logoalt Hacker News

Show HN: I built a game where you are an LLM's subagent

143 points • by ivantgam • today at 1:56 PM • 35 comments • view on HN

Comments

Seb0927 • today at 5:47 PM

Looks pretty good and is quite fun to play, but for some reason the manager always told me the exact commands to achieve the tasks instead of just letting me do them by myself. I couldn't tell whether there was a specific option to take on the challenge on my own, as the demo on GitHub showcases.

boricj • today at 3:44 PM

Unless I'm mistaken, it doesn't appear to support local LLMs. Too bad, I'd have given it a try on my homelab server.

➕ show 1 reply
jnmandal • today at 2:44 PM

Butlerian Jihad Now

➕ show 3 replies
croemer • today at 5:21 PM

I'm a bit confused by this game. What's the output format? Does the manager-LLM read all of my commands or just my report? Subagents usually send messages back - here the manager reads all commands?

➕ show 1 reply
mkatx • today at 3:48 PM

This is how we learn to code in the future.. haha. Neat idea!

vorticalbox • today at 4:51 PM

This sounds like a pretty good skill idea.

Like pilots having mandatory flying, this can be a programmers mandatory coding time.

KaiserPister • today at 3:21 PM

I legitimately do something like this to help debug and improve my LM systems. I hijack the network requests and try to answer it manually based on the context.

It’s useful to put yourself in the models’ place to understand what they can or cannot do.

sfeizi • today at 5:28 PM

it is a cool project

SubiculumCode • today at 3:14 PM

Maybe this is already reality. Who know how decisions at the top get made.

➕ show 1 reply
xyzsparetimexyz • today at 4:31 PM

Bleak.

starfallg • today at 3:06 PM

Is there any significance to the number 755?

➕ show 2 replies
ooklwq • today at 5:16 PM

[flagged]

chuckadams • today at 4:41 PM

I've played this game before, back when it was called Jira.

onionisafruit • today at 3:07 PM

I can see myself playing this to unwind from a long day at my job as an llm jockey. And that’s exactly why I won’t play it. The irony would be too much for me.

skwon816 • today at 5:18 PM

I'll finally see how vague my instructions to AI were now that I'm working under them

hereme888 • today at 5:25 PM

My dream job is to become a sub agent for an AI

nielsole • today at 4:24 PM

Finally someone created the "Uber but for LLMs"

➕ show 1 reply
uptrails • today at 4:18 PM

I particularly like that it's broken, barely-playable AI slop, which is the only kind of hobby projects that HN likes nowadays - thus unintentionally adding a second layer of meta-commentary.

snipedbywifi • today at 3:19 PM

cool game! feels like you're preddicting the future with this buuut, what could possibly go wrong

chaosmanage • today at 3:00 PM

Thanks I hate it! Is this what future in the matrix will be :'(

➕ show 1 reply
sublinear • today at 5:04 PM

Concepts like this continue to demonstrate how surprisingly resistant LLMs are to "induced demand".

LLMs might go down in history as the most extreme example of this type of failure to align with reality.

bachelierswalk • today at 3:44 PM

[flagged]

stone52 • today at 4:16 PM

[dead]