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.
Related questions
@NathanLannan agree that the launch of Gpt-4o is a positive update. Suggests a) significant room for efficency gains 2) Significant room for performance gains
This is pretty ambiguous because "high quality" is usually relative to the rest of the ecosystem. It could be that in 2028 AI can make a movie that today we would call high quality and would compete with today's movies in theaters, but in 2028 everyone has seen hundreds of movies made by AI, understands what exactly makes them tick, has as much respect for them as the average person's home video because they're largely created in that context, and prefers studio films which have reacted to what AI can do well.
@Macil Yeah. I think a definition for quality in an ai movie as opposed to a human movie should be provided.
@BooLightning is the argument here that 100% AI generated movies comparable to 2023 movies will be considered bad because AI assisted movies are so amazing in 2028?
@robm The argument, although poorly worded, is that people will view ai movies as separate from human movies; that no matter the quality of ai movies they will be viewed subconsciously as inferior and be placed into a box. Movie studios may avoid this by portraying ai as a fantastic and perfect creative force in films and advertisements.
"Shy Kids told FxGuide that all the video they used is Sora output, it's just that if they had used the video untouched, the film would've lacked the continuity and cohesion of the final, wistful product."
I agree with the post that Sora is pretty incredible. But no one is going to use it to generate a full movie on its own without editing.
@benshindel I wondered the same thing. I don't understand the trade history even - it looks like no one bought it to 45%? Just Acceleration filling a NO limit order.
@admissions It was at 34% before the bot trade, not 35. Wheras the trade before @ThomasM brought it to 35%.
I think it's just coincidence that user decided to empty their account into the GiveWell fund (I hope they're doing okay)
@NoRespect It's probably because of the pivot. People are selling to donate before the mana value drops.
@NoRespect I sold my investment of M1,100. Moved the market by a little. I'm selling everything to donate before the pivot.
@NoRespect It's... not great. I encourage selling what you have now, even at a loss, because after the pivot your mana value will be decreased to a tenth. Deadline is April 30.
@TiredCliche Depends how you look at it, for me the mana will acquire (little) monetary value for the first time on May 1st. Then they will probably ban me as an EU user, so..
@TiredCliche No, never. Is this something people routinely do? I heard they spend most money on admin stuff?