Will an AI be able to generate a high-quality, fully-functional video game to a prompt by market close?
Example:
An AI receiving the prompt “Create a complete, high-quality, rogue-like deckbuilding video game set in a dystopian future where humans and robots are at war.” should be able to handle all aspects of game creation from this prompt alone, resulting in a fully functional, high-quality video game.
Specifications:
For the purpose of this market:
AI-Generated Video Game: Refers to a complete and functional video game created by an AI without human intervention in the creative process. This includes game design, mechanics, graphics, sound, and coding.
High-Quality: The game should be comparable in quality and complexity to a mid to high-budget video game produced by humans in the same period. It should have a coherent plot, well-designed game mechanics, professional-level graphics and sound, and be largely free of bugs and glitches.
Fully-Functional: The game should be playable from start to finish, offering a complete gaming experience. Should be capable of making experiences that have playtimes comparable to human generated titles.
Conditions:
AI Generation: The AI should develop the game autonomously, including game design, storyline development, character creation, world-building, coding, and testing. Similar to how Midjourney generates a completed image, this AI should generate a completed game.
Verification: At least three reputable industry experts OR MYSELF must verify the AI's ability to handle game quality, functionality, and prompt robustness without human intervention.
Release: The AI doesn't need to publish the game to a store (ex: sell the game, create marketing copy, or deal with storefronts), we are just looking at creating the game.
Resolution:
The market will resolve as “YES” if a game meeting the above criteria is confirmed by market close. If no such game is confirmed by this date, or if the game does not meet all the specified criteria, the market will resolve as “NO”.
- I will NOT bet in this market.
- In the event that a technical anomaly or unforeseen issue leads to a discrepancy between the literal text and the intended meaning (spirit) of the question, the resolution of this market will align with the underlying intent of the question, ensuring the most fair and accurate outcome.
@DavidBolin ahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahaha 😭
@Stralor I've not done movies but I've done games, and while it was intense, the stories I've heard from friends involved in movies are truly insane.
@Odoacre the thing is that movies are intense, long days, usually gig work. games tend to be more "stable" fulltime employment (though, traditionally, games had insane "death marches" of long hours for years on end; people have died of heart failure in games from intense work. often.)
The film industry has a couple of advantages, though: their workers are unionized these days in the Western world (so costs can be estimated easier, though they're higher than games hour-for-hour bc of this) and we know how to make a film!! We make it up when we make games, even those of us who "know" how to make games really have no idea what makes a good game, the established formulas and best practices from film don't transfer well (yet?). So unless you're making FIFA 2060, the recipe isn't gonna be the same, the tech is moving out from underneath you constantly as you have to push to meet new expectations, and more often than not you're trying to combine music, painting, film, sculpture, and novels with server infrastructure, electronic hardware compatibility, and full force software development.
Below I've added a couple of lists of the "most expensive" to compare costs at the extreme end. Note, however, that while they appear roughly about equal, I suspect they're not even close. RDR2 famously took 1600 people 6-9 years to make, incl later platforms and releases. I heard it was secretly north of 2000 people (but the industry does a terrible job of crediting reliably). The cost estimates seem woefully low from where I'm standing; 1000 people/ year has an average burn rate of 100 million in the US, before marketing costs, so something is fishy with those "it cost 200 mil total" estimates.
It's safe to say that your average film takes 3-4 years, from concept to post production. These days, your average game takes 3-5* after concept and early proofs, though it used to be shorter. *Big caveat: this easily doubles if you do something like an MMO.
Even setting aside labor amounts, ability to conceptualize the project, the wide range of required skillsets, and lack of knowledge around best practices, there are two massive issues for an AI to overcome to successfully make a game.
The first is possibility space. A film is presented linearly. This will be difficult for an AI, as we're seeing with object permanence, but they're making improvements. Games are not presented linearly, and there must be a multitude of optional paths, interactions, and environments to consider interwoven with several layered mechanics that must be balanced, which all must have object permanence and state tracking.
The second major issue is implied by the first: the player. A film audience can react, but they do not affect what happens on the screen. A game's player is both audience and author, and an AI will have to understand not only the rules of the game but how that will lead to it actually being played. It can't pick that up like studying 1000s of hours of films and embedding tropes frame-by-frame because there's actual functionality required to be successful.
I better stop here. There are game generating AIs; I've got friends and colleagues working in that space. But they're narrow-framed generators, fitting a particular well-trod style of game such as platformer or puzzle and only able to do very small scales. To make a "High Quality" mid-to-high budget game, you need a generalized AI with a ton of power. I truly believe this is strong AGI or later territory.
@Stralor the cool thing about games is that to have a high quality game you don't have to push the envelope on the graphics. You mention some high cost games such as RDR2 but there are many games that are extremely high quality (at least in my opinion) that have been built by very small teams. Slay the spire is a great example.
@Odoacre oh definitely. I'm deep in the indie games space and also enjoy watching indie film from time to time. quality != fidelity at all but there is a minimum threshold in both the kind film and the kind of games we're talking about for this market
@Stralor the minimum for games is "mid-budget" which is probably much lower than the extremes of RDR2 or BG3.
You seem to assume only AAA games should be counted, but most games are indie games.
StS might even qualify as mid budget overall.
@Odoacre eh I read mid-budget as AA, so I'm thinking of like Hellblade and Divinity: Original Sin 2. Stuff that has significant financial support but isn't as massive as, say, Uncharted. Indie, even with an indie publisher and in the modern (lack of) definition doesn't qualify imo, so I'm thinking double digit millions+, mostly
@Stralor I don't think "a team of 10 made this" qualifies as mid-budget in any circumstance, games or film
@Stralor Even then, I don't think "hey AI, make me Stardew" or whatever dope ass indie game really is that much more solvable by an AI. Way too many moving parts.
I think the main challenge is not the moving parts per se, but making the game
at least somewhat fresh
interesting in balance terms
"fun" and entertaining
But those same challenges map to the movie market(s) also
@Odoacre fair enough! I'm very bearish on all of the "quality" games/ films AI markets. I'm the biggest NO holder on all of these by Sneaky, but also have significant holdings on Scott Alexander's
@Stralor I agree you need strong AGI to make a decent video game.
I think that is probably true for movies too. I think people are over extrapolating from text and image generation.
@Stralor Nope, it does not have to publish the game just generate it. Thanks for the question I will add clarification!