Why the conditional adds less than it looks: AGI compresses discovery, not validation. The parent market wants convincing human evidence by 2050, and a human lifespan trial takes about 21 years from enrolment at 65, so the trial that produces that evidence has to be enrolling by about 2029 whatever the AI does, and the median AGI date on Metaculus is later than that. What an AGI could change is the shortcut: a validated surrogate endpoint that regulators accept in place of mortality, which does not exist yet and would itself need years of mortality data to validate. The nearer gate is the same with or without AGI: the dog lifespan trials now running (TRIAD, Loyal's STAY study) read out around 2029 to 2031, because dogs enrolled at 8 reach median survival in 4.6 years rather than 21, and a drug that fails there is unlikely to get a human lifespan trial at all. The arithmetic with sources (mine): https://w0lph.github.io/k9/why-dogs.html?v=2

Why the path to human longevity runs through dogs

Computed from the cited life tables and trials: a lifespan answer takes 4.6 years in dogs and 21 in people; dogs share our environment and diseases, lifespan has already been extended in dogs, and a regulator has accepted lifespan extension as an indication for dog drugs.

well

I personally feel the notion that "AGI = cure for aging" ignores some unignorable possibilities: AGI could end up showing that aging is incurable in humans, or aging may turn out to be so complex that even ASI will never solve it.

@k1234567890y Or the takeoff is slow enough and AGI is achieved late enough that it simply isn't cured pre-2050. E.g. 20-30 year timelines until 30% GDP growth, and curing aging being relatively harder than expected could deliver a no resolution.

Still, all of that put together is less than a 50% chance imo, in large part because I have AGI timelines closer to ~5 years, and because I find significant software only RSI fairly likely. Also because aging is very unlikely out of reach for ASI in the long term.

@DavidHiggs Well, the common transhumanist assertion that "AGI/ASI = a cure for aging within years" is just a hypothesis, nothing more or less than this.

By the way, speaking of aging and AGI/ASI, I am starting to have a feeling: there's a chance that AGI and ASI will prove that aging as it is now is actually the best solution for maximizing lifespan.

What is a cure to aging defined as?

@AnishaZaveri however they're defining it in this market:
https://manifold.markets/DanielKilian/by-2050-will-there-be-a-cure-to-agi

By 2050, will there be a cure to aging?

28% chance.