.arunim.fyiI wish to revise this page. I have a vision of a well-constructed database that pulls tons of info from Epoch AI, mostly, and perhaps some others, and analyzes the factors I care most about. This is totally doable! Specifically, the database format allows for defining the actual relationships between things like models, companies, and all sorts of metrics. Reminder-to-self to use TypeIDs for this.
AI progress has been fast, as evidenced by many well-defined forecasts (see Forecasting AI (Overview)), as well as by far-less-well-defined vibes. Even more to the point, AI adoption has been remarkably fast (see The Rapid Adoption of Generative AI | NBER)
This makes it important to understand if the goal is to have a positive impact on the world—it would suck to try to do something great, succeed according to the goals I had when I set out, and then get steamrolled by the impact that AI has.
Given this, I’m curious to try to figure out:
- How can we ensure society adapts to rapid AI progress?
- How can we accurately measure AI progress? (see LLM evaluations)
- What capabilities do I expect each year?
- When will key milestones be hit and when?
- What does diffusion of quickly-developing technology look like?
- What Computers
Still Can’tWon’t Be Able To Do (ref) - Is RSI real? see models of rsi
What should I be doing in light of these expectations? I think that an 80/20 of this probably looks more like considering what a society or world looks like where intelligence is solved. If intelligence is no longer a bottleneck, what happens? Are there really other bottlenecks as the Economists™ claim, or are these bottlenecks broken through by the aforementioned intelligence once they come up?
Exponential vs Superexponential
it seems that a key question which determines power concentration / unipolar vs multipolarity is whether “power” will accumulate exponentially or superexponentially. and if exponentially, whether various actors will have the same exponent or different exponents.
if it’s superexponential, then the power will accumulate to a single actor.
if it’s exponential and one actor continuously has a larger exponent than the other actors, then the power will accumulate to the actor with the larger exponent.
if it’s exponential and actors have the same exponents or the exponents are noisy over time but similar for all actors, then the current power distribution will be maintained.
This all begs a question of how to define power and how to track whether it’s exponential or superexponential. The easiest proxy for the near future is to look at the revenue growth of the companies or the gdp growth of the countries, depending on what level of resolution you are interested in. So far, we have fairly strong evidence that the revenue growth of the companies is exponential, and the same goes for the gdp growth of the countries.
Seems like companies have had dissimilar growth rates. Anthropic is growing at 10x/year, OpenAI is growing at 4-5x/year, etc.
Can you be ASI pilled and believe that we stay at 2% growth? probably not. If that’s the case then ASI really isn’t all that special.
It seems like the default path right now is that Anthropic becomes the unipolar power. If things change, and e.g. Anthropic and OpenAI both begin to have 7x / year growth, then they will both hold a roughly similar amount of power in the long run, barring other factors.
Other Factors
labs and countries might sabotage each other such that the tallest sunflowers get cut down a notch, which would bring down their growth rates to something closer to their competitors. alternatively, they might preemptively stop the peers who are behind from sabotaging them or catching up, such that even in an exponential growth environment, the first mover advantage will still be maintained.
on the other hand, labs and countries might cooperate with each other to prevent such negative-sum dynamics and instead coordinate such that even second movers are afforded a large ratio of power in the long run, out of generosity from the first movers.