Flowchart post advises to always orient your strategy about what it'll take to double the size of your current value creation. Paul Graham's Do Things That Don't Scale is a coarse-grained version of this advice, pointing out that doubling a small fire is qualitatively different from doubling a large fire.
I guess that’s plausible, but then my main doom scenario would involve them getting leapfrogged by a different AI that has hit a rapid positive feedback loop of how to keep amplifying its consequentialist planning abilities.
My reasoning stems from believing that AI-space contains designs that can easily plan effective strategies to get the universe into virtually any configuration.
And they’re going to be low-complexity designs. Because engineering stuff in the universe isn’t a hard problem from a complexity theory perspective.
Why should the path from today to the first instantiation of such an algorithm be long?
So I think we can state properties of an unprecedented future that first-principles computer science can constrain, and historical trends can’t.
I think the mental model of needing “advances in chemistry” isn’t accurate about superintelligence. I think a ton of understanding of how to precisely engineer anything you want out of atoms just clicks from a tiny amount of observational data when you’re really good at reasoning.
I don’t know if LLM Ems can really be a significant factorizable part of the AI tech tree. If they have anything like today’s LLM limitations, they’re not as powerful as humans and ems. If they’re much more powerful than today’s LLMs, they’re likely to have powerful submodules that are qualitatively different from what we think of as LLMs.
I agree that rapid capability gain is a key part of the AI doom scenario.
During the Manhattan project, Feynman prevented an accident by pointing out that labs were storing too much uranium too close together. We’re not just lucky that the accident was prevented; we’re also lucky that if the accident had happened, the nuclear chain reaction wouldn’t have fed on the atmosphere.
We similarly depend on luck whenever a new AI capability gain such as LLM general-topic chatting emerges. We’re lucky that it’s not a capability that can feed on itself rapidly. Maybe we’ll keep being lucky when new AI advances happen, and each time it’ll keep being more like past human economic progress or like past human software development. But there’s also a significant chance that it could instead be more like a slightly-worse-than-nuclear-weapon scenario.
We just keep taking next steps of unknown magnitude into an attractor of superintelligent AI. At some point our steps will trigger a rapid positive-feedback slide where each step is dealing with very powerful and complex things that we’re far from being able to understand. I just don’t see why there’s more than 90% chance that this will proceed at a survivable pace.
My commentary on this grew into a separate post: Contra Hanson on AI Risk
I tweeted my notes of Eliezer's points with abridged clips.
Bravo.
Which 2+ outcomes from the list do you think are most likely to lead to your loss?