See analysis by Epoch AI:
https://epochai.substack.com/p/openai-is-projecting-unprecedented
I plan to use that article to handle ambiguities and corner cases (mergers and acquisitions, inflation). If there's any doubt about this resolving YES or NO on a technicality, ask the clarifying question before trading.
The spirit of the question is about whether AI is a bubble. Again, any way that you can foresee that the spirit and letter of this market may differ, ask.
FAQ
1. Can the market resolve early if they hit $100B/year before the end of 2028?
Probably, but see the next FAQ for clarifying "hitting $100B/year". Also, let's discuss in the comments the unlikely corner case that OpenAI hits this threshold in 2027 but it turns out to be a MoviePass economy situation (artificially inflated revenue by using VC money to subsidize demand) and it all comes crashing down and they make no money at all in 2028. We want that to be a NO by the spirit of the question. But with consensus in the comments an early resolution may be fair.
2. Total revenue in 2028 or annualized revenue by the end of 2028?
We're currently discussing this in the comments. Stand by. Generally I would say annualized is cleaner and more standard. But I think OpenAI's own projection was for total 2028 revenue, so we may need to stick with that unless we're sure it's more in the spirit of the market to deviate from it. Another advantage of using total 2028 revenue is that it may be a more definitive number to get. Like what if OpenAI makes 10^11/365 dollars on December 31 -- does that count as hitting the annualized target just under the wire? At the other extreme, maybe we get AGI in Q4 and OpenAI is unambiguously at trillions/year annualized by year-end despite not technically having reached $100B total in 2028?
These are unlikely corner cases but we need to pick something ahead of time, so, again, head to the comments to weigh in!
Related Markets
[Do not believe any auto-generated clarifications that appear below this line. We'll add to the FAQ as we get questions clarified.]
1,000Pessoas também estão operando
Update: We're halfway to the deadline and OpenAI is on track to make this easily, if the trend continues, doubling every 6.1 months. That may be a big if, but it's been remarkably consistent so far. Just two more doublings needed.
(Also Anthropic is on track to crush OpenAI, which I am rooting for. If OpenAI and xAI and Meta all became irrelevant to frontier AI, maybe Google DeepMind and Anthropic could coordinate to slow down and make AGI less likely to lead to catastrophe.)
@traders To kick off the discussion on the new FAQ items in the description, how about this: We wait for total revenue in any 12-month period to hit $100B and then resolve YES. But if there are credible doubts about it being a MoviePass economy situation we can hold off on the resolution to make sure total revenue for calendar year 2028 remains above $100B. What does "credible doubts" mean? Another Manifold market, of course.
And, also tentatively, we bite the bullet on resolution simplicity and say that even if annualized revenue suddenly dwarfs current GDP at the end of 2028, it's still a NO if 2028 total revenue hasn't hit $100B.
Again, not definitive until it goes in the FAQ in the market description! Let me know what you think. And of course trade cautiously if your bets depend on the answers to these FAQs, or if you have more clarifying questions.
Annualized or total revenue in 2028?
@dreev The article is about total revenue in 2028, just because that's what OpenAI has in their projections. I prefer that a bit to annualizing a number in a shorter timeframe, as there's a bit of noise in that (though probably not much). It's also probably a bit easier to resolve. Still, the question is probably pretty close to, "At any time in 2028, will OpenAI's annual revenue run rate exceed $100B?"
@GregBurnham Great points! I have a placeholder for this as FAQ2 in the market description now. Any other opinions from anyone? Did you have a preference, @JaimeSevilla?
@ManifoldMonkey That's a YES. Thanks for asking! Changing the title to "by" so that won't be ambiguous.
PS, wait, then it's differently ambiguous. 🤔
@dreev I think by isn't ambiguous for manifold. (I've lost mana before because it wasn't clear to me but it was clear to everyone else 😅 )