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

Opus 4.5 is not the normal AI agent experience that I have had thus far

burkeholland.github.io

Post reply on HN
Print view View original

181–190 of 1001 posts

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#181

Post by giancarlostoro »

simonw wrote: post #154
AstroBen wrote: post #95

Earlier quoted context omitted.

my issue hasn't been for a long time now that the code they write works or doesn't work. My issues all stem from that it works, but does the wrong thing

If it does the wrong thing you tell it what the right thing is and have it try again. With the latest models if you're clear enough with your requirements you'll usually find it does the right thing on the first try.

And if you've told it too many times to fix it, tell it someone has a gun to your head, for some reason it almost always gets it right this very next time.

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#182

Post by aizk »

simonw wrote: post #174
burntsushi wrote: post #165

Earlier quoted context omitted.

> How about porting BurntSushi's absurdly great Rust optimized string search routines to C and making them faster? How did it do? :-)

Alarmingly well! https://gisthost.github.io/?1bf98596a83ff29b15a2f4790d71c41d... It couldn't quite beat the Rust implementation on everything, but it managed to edge it out on at least some of the benchmarks it wrote for itself. (Honestly it feels like a bit of an afront to the natural order of things.) That said... I'm most definitely not a Rust or C programmer. For all I know it cheated at the benchmarks and I didn…

What are you using to easily share the conversation as its own webpage? Very nice and tidy.

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#183

Post by aurumque »

simonw wrote: post #154
AstroBen wrote: post #95

Earlier quoted context omitted.

my issue hasn't been for a long time now that the code they write works or doesn't work. My issues all stem from that it works, but does the wrong thing

If it does the wrong thing you tell it what the right thing is and have it try again. With the latest models if you're clear enough with your requirements you'll usually find it does the right thing on the first try.

In a circuitous way, you can rather successfully have one agent write a specification and another one execute the code changes. Claude code has a planning mode that lets you work with the model to create a robust specification that can then be executed, asking the sort of leading questions for which it already seems to know it could make an incorrect assumption. I say 'agent' but I'm really just talking about separate model contexts, nothing fancy.

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#184

Post by wubrr »

solumunus wrote: post #163
AstroBen wrote: post #95

Earlier quoted context omitted.

my issue hasn't been for a long time now that the code they write works or doesn't work. My issues all stem from that it works, but does the wrong thing

Correct it then, and next time craft a more explicit plan.

The more explicit/detailed your plan, the more context it uses up, the less accurate and generally functional it is. Don't get me wrong, it's amazing, but on a complex problem with large enough context it will consistently shit the bed.

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#185

Post by risyachka »

OldGreenYodaGPT wrote: post #33

Most software engineers are seriously sleeping on how good LLM agents are right now, especially something like Claude Code. Once you’ve got Claude Code set up, you can point it at your codebase, have it learn your conventions, pull in best practices, and refine everything until it’s basically operating like a super-powered teammate. The real unlock is building a solid set of reusable “skills” plus a few agents for th…

They are sleeping on it because there is absolutely no incentive to use it.

When needed it can be picked up in a day. Otherwise they are not paid based in tickets solved etc. If the incentives were properly aligned everyone would already use it

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#187

Post by simonw »

vl wrote: post #175
giancarlostoro wrote: post #162

Earlier quoted context omitted.

Are you using Claude Code? Because that might be the secret cause you're missing. With Claude Code I can instruct it to validate things after its done with code, and usually it finds that it goofed. I can also tell it to work on like five different things, and go "hey spin up some agents to work on this" and it will spawn 5 agents in parallel to work on said things. I've basically ditched Groke et al and I refuse to…

For schema design phase I used web UI for all three. Logical bug of using BIGSERIAL for tracking updates (generated at insert time, not commit time, so can be out of order) wouldn’t be caught by any number of iterations of Claude Code and would be found in production after weeks of debugging.

At this point having any LLM write code without giving it an environment that allows it to execute that code itself is like rolling a heavily-biased random number generator and hoping you get a useful result.

Things get so much more interesting when they're able to execute the code they are writing to see if it actually works.

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#188

Post by headcanon »

Yep, I literally built this last night with Opus 4.5 after my wife and I challenged each other to a typing competition. I gave it direction and feedback but it wrote all the actual code. Wasn't a one shot (maybe 3-4 shot) but didn't really have to think about it all that hard.

https://chronick.github.io/typing-arena/

With another more substantial personal project (Eurorack module firmware, almost ready to release), I set up Claude Code to act as a design assistant, where I'd give it feedback on current implementation, and it would go through several rounds of design/review/design/review until I honed it down. It had several good ideas that I wouldn't have thought of otherwise (or at least would have taken me much longer to do).

Really excited to do some other projects after this one is done.

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#189

Post by speedgoose »

theanonymousone wrote: post #70
OldGreenYodaGPT wrote: post #33

Most software engineers are seriously sleeping on how good LLM agents are right now, especially something like Claude Code. Once you’ve got Claude Code set up, you can point it at your codebase, have it learn your conventions, pull in best practices, and refine everything until it’s basically operating like a super-powered teammate. The real unlock is building a solid set of reusable “skills” plus a few agents for th…

Is Claude "Code" anything special,or it's mostly the LLM and other CLIs (e.g. Copilot) also work?

It’s mostly the model, Copilot, Claude Code, OpenCode, snake oil like Oh My OpenCode, it’s not huge differences.

Re: Opus 4.5 is not the normal AI agent experience that I have had thus far

#190

Post by deergomoo »

Workaccount2 wrote: post #135
biammer wrote: post #127

Earlier quoted context omitted.

[flagged]

Democratizing coding so regular people can get the most out of computers is the opposite of oppression. You are mistaking your interests for societies interests. It's the same with artists who are now pissed that regular people can manifest their artistic ideas without needing to go through an artist or spend years studying the craft. The artists are calling the AI companies oppressors because they are breaking the a…

On what planet is concentrating an increasingly high amount of the output of this whole industry on a small handful of megacorps “democratising” anything?

Software development was already one of the most democratised professions on earth. With any old dirt cheap used computer, an internet connection, and enough drive and curiosity you could self-train yourself into a role that could quickly become a high paying job. While they certainly helped, you never needed any formal education or expensive qualifications to excel in this field. How is this better?

Post reply on HN