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

Eight more months of agents

crawshaw.io

Post reply on HN
Print view View original

111–120 of 250 posts

Re: Eight more months of agents

#111

Post by baq »

skydhash wrote: post #98
bitwize wrote: post #69

Earlier quoted context omitted.

But there's a difference. Being opposed to AI-generated art/music/writing is valid because humans still contribute something extraordinarily meaningful when they do it themselves. There's no market for AI-generated music, and AI-generated art and writing tends to get called out right away when it's detected. People want the human expression in human-generated art, and the AI stuff is a weak placeholder at best. For s…

Ok, I’ll bite. What’s there to learn that you can tie directly to an increase of productivity? I can say “learn how to use vim makeprg feature so that you can jump directly to errors reported by the build and tool” and it’s very clear where the ROI. But all the AI hypers are selling are hope, prayers, and rituals.

I’m not an AI hyper, I just don’t code manually anymore. Tickets take about as much time to close as before, but the code shipped now has higher test coverage, higher performance, better concurrency error handling, less follow-up refactor PRs, less escapes to staging/prod and better documentation; some of it is now also modeled in a model checker.

Re: Eight more months of agents

#112

Post by skydhash »

baq wrote: post #99
skydhash wrote: post #90

Earlier quoted context omitted.

It doesn’t matter how fast you run if it’s not the correct direction.

Good LLM wielders run in widening circles and get to the goal faster than good old school programmers running in a straight line

I still haven’t seen those mythical LLM wielders in the wild. While I’m using tools like curl, jq, cmus, calibre, openbsd,… that has been most certainly created by those old school programmers.

Re: Eight more months of agents

#113

Post by pseudosavant »

baq wrote: post #94
hoistbypetard wrote: post #93

> To me that statement is as obvious as "water is wet". Water is not wet. Water makes things wet. Perhaps the inaccuracy of that statement should be taken as a hint that the other statements that you hold on the same level are worthy of reconsideration.

The good old classic technically correct and completely besides the point observation.

Yes, technically HN is full of these kinds of corrections... but HN isn't actually wet.

Re: Eight more months of agents

#114

Post by skydhash »

baq wrote: post #111
skydhash wrote: post #98

Earlier quoted context omitted.

Ok, I’ll bite. What’s there to learn that you can tie directly to an increase of productivity? I can say “learn how to use vim makeprg feature so that you can jump directly to errors reported by the build and tool” and it’s very clear where the ROI. But all the AI hypers are selling are hope, prayers, and rituals.

I’m not an AI hyper, I just don’t code manually anymore. Tickets take about as much time to close as before, but the code shipped now has higher test coverage, higher performance, better concurrency error handling, less follow-up refactor PRs, less escapes to staging/prod and better documentation; some of it is now also modeled in a model checker.

So the code was an unknown (to the world) X quality, but now it’s X+k quality? How does that help me exactly?

Re: Eight more months of agents

#115

Post by swyx »

baq wrote: post #94
hoistbypetard wrote: post #93

> To me that statement is as obvious as "water is wet". Water is not wet. Water makes things wet. Perhaps the inaccuracy of that statement should be taken as a hint that the other statements that you hold on the same level are worthy of reconsideration.

The good old classic technically correct and completely besides the point observation.

username checks out

Re: Eight more months of agents

#117

Post by 0xbadcafebee »

Local models are decent now. Qwen3 coder is pretty good and decent speed. I use smaller models (qwen2.5:1.5b) with keyboard shortcuts and speech to text to ask for man page entries, and get 'em back faster than my internet connection and a "robust" frontier model does. And web search/RAG hides a multitude of sins.

"Using anything other than the frontier models is actively harmful" - so how come I'm getting solid results from Copilot and Haiku/Flash? Observe, Orient, Decide, Act, Review, Modify, Repeat. Loops with fancy heuristics, optimized prompts, and decent tools, have good results with most models released in the past year.

Re: Eight more months of agents

#118

Post by uludag »

baq wrote: post #99
skydhash wrote: post #90

Earlier quoted context omitted.

It doesn’t matter how fast you run if it’s not the correct direction.

Good LLM wielders run in widening circles and get to the goal faster than good old school programmers running in a straight line

I try to avoid LLMs as much as I can in my role as SWE. I'm not ideologically opposed to switching, I just don't have any pressing need.

There are people I work with who are deep in the AI ecosystem and it's obvious what tools they're using It would not be uncharitable in any way to characterize their work as pure slop that doesn't work, buggy, untested adequately, etc.

The moment I start to feel behind I'll gladly start adopting agentic AI tools, but as things stand now, I'm not seeing any pressing need.

Comments like these make me feel like I'm being gaslit.

Re: Eight more months of agents

#119

Post by dmos62 »

gip wrote: post #110
dmos62 wrote: post #107

Earlier quoted context omitted.

Curious what kind of agent did you build? I'm building a programming agent myself, it's intentionally archaic in that you run it by constantly copy-pasting from-to fresh ChatGPT sessions (: I'm finding it challenging to have it do good context management: I'm trying to solve this by declaring parts of code or spec as "collections" with an overview md file attached that acts like a map of why/where/what, but that can'…

Send an DM on twitter to @edfixyz (one of my account) and I'll reply with a link to the website tomorrow to give you a sense. Can't share a link here, it will kill my backend.

Care to send me a link on my email? It's in my about. I don't use my X account and can't seem to login, the verification SMS never arrives.
Post reply on HN