IE "make me a 120 minute Star Trek / Star Wars crossover". It should be more or less comparable to a big-budget studio film, although it doesn't have to pass a full Turing Test as long as it's pretty good. The AI doesn't have to be available to the public, as long as it's confirmed to exist.
On June 30th, 2027, I'll try to ask @ScottAlexander if he thinks such a capability already exists. If he's unavailable/unwilling to adjudicate, I'll defer to my best understanding of the criteria. I reserve the right to resolve N/A or wait until Scott clarifies the criteria if there is significant ambiguity. I won't bet in this market.
See also:
1,000
1.00@brp you can't prove the tech exists without doing it imo (particularly in AI) and the economic hurdles are part of the challenge. I'd say it requires someone to do it, but it doesn't matter how much they spend if they choose to do so. (I don't think it will work either way 😉)
@DanMan314 No, the main market is about "In 2028". Left a comment there asking to update the close date.
@firstuserhere I interpreted the market the way @Gen did. Scott assessed his last 5 year predictions at the beginning of the year. Probably good to get clarified.
@DanMan314 I have never read anything by Scott. Stuff that is intended, should be included in the market. You're not expected to read my blog etc to interpret my markets.
@firstuserhere That’s why I clarified “at the beginning of 2026” for my market. I agree it’s ambiguous, just sharing how I interpreted it.
@firstuserhere Yep, you phrased it just right! I interpret scott's markets about "In 2028" to be about things that are possible IN 2028, not BY 2028.
@firstuserhere I have created this market : https://manifold.markets/Zardoru/scott-alexanders-in-2028-ai-movie-m
@firstuserhere Scott always scores his predictions at the beginning of the year.
It is about Jan 1, 2028.