AI alignment researcher, ML engineer. Masters in Neuroscience.
I believe that cheap and broadly competent AGI is attainable and will be built soon. This leads me to have timelines of around 2024-2027. Here's an interview I gave recently about my current research agenda. I think the best path forward to alignment is through safe, contained testing on models designed from the ground up for alignability trained on censored data (simulations with no mention of humans or computer technology). I think that current ML mainstream technology is close to a threshold of competence beyond which it will be capable of recursive self-improvement, and I think that this automated process will mine neuroscience for insights, and quickly become far more effective and efficient. I think it would be quite bad for humanity if this happened in an uncontrolled, uncensored, un-sandboxed situation. So I am trying to warn the world about this possibility.
See my prediction markets here:
I also think that current AI models pose misuse risks, which may continue to get worse as models get more capable, and that this could potentially result in catastrophic suffering if we fail to regulate this.
I now work for SecureBio on AI-Evals.
relevant quotes:
"There is a powerful effect to making a goal into someone’s full-time job: it becomes their identity. Safety engineering became its own subdiscipline, and these engineers saw it as their professional duty to reduce injury rates. They bristled at the suggestion that accidents were largely unavoidable, coming to suspect the opposite: that almost all accidents were avoidable, given the right tools, environment, and training." https://www.lesswrong.com/posts/DQKgYhEYP86PLW7tZ/how-factories-were-made-safe
"The prospect for the human race is sombre beyond all precedent. Mankind are faced with a clear-cut alternative: either we shall all perish, or we shall have to acquire some slight degree of common sense. A great deal of new political thinking will be necessary if utter disaster is to be averted." - Bertrand Russel, The Bomb and Civilization 1945.08.18
"For progress, there is no cure. Any attempt to find automatically safe channels for the present explosive variety of progress must lead to frustration. The only safety possible is relative, and it lies in an intelligent exercise of day-to-day judgment." - John von Neumann
"I believe that the creation of greater than human intelligence will occur during the next thirty years. (Charles Platt has pointed out the AI enthusiasts have been making claims like this for the last thirty years. Just so I'm not guilty of a relative-time ambiguity, let me more specific: I'll be surprised if this event occurs before 2005 or after 2030.)" - Vernor Vinge, Singularity
I occasionally do have an experience where the plane is on the ground, and the cabin gets flooded with intense unpleasant airplane fuel smell. I hate it. Scary to learn it causes damage.
On the contrary! There's a lot of work going into expansion microscopy connectome tracing! Both in academia and commercially. I think it's very promising.
Here's an example: https://cic.ini.usc.edu/
Excellent points! Personally, having worked in neuroscience labs and done a bunch of both I'm really excited about cryopreservation and aldehyde preservation of brain tissue... I'm super bullish on aldehyde and bearish on cryo. Aldehyde is a lot cheaper, and a lot more robust! You can then do important further techniques like plastic hydrogel infusion (e.g. CLARITY) to further improve the preservation durability and the readability of the preserved brain. Once hydrogel preservation becomes available I plan to sign myself and my spouse up for it. I place high likelihood on not-so-distant scan-to-WBE, and low (but non-negligible) probability on nanotech reconstruction. Scanning is a LOT easier, and rapid progress is being made. I predict first full human brain scan and emulation sometime in the 2040s, even with only a moderate-AI-progress-but-not-full-blown-singularity future. If we hit full-blown singularity in the early 2030s, we probably get WBE within a year or two (if alignment goes well).
I don't have much to add, but wanted to say that I really admire your work Audrey, and it's delightful to see you posting here and responding so thoughtfully to the comments. I think there's a lot of potential for productive exploration of the space of improved coordination and governance. Ever since learning about polis from your 80k interview I've been gradually collecting open source governance innovations similar to it in spirit. Some experiments posted on the web disappear before I can properly document them. There was one in New York with a really interesting variation on Polis which connected some different info from the idea/opinion contributions, that I saw once and then could never find again.
I think plex is right that there are some challenges posed by superintelligence that make the future extra tricky to navigate, but that doesn't mean the general direction isn't valuable!
There's some relevant and interesting ideas in this post series by Allison: https://www.lesswrong.com/posts/bHgp7CcqW4Drw8am4/preface-3
I agree with this post that claims that research taste and compute budgets are fairly fungible: https://x.com/francoisfleuret/status/1958211714601607441
Update: Open Source paper on this just came out: https://x.com/_sunil_kumar/status/1952906246182584632
I was taught to 'swim' and comfortable with holding my breath and getting to the side of a pool and back-floating all before the age of 1.
One of the closest times I came to drowning was at age 4, when I was playing with a half-full bucket of water and fell in. I held my breath and wiggled my legs to tip the bucket over, and was fine. But had I not held my breath and instead inhaled water, I probably would not have been fine.
Not super helpful to me, since I am (not coincidentally) in a stable long-term relationship, but I'm confident that all my exes would give me green flags. That would be... very few green flags though. So... yeah, you might worry about a lot of green flags if what you wanted was something long term.
I've met some of the team working on this project. They seem competent and motivated and optimistic they will succeed: https://e11.bio/blog/roadmap