HNHacker News
TopNewBestAskShowJobs

rco8786

9,214 karma · joined December 13, 2010

submissionscomments
rco8786··on I made ChatGPT and Google tell I'm a competitive hot-dog-eating world champion
> Has anyone seen a production system that actually does claim-level verification before generation?

"Claim level" no, but search engines have been scoring sources on reliability and authority for decades now.

rco8786··on Anthropic tries to hide Claude's AI actions. Devs hate it
> that will go live very soon.

I hope that you are successful!

rco8786··on Anthropic tries to hide Claude's AI actions. Devs hate it
We'll have to solve for that sometime soon-ish I think. Claude Code has at least some sort of token estimation built-in to it now. I asked it to kick off a large agent team (~100 agents) to rewrite a bunch of SQL queries, one per agent. It did the first 10 or so, then reported back that it would cost too much to do it this way...so it "took the reins" without my permission and tried to convert each query using only the main agent and abandoned the teams. The results were bad.

But in any case, we're definitely coming up on the need for that.

rco8786··on Anthropic tries to hide Claude's AI actions. Devs hate it
Can you explain more? I'm talking about LLM/agent behavior in a generalized sense, even though I used claude code as the example here.

What is Codex doing differently to solve for this problem?

rco8786··on Anthropic tries to hide Claude's AI actions. Devs hate it
The corrective agent has the exact same percentage chance at making the mistake. "Correcting" an assumption that was previously correct into an incorrect one.

If a singular agent has a 1% chance of making an incorrect assumption, then 10 agents have that same 1% chance in aggregate.

rco8786··on Anthropic tries to hide Claude's AI actions. Devs hate it
Totally agreed. Those assumptions often compound as well. So the AI makes one wrong decision early in the process and it affects N downstream assumptions. When they finally finish their process they've built the wrong thing. This happens with one process running. Even on latest Opus models I have to babysit and correct and redirect claude code constantly. There's zero chance that 5 claude codes running for hours without my input are going to build the thing I actually need.

And at the end of the day it's not the agents who are accountable for the code running in the production. It's the human engineers.

rco8786··on Anthropic tries to hide Claude's AI actions. Devs hate it
That is also my experience. Doesn't even have to be a 10 year old codebase. Even a 1 year old codebase. Any one that is a serious product that is deployed in production with customers who rely on it.

Not to say that there's no value in AI written code in these codebases, because there is plenty. But this whole thing where 6 agents run overnight and "tada" in the morning with production ready code is...not real.

rco8786··on Anthropic tries to hide Claude's AI actions. Devs hate it
> it's about catching when it goes off the rails before it makes a mess

The latest "meta" in AI programming appears to be agent teams (or swarms or clusters or whatever) that are designed to run for long periods of time autonomously.

Through that lens, these changes make more sense. They're not designing UX for a human sitting there watching the agent work. They're designing for horizontally scaling agents that work in uninterrupted stretches where the only thing that matters is the final output, not the steps it took to get there.

That said, I agree with you in the sense that the "going off the rails" problem is very much not solved even on the latest models. It's not clear to me how we can trust a team of AI agents working autonomously to actually build the right thing.

rco8786··on Building SQLite with a small swarm
Not a single comment about whether it actually works or not?
rco8786··on GPT-5 outperforms federal judges in legal reasoning experiment
I don't know if I'm comfortable with any of this at all, but seems like having AI do "front line" judgments with a thinner appeals layer available powered by human judges would catch those edge cases pretty well.
rco8786··on Show HN: Total Recall – write-gated memory for Claude Code
Just +1ing on the readme. It's clearly an LLM dump that repeats a lot of the same points multiple times and is missing a very, very important section:

How to use Total Recall. It explains all of the concepts, but lacks a "quick start" section for someone who is ready to get up and running.

rco8786··on MIT Technology Review has confirmed that posts on Moltbook were fake
Not OP. But embeddings are the internal matrix representations of tokens that LLMs use to do their work. If tokens are the native language that humans use, embeddings are the native language that LLMs use.

OP, I think, is saying that once LLMs start communicating natively without tokens is when they shed the need for humans or human-level communication.

Not sure I 100% agree, because embeddings from one LLM are not (currently) understood by another LLM and tokens provide a convenient translation layer. But I think there's some grain of truth to what they're saying.

rco8786··on LLMs as the new high level language
The thing that’s always missing from these critiques isnt about code quality or LoC or slop.

The issue is that if you fire off 10 agents to work autonomously for an extended period of time at least 9 of them will build the WRONG THING.

The problem is context management and decision making based on that context. LLMs will always make assumptions about what you want, and the more assumptions they make the higher the likelihood that one or more of them is wrong.

rco8786··on Outsourcing thinking
I didn't mean to imply otherwise. Only to point out that in the early days of the internet, even into the 00s, teachers had a Hard No rule on any internet source.

I graduated high school in '04 and even then I was only allowed to use this system called "Galileo" which was basically a curated listed of encyclopedic articles specifically meant for education and research.

rco8786··on We tasked Opus 4.6 using agent teams to build a C Compiler
> Claude will work autonomously to solve whatever problem I give it. So it’s important that the task verifier is nearly perfect, otherwise Claude will solve the wrong problem.

I think this is the fundamental thing here with AI. You can spin up infinite agents that can all do....stuff. But how do you keep them from doing the wrong stuff?

Is writing an airtight spec and test harness easier or less time consuming than just keeping a human in the loop and verifying and redirecting as the agents work?

It all still comes back to context management.

Very cool demonstration of the tech though.

rco8786··on Agent Skills
Yep. I have an incredibly hard time getting them to use Skills at all, even when asked.

I saw someone's analysis a few days ago and they found that their agents were more accurate when just dumping the skill context directly into AGENTS.md

rco8786··on Outsourcing thinking
This implies that people had better mathematical intuition, on average, pre calculator which seems difficult to believe.
rco8786··on Outsourcing thinking
Right so apply the exact same logic to LLMs as you did to the internet.

At first the internet was unreliable. Nobody could trust the information it gave you. So teachers insisted that students only use their trusted sources. But eventually the internet matured and now it would be seen as ridiculous for a teacher to tell a student not to do research on the internet.

Now replace "the internet" with "LLMs".

rco8786··on Outsourcing thinking
I'll say that I'm still kinda on the fence here, but I will point out that your argument is exactly the same as the argument against calculators back in the 70s/80s, computers and the internet in the 90s, etc.
rco8786··on Vitamin D and Omega-3 have a larger effect on depression than antidepressants
I don't relate at all to the latter part of your question, so by process of elimination it must be the former :)
rco8786··on Vitamin D and Omega-3 have a larger effect on depression than antidepressants
> except it killed my libido

Similar experience. Apparently pretty much ubiquitous with SSRIs

rco8786··on Show HN posts p/month more than doubled in the last year
The Show HN prefix is submitting it that way.

> To post [to Show HN], submit a story whose title begins with "Show HN".

rco8786··on Get-Shit-Done
This looks interesting, anyone tried it yet?
rco8786··on Claude Code's new hidden feature: Swarms
Is this significantly different that the subagents that are already in CC?
rco8786··on Unrolling the Codex agent loop
Think of it as Just-In-Time context injection/enhancement
rco8786··on Unrolling the Codex agent loop
That’s a separate tool though. You don’t want to have to open another terminal to git diff every 30 seconds and then give feedback. Much better UX when it’s inline.

My main hooks are desktop notifications when Claude requires input or finishes a task. So I can go do other things while it churns and know immediately when it needs me.

rco8786··on Scott Adams has died
I had this exact experience. Growing up I had nothing but good memories of reading Dilbert over my breakfast cereal, and then laughing as I got into the workforce and realized how accurate the satire was. And then seeing what "he" was actually like just completely threw me for a loop.
rco8786··on Code and Let Live
This is cool but like...a docker container running locally works fine too?
rco8786··on Google: Don't make "bite-sized" content for LLMs
So this article itself is literally content chunking.

> So you end up with short paragraphs, sometimes with just one or two sentences

The average number of sentences per paragraph in the article is... 2.4

rco8786··on Eat Real Food
Most of it seems fine, although eating even more meat than we already do is a bit perplexing.

The new "guidelines" for alcohol are pretty laughable though. I say that as someone who enjoys his fair share of beers. “The implication is don’t have it for breakfast," <- direct quote from celebrity Dr Oz during the press conference.

← PreviousPage 4 of 34Next →