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

Claude Opus 4.6

anthropic.com

Post reply on HN
Print view View original

341–350 of 1001 posts

Re: Claude Opus 4.6

#341

Post by fragmede »

stkai wrote: post #232
simonw wrote: post #40

The bicycle frame is a bit wonky but the pelican itself is great: https://gist.github.com/simonw/a6806ce41b4c721e240a4548ecdbe...

Would love to find out they're overfitting for pelican drawings.

The estimation I did 4 months ago:

> there are approximately 200k common nouns in English, and then we square that, we get 40 billion combinations. At one second per, that's ~1200 years, but then if we parallelize it on a supercomputer that can do 100,000 per second that would only take 3 days. Given that ChatGPT was trained on all of the Internet and every book written, I'm not sure that still seems infeasible.

https://news.ycombinator.com/item?id=45455786

Re: Claude Opus 4.6

#344

Post by skydhash »

sweetheart wrote: post #112
gjsman-1000 wrote: post #20

Earlier quoted context omitted.

Also explains why Claude Code is a React app outputting to a Terminal. (Seriously.)

React's core is agnostic when it comes to the actual rendering interface. It's just all the fancy algos for diffing and updating the underlying tree. Using it for rendering a TUI is a very reasonable application of the technology.

The terminal UI is not a tree structure that you can diff. It’s a 2D cells of characters, where every manipulation is a stream of texts. Refreshing or diffing that makes no sense.

Re: Claude Opus 4.6

#345

Post by jmalicki »

Der_Einzige wrote: post #209
jmalicki wrote: post #109

Earlier quoted context omitted.

A ton of GPU kernels are hugely inefficient. Not saying the numbers are realistic, but look at the 100s of times of gain in the Anthropic performance takehome exam that floated around on here. And if you've worked with pytorch models a lot, having custom fused kernels can be huge. For instance, look at the kind of gains to be had when FlashAttention came out. This isn't just quantization, it's actually just better op…

"This isn't X, it's Y" with extra steps.

I'm flattered you think I wrote as well as an AI.

Re: Claude Opus 4.6

#346

Post by aggie »

legitster wrote: post #199
lukebechtel wrote: post #146

Earlier quoted context omitted.

Why? I use it for all and love it. That doesn't mean you have to, but I'm curious why you think it's behind in the personal assistant game.

I have three specific use cases where I try both but ChatGPT wins: - Recipes and cooking: ChatGPT just has way more detailed and practical advice. It also thinks outside of the box much more, whereas Claude gets stuck in a rut and sticks very closely to your prompt. And ChatGPT's easier to understand/skim writing style really comes in useful. - Travel and itinerary: Again, ChatGPT can anticipate details much more, an…

How does #3 square with Anthropic's literal warehouse full of books we've seen from the copyright case? Did OpenAI scan more books? Or did they take a shadier route of training on digital books despite copyright issues, but end up with a deeper library?

Re: Claude Opus 4.6

#347

Post by elAhmo »

ifwinterco wrote: post #340
wasmainiac wrote: post #265

Earlier quoted context omitted.

Dumb question. Can these benchmarks be trusted when the model performance tends to vary depending on the hours and load on OpenAI’s servers? How do I know I’m not getting a severe penalty for chatting at the wrong time. Or even, are the models best after launch then slowly eroded away at to more economical settings after the hype wears off?

On benchmarks GPT 5.2 was roughly equivalent to Opus 4.5 but most people who've used both for SWE stuff would say that Opus 4.5 is/was noticeably better

I mostly used Sonnet/Opus 4.x in the past months, but 5.2 Codex seemed to be on par or better for my use case in the past month. I tried a few models here and there but always went back to Claude, but with 5.2 Codex for the first time I felt it was very competitive, if not better.

Curious to see how things will be with 5.3 and 4.6

Re: Claude Opus 4.6

#348

Post by losvedir »

nickstinemates wrote: post #313
exe34 wrote: post #178

Earlier quoted context omitted.

Can Claude fix the flicker in Claude yet?

[flagged]

Oh, is that what the issue is? I've seen the "flicker" thing as a meme, but as someone who uses Claude Code I've never noticed. I use ghostty mostly, so maybe it's not an issue with ghostty? Or maybe I just haven't noticed it.

Re: Claude Opus 4.6

#349

Post by elAhmo »

jsheard wrote: post #38
blibble wrote: post #17

> We build Claude with Claude. Our engineers write code with Claude Code every day well that explains quite a bit

CC has >6000 open issues, despite their bot auto-culling them after 60 days of inactivity. It was ~5800 when I looked just a few days ago so they seem to be accelerating towards some kind of bug singularity.

Insane to think that a relatively simple CLI tool has so many open issues...

Re: Claude Opus 4.6

#350

Post by Aurornis »

runarberg wrote: post #236
Aurornis wrote: post #138

Earlier quoted context omitted.

> A year or more ago, I read that both Anthropic and OpenAI were losing money on every single request even for their paid subscribers This gets repeated everywhere but I don't think it's true. The company is unprofitable overall, but I don't see any reason to believe that their per-token inference costs are below the marginal cost of computing those tokens. It is true that the company is unprofitable overall when you…

I can see a case for omitting R&D when talking about profitability, but training makes no sense. Training is what makes the model, omitting it is like omitting the cost of running the production facility of a car manufacturer. If AI companies stop training they will stop producing models, and they will run out of a products to sell.

It depends on what you're talking about

If you're looking at overall profitability, you include everything

If you're talking about unit economics of producing tokens, you only include the marginal cost of each token against the marginal revenue of selling that token

Post reply on HN