Are LLM merge rates not getting better? • Hacker News Board

Re: Are LLM merge rates not getting better?

#171

Post by pornel » Sat, Mar 14, 2026, 5:34 PM UTC

zx8080 wrote: post #149

aerhardt wrote: post #61

I feel that two things are true at the same time: 1) Something happened during 2025 that made the models (or crucially, the wrapping terminal-based apps like Claude Code or Codex) much better. I only type in the terminal anymore. 2) The quality of the code is still quite often terrible. Quadruple-nested control flow abounds. Software architecture in rather small scopes is unsound. People say AI is “good at front end”…

> People say AI is “good at front end” but I see the worst kind of atrocities there It's commonly universal to say "AI is great in X", where one is not professional in X. It's because that's how AI is designed: to output tokens according to stats, not logic, not semantic, and not meaning: stats.

Gell-Mann Amnesia for code quality.

Re: Are LLM merge rates not getting better?

#172

Post by jmalicki » Sun, Mar 15, 2026, 12:59 AM UTC

0x3f wrote: post #160

jmalicki wrote: post #156

Earlier quoted context omitted.

You can see it happen in reverse with X/Twitter. Did reducing waste affect the user experience or uptime of Twitter? Not really. But advertising revenues plummeted, because those extra employees were mostly not about the user experience or keeping the website up, they were about servicing the advertisers that brought the company revenue.

I thought advertising revenues plummeted mostly for content/optics/PR reasons, not ad-buyer-facing feature reasons.

Content moderation was an ad-buyer-facing feature. I really see no evidence Musk actually understood that. When he took it away he was all like surprised Pikachu face that advertisers left.

Re: Are LLM merge rates not getting better?

#173

Post by tokenbar » Sun, Mar 15, 2026, 2:16 PM UTC

butILoveLife wrote: post #97

AussieWog93 wrote: post #43

Earlier quoted context omitted.

I'd agree with you on 4.5 to 4.6, but going from gpt-5 or 4.0 to 4.5 was night and day.

GPT5 added the router, which was def a downgrade. 4.5 was probably the best non-COT model humanity has made. But too expensive to run.

[dead]

Re: Are LLM merge rates not getting better?

#174

Post by dennarucker » Tue, Mar 17, 2026, 4:17 AM UTC

[dead]

Re: Are LLM merge rates not getting better?

#175

Post by techcam » Tue, Mar 17, 2026, 5:59 PM UTC

I’ve been noticing the same — a lot of failures aren’t obvious “jailbreaks,” they’re just subtle prompt structure issues that only show up in production.