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 just want to comment that I think Minsky’s community of mind is a better overall model of agency than predictive coding. I think predictive coding does a great job of describing the portions of the brain responsible for perceiving and predicting the environment. It also does pretty well at predicting and refining the effects of one’s actions on the environment. It doesn’t do well at all with describing the remaining key piece: goal setting based on expected value predictions by competing subagents.
I think there’s a fair amount of neuroscience evidence pointing towards human planning processes being made up of subagents arguing for different plans. These subagents are themselves made up of dynamically fluctuating teams of sub-sub-agents according to certain physical parameters of the cortex. So, the sub-agents are kinda like competing political parties, that can fracture or join dynamically to adapt to different contexts.
Also, it’s important to keep in mind that actually the subagents don’t just receive maximum reward for being accurate. They actually receive higher rewards for things turning out unexpectedly better than was predicted. This slightly complicated surprise-enhanced-reward mechanism is common across mammals and birds, was discovered by behaviorists quite a while back (see: reinforcement schedules, for optimizing unpredictability to maximize behavior change. Also, see surprisal and dopamine). So yeah, not just 100% predictive coding, despite that claim persistently being made by the most enthusiastic predictive coding adherents. They argue for that, but I think their arguments are trying to turn a system that 90% agrees with them into one that 100% agrees with them by adding in a bunch of confusing epicycles that don’t match the data well.