Hacker News Reader

Comment by madrox | Hacker News Reader

Having spent the last three months with Agents I’m starting to develop a saying: “Any sufficiently advanced agent is indistinguishable from a DSL”

I’m becoming convinced Agents are mostly novelty. They aren’t Turing complete. They’re fun to implement, but any algorithm they run to solve a task can’t be inspected and can go awry with complex logic. You can accomplish the same thing by asking the LLM to write code in whatever language, and it’s more reliable. There is more reluctance to do that though, because the idea of a BE environment that runs user generated scripts is scary.

View on HN

What have you been trying to use agents for?

I think most of the actual value created short/medium-term with agents will be in entertainment where things going awry is OK and no action they take can really be catastrophic.

Some people are already spending a ton of time on character.ai (https://beta.character.ai). There are sims-like games like ai-town (https://www.convex.dev/ai-town). My friend and I made an open-source murder mystery novella with agents (https://gron.games).

We found that a lot of the actions they take are pretty low stakes, so the correctness issues are much less important (in ai-town, they just talk with each other, in our game, they just beat you or run away if you interrogate them poorly).

My current work is using LLMs in fintech, but my background is in entertainment and has been where most of side projects have gone. LLMs are great generative tools, but Agents are reasoning tools. They implicitly execute an algorithm and interpret the results. An agent will never reason as well as imperative code.

I think this is why agents appear suited to entertainment, but this is almost a different topic that I could rant about for hours. There’s a whimsical element of randomness in what the character may do that the programmer does not need to explicitly design. However, I think that whimsy has limited reach, so while the developer may be more amused by their own creation it has limited appeal.

Curious how you went with this approach.

I too have been playing with agents. But I’ve explicitly biased away from having them do things that can achieved directly with code.

Stringing together tool selection into steps with reasoning imo is the crucial brain layer. Code is one of the tools. I’ve also gone with a kind of preemptive memory, where the brain gets injected with relevant memory automatically, instead of it requesting for it. Same for retrieval from the internet.

But wanted to understand how you’re thinking about.

You can string tools together in code as well, and have an LLM reinterpret the final results in text if you want. There’s the middle bit when you ask the Agent to reason that works better if you just ask it to write code and give it functions to call instead of tools.

Reply on news.ycombinator.com