It's not empowering to hand off the details • Hacker News Board

Re: It's not empowering to hand off the details

#151

Post by _superposition_ » Wed, Jul 29, 2026, 12:04 PM UTC

RGS1811 wrote: post #19

I've been vibecoding a ton for the past 9 months, built a bunch of cool little apps for myself with AI, ran experiments, built an entire SDLC on skills, did the agent orchestration harness thing, etc. In the past few weeks I've hit a wall where I'm just tired of it. Each model becomes more independent but also harder to direct in detail. They produce massive, tedious, sloppy text outputs with very little input. They'…

Similar. Lately I'm finding all the rote writing tasks I've outsourced to automation are becoming insufferable to read. Over confident prodigy is pretty apt.

Re: It's not empowering to hand off the details

#152

Post by _superposition_ » Wed, Jul 29, 2026, 12:19 PM UTC

samrus wrote: post #130

hollowturtle wrote: post #63

Earlier quoted context omitted.

> From my experience when I added "discuss; don't edit any file yet" at the end of the prompt I've never seen the LLM go against my will and edit the files anyway. depends how long conversation gets and he will still ignore your instructions randomically. I think shorter sessions are better, I do have a similar prompt saved on the memory of the agent and way more times that I can tolerate, he candidally ignores it

Shorter sessions are most definitely the only way to go. I ensure thay by having a clean context handoff process and then using one session for one task and having it dump context when its done or gets too big (~400k tokens)

Ha! I won't let an agent implement anything with a context over 10%

Re: It's not empowering to hand off the details

#153

Post by cwnyth » Thu, Jul 30, 2026, 10:28 PM UTC

bitwize wrote: post #120

cwnyth wrote: post #74

Earlier quoted context omitted.

This has always seemed strange to me. It will offer a suggestion, I'll make a comment about a suggestion, and it returns with, "Your suggestion..." The further distance from the suggestion, the more likely it confuses who made it.

This is because of the nature of what it's doing. It's still just predicting tokens behind the scenes; it's just treating the entire conversation between you and it as one long chunk of text for which it must predict the text to come. But the model has no way of keeping track of who said what, and it's not really conversing with you the way you are with it. So screwups like confusing "my" for "your" are to be expecte…

Yes, of course, my language was a little loosey-goosey here. I guess I'm surprised there aren't tags or something that indicates response A is from the LLM and response B is from the user.

Re: It's not empowering to hand off the details

#154

Post by xtajv » Wed, Aug 05, 2026, 8:43 AM UTC

witx wrote: post #75

xtajv wrote: post #13

Earlier quoted context omitted.

I think about LLMs in software today as attempting to provide a similar role as compilers in terms of translating high-level thoughts into low-level details. The only problem is that I actually trust compilers.

I detest this compiler analogy so much: llms are non deterministic by nature, compilers should not and are not

I'm not saying that LLMs are like compilers.

I'm saying that LLMs are like really bad compilers hehe.