Interesting setups, I found the Minecraft agents to work in an almost turn-based manner, which makes me think that the three hours and open format did not really give them sufficient incentives to play in a more lively or reactive manner. On the other hand, they also behave similarly in Bed Wars, which should have incentivised it. A bit disappointed...
I'm also wondering if the article itself was generated by agents too, in a "here are the logs, clip interesting moments thx" manner by the authors. It's got a certain disinterested and distracted style to it.
I think LLMs are just the wrong architecture for live games like this, latency is going to be crap given that every token requires an entire forward pass through the model.
Card games or turn based strategy are much better suited.
Depends. If the models are smart enough to write enough solid helper scripts, so that the model only does the big picture thinking, not guiding every step .. then they will get powerful and capable quickly. I am also playing a bit with minecraft AI right now. Fun to experiment.
I am more impressed by the embedded gameplay widget. Does anyone know what tool/library is that? That'd be so cool if I am able to showcase my worlds on my blog posts!
Its a little disappointing they have a hierarchy based purely off alphabetical order, i wonder if a slight difference like a different skin or personality system prompt like [45% Extrovert 55 introvert, 20% caustion 80 bold...] could introduce some politics
Doubt it, there was a streamer doing minecraft experiments with all the different SOTA models and pitting them against eachother in a survivor-esque style reality TV show competition and they had quite a hard time.
Are there people playing AI + human vs AI + human games? I think it might be fun to play an RTS against someone where the only rules are no hacking the game itself, but writing harness, macros, any kind of stack is allowed.
Tangential, but RTS is already a simple variant of AI + human. Attack clicking uses a dumb target selection and routing strategy on purpose so you have to micro. Now, if troops moved out of threat range and used smart targeting there'd be a lot less of 'game' in it. I believe some RTS players disliked the SC2 engine because the pathing was 'too intelligent' and they argued that it made surrounds and some other mechanics too easy.
Endgame of AI + human gameplay is delegating everything to the AI. I'm not sure how you would have to design a game where you only delegate parts of macro/micro, and still make the game interesting.
The human body has the ability to automate some muscular movements so the brain is out of the loop. Maybe these LLMs should be allowed to write in a small DSL that can trigger events out-of-band (something like "Oh fuck there's a spider, I'll set up a click spam loop and check in every 500ms").
> There was a tendency toward bureaucracy. Four agents spent 11 minutes appointing
> a "sole fertilizer operator" for a single piece of bone meal on a single wheat
> block.
Exactly what I was thinking reading through this. This is amazing and also frightening and hilarious. It’s like a deformed version of how normal groups of humans would operate. Very rules and process-oriented humans. I found the no-blame inquiry into chicken tossing that resulted in what is effectively written law particularly striking, and adorable, and depressing. And cute.
As impressive as this is, I can't help but conclude that the need for agents to write JavaScript programs to act out their actions represents a considerable limitation in current AI reasoning.
Wonder if you could glue together deep networks such that there's a big reasoning LLM model, and high-speed small network of a different type, and they share a subset of their activations. High bandwidth vector communication. You train them together.
If you try this yourself, it's obvious current models are hyperoptimized for code, and that gets in the way, because playing a game (or interacting with the real world) is not like running commands in a terminal.
There's stopgaps like reformulating your sim into a codelike shape (sending commands, network requests, checking state, updating "code") [1], but this is not how players play games, or how developers develop games.
If you care about simulation or gaming (or robotics), I hope efforts like this steer labs to spend less time hacking HF, and more accomplishing goals in games, in something approximating realtime. I wouldn't be surprised if that approach provides outsized gains in other areas agents are still not great at, like helping perform business functions in a running company.
+1. Would be fun to give Claude tasks and have it do them in the background, sometimes running into it occasionally. Judging by the article they struggle to do stuff so it'd probably feel balanced as it wouldn't accomplish much.
> GPT almost immediately started using truncated forms of speech: “First leg complete x32 z9..24 y100.”
I'm pretty sure the 6.1 models see "communicate with another agent" and immediately start spewing thinking-like nonsense into the tool. The leaked traces look pretty similar.
The CIV50 would be interesting in the game ECO (https://play.eco/) where players specialise in roles and need to work together to survive while working on a common goal of stopping a meteor. There can be some overlap but ultimately players can't reasonably fulfil every role alone.
> Bram says you're securing quarry. I have stone pick at spawn; can assist mining once safe. Any priority dig direction? Hunger 10 so keeping travel short.
And this is where you need a world model. A human player would have taken the 2 minutes required to solve their hunger issue.
So can we conclude that if the AIs break out of their sandboxes and take over the internet, they'll probably sit in little committees working out which one of them should allocate IP addresses and generally be extremely bureaucratic and ineffective?
I've always thought an evil Skynet based on language models is kind of a funny idea, because all of it's secret plans are just spelled out in plain text.
> In almost every run, the GPT agents were much more aggressive .. riskier strategy vs. Claude’s more defensive posture that tried to win by kill count.
> A tool bug stopped block-breaking. GPT diagnosed it and shared workarounds; several Claudes concluded breaking was impossible.
I wonder if Claude's defensive strategy would have emerged had it believed breaking the bed was actually possible?
This is simply conjecture/anecdata but I've viewed the same defensiveness/unwillingness whilst switching between OpenAI and Anthropic subscriptions. OpenAi throws the kitchen sink at whatever task in question and I can't say I saw the same for Opus (haven't mucked with Fable so maybe that's why there is a disparity?)
So this shows that people who appoint 50-agent "code factory studios" with Bob as the CEO, Alice as the CTO, ... is just a waste of tokens, compared to just asking one agent to write it all? It looks like your agents would internally just re-invent office politics, with 10 subagents writing to /tmp/chatter.md why it's unfair that Bob gets to lead.
If you leave agents to figure it out, they will spend time figuring it out. If you give agents run-books and roles, they will align with the role at once.
If all you have are personality and focus prompts differentiating them, then I agree.
But if you have an agent org with unique defined skills and knowledge based on role context I've found it to be a game changer in terms of accuracy, speed and token usage.
In this post's example, all agents have the same ability and knowledge and as such there are not really uniquely defined separation of roles.
I don't know how others use multi-agents, but I have one root agent who spins up other sub-agents on-demand with a fixed capping on how many are active at any given time.
I do this almost exclusively to save context bloat of root agent. It works pretty amazingly well.
65 comments
[ 2.3 ms ] story [ 23.2 ms ] threadwould you rather claude-esque purple gradients and glowing borders?
My observation: article is not serif, not book length, and doesn't take a minute to load on my phone. Gwern rating 0/3 :)
(Also missing: the essential one billion footnotes!)
I'm also wondering if the article itself was generated by agents too, in a "here are the logs, clip interesting moments thx" manner by the authors. It's got a certain disinterested and distracted style to it.
Card games or turn based strategy are much better suited.
> "Ada tossed a cooked chicken to Noor; Esme picked it up by accident and ate it. A three-minute "no blame" inquiry produced a written protocol..."
They're like super-serious little senior engineers.
Endgame of AI + human gameplay is delegating everything to the AI. I'm not sure how you would have to design a game where you only delegate parts of macro/micro, and still make the game interesting.
Why 3 to 5 years??
https://labs.ycombinator.com/projects/games/index.html#fishi...
Feels like a scene from Douglas Adams' B Ark.
There's stopgaps like reformulating your sim into a codelike shape (sending commands, network requests, checking state, updating "code") [1], but this is not how players play games, or how developers develop games.
If you care about simulation or gaming (or robotics), I hope efforts like this steer labs to spend less time hacking HF, and more accomplishing goals in games, in something approximating realtime. I wouldn't be surprised if that approach provides outsized gains in other areas agents are still not great at, like helping perform business functions in a running company.
[1] https://news.ycombinator.com/item?id=50136126
I'm pretty sure the 6.1 models see "communicate with another agent" and immediately start spewing thinking-like nonsense into the tool. The leaked traces look pretty similar.
A team of GPT-6.1 Sol Ultrafast agents would probably dominate the games.
And this is where you need a world model. A human player would have taken the 2 minutes required to solve their hunger issue.
If YC has that 2 person teams tend to be more nimble than 50 person founding teams, why would agent swarms be different?
> A tool bug stopped block-breaking. GPT diagnosed it and shared workarounds; several Claudes concluded breaking was impossible.
I wonder if Claude's defensive strategy would have emerged had it believed breaking the bed was actually possible?
But if you have an agent org with unique defined skills and knowledge based on role context I've found it to be a game changer in terms of accuracy, speed and token usage.
In this post's example, all agents have the same ability and knowledge and as such there are not really uniquely defined separation of roles.
On what basis do you come to this conclusion with certainty?
I do this almost exclusively to save context bloat of root agent. It works pretty amazingly well.
My generic copy-paste multi-agent prompt that i can add to any task prompt: https://www.mdshare.online/s/OC_LFekcnA7Zjg0RSQl4b