Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

LLMs corrupt your documents when you delegate

arxiv.org

Post reply on HN
Print view View original

191–200 of 235 posts

Re: LLMs corrupt your documents when you delegate

#191

Post by deadbabe »

ofjcihen wrote: post #116
deadbabe wrote: post #65

Earlier quoted context omitted.

> we who use them a lot also get more proficient and learn where they can be used effectively vs not I think it’s always been obvious where an LLM could be used effectively and where it cannot, if you understand how they work and don’t see them as magical. The “increase in proficiency” is mostly people coming back to reality and being more intentional about LLM usage. There are no surprise discoveries here. One does…

I think you hit the nail on the head. I had been in this space for a little bit before it really became popular. I haven’t seen incredible gains in model competency. What I have seen though is people figuring out what works and what doesn’t.

It’s pretty telling that ignoring LLMs entirely for a few years and then jumping in last minute after everyone has struggled through figuring out how to use them still puts you on the same level very quickly.

Re: LLMs corrupt your documents when you delegate

#193

Post by handoflixue »

> Our main experiment is a round-trip relay with N = 10 consecutive round-trips per environment, simulating 20 delegated interactions. In each interaction, the model receives all work environment documents as text in its context window in a single turn

The LLM isn't being given an actual file system they can work with - they're expected to receive the document as text in the prompt, perform a task, and then re-output text into the conversation?

Maybe I'm misunderstanding the methodology, but this feels a lot like the human game of Telephone - or perhaps, asking one to do a similar editing task using only Microsoft Outlook with copy/paste disabled.

I'd imagine that one gets radically different results if one uses the appropriate desktop tools, just like humans do much better outside games of Telephone.

Re: LLMs corrupt your documents when you delegate

#194

Post by embedding-shape »

deadbabe wrote: post #191
ofjcihen wrote: post #116

Earlier quoted context omitted.

I think you hit the nail on the head. I had been in this space for a little bit before it really became popular. I haven’t seen incredible gains in model competency. What I have seen though is people figuring out what works and what doesn’t.

It’s pretty telling that ignoring LLMs entirely for a few years and then jumping in last minute after everyone has struggled through figuring out how to use them still puts you on the same level very quickly.

> then jumping in last minute after everyone has struggled through figuring out how to use them still puts you on the same level very quickly

Does it actually though?

I've used agents for quite some time now, if someone who never used agents before want to put this to the test somehow, I'm open to try to measure this, reach out via email :)

Re: LLMs corrupt your documents when you delegate

#195

Post by fluidcruft »

wfurney wrote: post #185
TeMPOraL wrote: post #179

Earlier quoted context omitted.

Most people also understand that, because they're not "frequent" users of a thing, they absolutely suck at using it, and set their expectations accordingly. In particular, they realize that doing anything non-trivial with the thing requires them to spend some learning and practice time, or asking/hiring a "frequent" user to do it for them. So the reasonable response to being told you're holding your scissors wrong is…

I’m interested in the “non-trivial” point as well, this seems to be a common refrain from the anti-LLM tech crowd, “LLMs aren’t good at doing anything non-trivial”, well is that really the case or is it just harder and one needs to put in more practice for more complicated tasks? I don’t have an example off hand, but I know that it’s easy to dismiss something an LLM does as trivial if your work is extremely marginal.…

On the one hand there's simonw's concept of a "frequent LLM user" and then there's the actual vast majority of people using ChatGPT web app or one of the various Office CoPilots.

Re: LLMs corrupt your documents when you delegate

#199

Post by nullsanity »

kristjansson wrote: post #117
simonw wrote: post #20

I'm suspicious of their results with regards to tool usage. It's unsurprising that round-tripping long content through an LLM results in corruption. Frequent LLM users already know not to do that. They claim that tool use didn't help, which surprised me... but they also said: > To test this, we implemented a basic agentic harness (Yao et al., 2022) with file reading, writing, and code execution tools (Appendix M). We…

Only sort of related, but I would love to see a harness with ed as the primary file editing / reading tool. Half the bash Claude runs seems to be sed anyway, having some state persist in ed would seem to help. What does one do when a full editor consumes too much bandwidth^H tokens? Use ed, the standard editor!

I'm not sure you understand how those terminal programs are rendered - but the amount of control code data sent to Claude would be way, way more than using command line sed.

Re: LLMs corrupt your documents when you delegate

#200

Post by emp17344 »

ACCount37 wrote: post #118
tommyage wrote: post #82

Earlier quoted context omitted.

All of software is hard-coded algorithm. If you differ between AI source code and engineer source code say so. "Getting things done" is a business need. Which things get translated to a deterministic language executable by a computer is code. There are entire languages dedicated for lesser engineers/domain experts to formulate business requirements. Anyhow; What's your point? That we received a framework for "soft al…

AI is more "grown" than it is "hard-coded". It's sideways to normal software - the way DSP is sideways to normal software but somehow even worse. It all comes down to hard logic eventually, but that "eventually" has teeth. None of the interesting behaviors of AI systems live in "engine.py". My point is: there are tasks where the choices are to use AI, use a meatbag, or suck forever. The "use AI" option going to be fl…

The contempt you people have for your fellow human beings is palpable. Your point is obfuscated because your objective seems to be to degrade others.
Post reply on HN