EG "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.
1,000How high quality?
People already can make 2 minute or more films via prompting.
Shouldn't be that hard to make a decent movie by orchestrating current tools. Probably costs a few thousand dollars to make so there's not a large market, but by 2028 the costs should maybe be 10s of dollars and similar quality as 4o image gen but for a movie.
@ChinmayTheMathGuy says it right there in the description:
"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."
I don't think any AI system can make even a 2 minute clip that is at the level of a 2 minute clip of a big-budget studio film.
No AI system can currently even make a 2 minute clip with matching audio, video, dialogue, consistent characters, etc, from a single prompt, let alone one at the quality of a big-budget studio film.
In any case, feel free to bet YES if you think so.
@bens I mean, image generation took a big leap with the OpenAI release yesterday, and still I wouldn't even say a single IMAGE is anywhere near the level of a "professional studio photograph" or "National Geographic Cover" or something. There are levels to this stuff.
@bens also keep in mind we are now 2 years and 8 months away from market resolution.
@ElliotDavies It's a positive update from heavily below 40% to still significantly below 40%. And a very small update tbh, since this only resolves yes realistically if meaningful rsi kicks in pre-resolution date, which has ~almost nothing to do with image gen progress.
@ElliotDavies if you were making a five minute movie, how much time would 4o’s tools save you vs tools from last week?
@GG I think object permanence would save at least *some* time. Say you're using it to make a storyboard, it helps a lot if stuff is consistent.
@ElliotDavies If it was an update for you, you haven't been paying attention. OpenAI announced their current model around a year ago. The only "update" is that OpenAI finally shipped due to pressure from Google Gemini. Maybe this will spark an AI-Race, but given current trends, it is overwhelmingly likely that the first model capable of generating short-episodes (1-5 minutes) will come out of China in the next 12 months.
I don't understand how the market was at 50% prior to 2024 and then dropped by 15% afterward, even though major advancements in video generation came after this period. 50–60% is a fair estimate for the odds of having such technology by 2028. Look at the rate of advancement in image generation and extrapolate that to video.
Many will probably object, saying that video still has a long way to go before it can generate a two-hour Hollywood level film, or that models are already hitting a wall and won't be sufficient before 2028. I beg to differ. I think diffusion models will be phased out in favor of autoregressive multimodal models (similar to the recent 4o native image generation, but now imagine this for video).
There's a fair chance this market will resolve YES by 2028 because of the rate of advancement, but I wouldn’t say I’m near certain. As mentioned, 50–60% odds, but mid 30s are way too low, imo.
@LukaChrelashvili Stock markets don't go up when a company does well and don't go down when a company does poorly. They go up when a company does better than expected and down when a company does worse than expected.
Applying that same logic to prediction markets gives the obvious starting point assumption that although AI has improved, it has not improved at the trajectory people predicted in the past.
@LukaChrelashvili You could also look into what points Manifold's loan system changed, and points when lots of people got lots of mana (such as the new year). If people have more spare mana and particularly if loans flow more freely, people become more willing to push long term markets like this away from 50%.
@Tumbles there were also a couple of very bullish traders holding the market high until they departed in late 2023
@Balasar Yeah, I have fast AGI timelines (~50% around end of decade), but this kind of market is still too fast/over-hyped. Lots of just crazy over-optimism on this site XD
@GazDownright ~Any Michael Bay movie is “more or less comparable to a big-budget studio film.” I get that this market is technically inherently subjective, but in practice a lot of outcomes will be extremely obvious.
@DavidHiggs There several big-budget studio films that are of very poor quality. I think AI is already capable of making big-budget-looking absolute swill. If AI is able to create a masterpiece, however, is a more interesting question, imo.
@GazDownright I have yet to see a single example of plausibly “big-budget” looking anything from AI in terms of video, and that’s without audio.
@DavidHiggs I think current AI is approaching the level where it could make a superior version of Cats already