601–610 of 819 posts
Earlier quoted context omitted.
It still says "Perfect!" about its own work far too often.
In fairness that sounds like me when I code. It's either "Perfect!" or "Genius!". Or conversely "I'm a complete idiot!"
Wonder if I could Claude to do that
Earlier quoted context omitted.
Why is this getting downvoted? It was hilarious! I actually added a fun thing to my user-wide CLAUDE.md, basically saying that it should come up with a funny insult every time I come up with an idea that wasn't technically sound (I got the prompt from someone else). It seems to be disobeying me, because I refuse to believe that I don't have bad ideas. Or some other prompt is overriding it.
This is brilliant! Can you give me some pointers? Ie : if I make a request that seems dumb tell me custom instruction?
What I'm not liking is it seems even... lazier... than previously. By which I mean the classic "This is getting complicated so..." (followed by cop-out, dropping the original task and motivation).
There's also a bug where compaction becomes impossible. ("conversation too long" and its advice on how to fix doesn't work)
So far I'm liking that it seems to follow my CLAUDE.md instructions better, doing more frequent checkins with me to ask me to review what it's done, etc, and taking my advice more. What I'm not liking is it seems even... lazier... than previously. By which I mean the classic "This is getting complicated so..." (followed by cop-out, dropping the original task and motivation). There's also a bug where compaction become…
I have seen this issue with every model so far
Earlier quoted context omitted.
Have you thought about benchmarking models a month or two after release to see how it competes vs the day 1 release
For that to be useful I'd need to be running much better benchmarks - anything less than a few hundred numerically scored tasks would be unlikely to reliably identity differences. An organization like Artificial Analysis would be a better fit for that kind of investigation: https://artificialanalysis.ai/
From https://news.ycombinator.com/item?id=40859434 :
> E.g promptfoo and chainforge have multi-LLM workflows.
> Promptfoo has a YAML configuration for prompts, providers,: https://www.promptfoo.dev/docs/configuration/guide/
openai/evals//docs/build-eval.md: https://github.com/openai/evals/blob/main/docs/build-eval.md
From https://news.ycombinator.com/item?id=45267271 ;
> API facades like OpenLLM and model routers like OpenRouter have standard interfaces for many or most LLM inputs and outputs. Tools like Promptfoo, ChainForge, and LocalAI also all have abstractions over many models.
> What are the open standards for representing LLM inputs, and outputs?
> W3C PROV has prov:Entity, prov:Activity, and prov:Agent for modeling AI provenance: who or what did what when.
> LLM evals could be represented in W3C EARL Evaluation and Reporting Language
"Can Large Language Models Emulate Judicial Decision-Making? [Paper]" https://news.ycombinator.com/item?id=42927611
"California governor signs AI transparency bill into law" (2025) https://news.ycombinator.com/item?id=45418428 :
Is this the first of its sort?:
> CalCompute
Price is playing a big role in my AI usage for coding. I am using Grok Code Fast as it's super cheap. Next to it GPT-5 Codex. If you are paying for model use out of pocket Claude prices are super expensive. With better tooling setup those less smart (and often faster) models can give you better results. I am going to give this another shot but it will cost me $50 just to try it on a real project :(
I'm paying $90(?) a month for the Max and it holds up for about an hour or so of in depth coding before it kicks in the 5-hour window lockout (so effectively about 4 hours of time when I can't run it). Kinda frustrating, even with efficient prompt and context length conservation techniques. I'm going to test this new sonnet 4.5, now but it'll probably be just as quick to gobble my credits.
Earlier quoted context omitted.
Gemini 2.5 Pro = Long context king, image input king GPT-5 = Overengineering/complexity/"enterprise" king Claude = "Get straightforwaed shit done efficiently" king
On the plus side, GPT5 is very malleable, so you CAN prompt it away from that, whereas it's very hard to prompt Claude into producing hard code: even with a nearly file by file breakdown of a task, it'll occasionally run into an obstacle and just give up and make a mock or top implementation, basically diverge from the entire plan, then do its own version.
"Differentiate y(x) w.r.t x, gamma and beta."
It not only produced the correct result, but it understood the context - I didn't tell it the context was layer norm, back-propagation and matrices.
This release is a step function for my use cases.
My screenshot came from here: https://docs.pytorch.org/docs/stable/generated/torch.nn.Laye...
Price is playing a big role in my AI usage for coding. I am using Grok Code Fast as it's super cheap. Next to it GPT-5 Codex. If you are paying for model use out of pocket Claude prices are super expensive. With better tooling setup those less smart (and often faster) models can give you better results. I am going to give this another shot but it will cost me $50 just to try it on a real project :(