"Beat Factorio" means constructing a rocket.
"Figure out how" means roughly that it wasn't hand-held through the process. Roughly, the supervised part of the AI's training process must have been plausible in a world where Factorio didn't exist. (But it's allowed unlimited unsupervised play of Factorio.)
Details:
Examples of things that are okay:
a human can choose a favorable map seed
the AI can play Factorio unsupervised for any length of time
the AI can access the wiki (or other online resources) during its unsupervised play
Examples of things that aren't okay:
no knowledge of Factorio-in-particular (e.g. recipes, blueprints, "muscle memory") can have been programmed into the AI
the AI must not have had any supervised training on Factorio specifically (e.g. watching videos of people playing)
the AI's reward function must not encode any deep knowledge about Factorio specifically (rewarding novelty is fine; rewarding it each time it constructs a new kind of item is... kinda borderline, but I lean not-fine?; rewarding it each time it constructs a new kind of item along the critical path is not fine)
Every January, I'll check whether there's any credible claim that somebody's done anything-like-this; if so, I'll dig in to see if it meets my "figuring out" criteria; if so, I'll resolve this market YES. Otherwise, on 2030-01-01, I'll resolve it NO.
Update 2025-03-05 (PST) (AI summary of creator comment): Factorio Version Update
Any version is acceptable: The experiment may be conducted using any version of Factorio.
Update 2025-05-05 (PST) (AI summary of creator comment): * Factorio-related data included in the AI's training set due to broad web scraping is acceptable.
Specialized post-training or fine-tuning specifically for Factorio is not acceptable.
The guiding principle is that the AI was not intentionally designed or fine-tuned to be proficient at Factorio.
Update 2026-09-08 (PST) (AI summary of creator comment): * Launching a rocket in the Space Age expansion does not count.
Beating Space Age (completing the expansion) is accepted as a YES resolution.
Update 2026-09-09 (PST) (AI summary of creator comment): * An AI victory with enemies turned off is acceptable.
1,000People are also trading
Welp, I think that's it.
@OptimizationProcess I am somewhat skeptical that this run was real (it was done out of public view, unlike https://www.twitch.tv/danielv1234 ), so I think you should leave unresolved until further information emerges. Good chance this will soon see a replication with more complete public evidence.
@ICRainbow Very reasonable to ask! I would take a YES at 1% (it was higher before Mira posted their video), but I don't wanna deal with the transaction cost and counterparty risk of an OTC bet with a Manifold stranger.
Rules clarification: this market was about vanilla Factorio because Space Age didn't exist yet. Launching a rocket in Space Age doesn't count, since it's so much easier than in the base game. (At time of writing, I believe Astra has done this in a way that's at least 80% of the way to a YES resolution by my lights.) I will also accept beating Space Age as a YES resolution, since that's at least approximately the same challenge level.
Looking at how much Claude is struggling with pokemon - a much simpler game with fewer decisions and less need for spatial thinking, I'm sceptical. If it happens relatively soon, I doubt it'll be achieved by general purpose AIs like LLMs.
For those betting Yes on before 2027 - do you know of any projects for game playing AIs? Genuinely curious if anything looks promising.
@ProjectVictory Claude's struggles with Pokemon feels like it could be a few "simple" advancements away from easily beating the game. Improve its vision, long-term memory, and spatial reasoning, and it should not have any issues beating Pokemon. And from there it should be able to beat any other slower paced game - just scale it up.
Of course, any one of those advancements may prove to be very difficult. My intuition tells me they will all be solved within 5 years, and possibly much sooner.
The biggest issue seems to be the no Factorio in the training set clause. Noone training a foundation model is going to go to the trouble of excluding Factorio videos from their data set just for the sake of this market.
@ThisProfileDoesntExist Good point. To resolve ambiguity in what counts as having information "programmed into the AI": I declare it to be okay if there's Factorio stuff in the training set due to broad web scraping -- just no specialized Factorio post-training. Roughly speaking, nobody was designing the AI to be good at Factorio.
In the case of a generalised game-playing AI (which the criteria seem to describe) there needs to be a generalised reward mechanism. This is not trivial and I wonder what people think it might look like.
I think curiosity-based reward is a good candidate but I'm not sure it would result in deep mastery of arbitrary games.
@ThisProfileDoesntExist having read the wiki, what would make it want to follow the objectives? What would its inner motives be in general?
@Tomoffer reading the description again I guess you'd expect an auto gpt style system to be given the explicit natural language instruction to "meet all objectives" or something
I would expect Factorio to first be beaten by either:
A giant multimodal transformer model trained on the whole internet including the parts of it that are about Factorio, which fails your criterion of not watching any videos of people playing Factorio.
Some sort of narrow RL model trained on Factorio specifically, which is probably going to be rewarded for collecting items along the critical path.
I.e. if the team making an AI beat Factorio doesn't specifically have your rules in mind, I expect your rules to be violated and the AI to not count for this market.
@Multicore If you train a game-playing AI on every game that exists, you'll have to create a brand new game to evaluate whether that knowledge has generalized. You're probably right that the first AI to do this will be trained specifically for it, but there is a motivation to create an AI that can work out games in a blind playthrough (as it is a step in the direction of AGI). An Alpha Zero approach wouldn't count for this market, AFAICT, but it does show that work is being done towards that end.
Imo, at a glance this mostly depends on one of two things happening
Some company decides to try training a model to play more complicated arbitrary games and Factorio is one of them
Plausible, but I'd maybe say 30% that it would get chosen for the list. But even if they chose Factorio for the list of games to run it on, they may not care so much about deliberately getting it to beat Factorio
ex: they may just do simpler shorter term goals, especially for an early attempt. Like Minecraft they may do 'build a cool mansion' and Factorio 'automate a factory producing green science'.
This might just be it being a shorter term reasoner, but also might be them just not bothering to apply it for that much (so it could be possible in principal but just not done)
We get generalist enough and fast enough agents you can run on your computer. Like if it suddenly became a lot more cheaper to train a model like this, so a 'bunch' more hobbyists can enter the arena.
One method is GPT model hooked up to code interpreter and just let it generate code to solve for solutions ;d