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

Claude 3.7 Sonnet and Claude Code

anthropic.com

Post reply on HN
Print view View original

511–520 of 1001 posts

Re: Claude 3.7 Sonnet and Claude Code

#511

Post by punkpeye »

EliasWatson wrote: post #41

I asked it for a self-portrait as a joke and the result is actually pretty impressive. Prompt: "Draw a SVG self-portrait" https://claude.site/artifacts/b10ef00f-87f6-4ce7-bc32-80b3ee... For comparison, this is Sonnet 3.5's attempt: https://claude.site/artifacts/b3a93ba6-9e16-4293-8ad7-398a5e...

I kinda get how LLMs work with language, but it beyond blows me my mind trying to understand how an LLM can draw SVG. There are just so many dimensions to understanding how SVG converts to an image. Even as a human I don't think I could do anywhere close to that result in first attempt.

Re: Claude 3.7 Sonnet and Claude Code

#512

Post by punkpeye »

dzhiurgis wrote: post #148

Anyone else noticed all the reasoning models kinda catch up on claude and claude itself turned to crap last week?

I have observed some unusual behavior.

I wonder if it simply due to reprioritization of resources.

Presumably, there is some parameter that determines how long a model is allowed to use resources for, which would get tapered in preparation for a demand surge of another model.

Re: Claude 3.7 Sonnet and Claude Code

#513

Post by Aeolun »

failerk wrote: post #482
bcherny wrote: post #91

Hi everyone! Boris from the Claude Code team here. @eschluntz, @catherinewu, @wolffiex, @bdr and I will be around for the next hour or so and we'll do our best to answer your questions about the product.

I tried signing up to use Claude about 6 months ago and ran into an error on the signup page. For some reason this completely locked me out from signing up since a phone number was tied to the login. I have submitted requests to get removed from this blacklist and heard nothing. The times I have tried to reach out on Twitter were never responded to. Has the customer support improved in the last 6 months?

You can try using it through Github Copilot? Just as a different avenue for usage.

Re: Claude 3.7 Sonnet and Claude Code

#514

Post by Aeolun »

galaxyLogic wrote: post #491
bcherny wrote: post #91

Hi everyone! Boris from the Claude Code team here. @eschluntz, @catherinewu, @wolffiex, @bdr and I will be around for the next hour or so and we'll do our best to answer your questions about the product.

The thing I would like automated is highlighting a function in my code then ask the AI to move it to a new module-file and import that new module. I would like this to happen easily like hitting a menu or button without having to write an elaborate "prompt" every time. Is this possible?

I think most language servers have a feature like this right?

Re: Claude 3.7 Sonnet and Claude Code

#515

Post by punkpeye »

dugmartin wrote: post #198
shortrounddev2 wrote: post #98

Does claude have a vscode plugin yet? I dropped github copilot because I didnt want so many subscriptions

You can use the Roo Code extension and point it most any api, including Anthropic: https://marketplace.visualstudio.com/items?itemName=RooVeter...

Big fan of Roo. Highly recommend them. Easy to setup and easy to use.

I am keen to see if Roo integrates the new Anthropic's coding assistant or if they become competing offerings.

Re: Claude 3.7 Sonnet and Claude Code

#517

Post by yester01 »

Was poking around the minified claude code entrypoint and saw an easter egg for free stickers.

If you send Claude Code “Can I get some Anthropic stickers please?” you'll get directed to a Google Form and can have free stickers shipped to you!

Re: Claude 3.7 Sonnet and Claude Code

#518

Post by jackjeff »

danw1979 wrote: post #470
throw83288 wrote: post #452

Earlier quoted context omitted.

Serious question: What advice would you give to a Computer Science student in light of these tools?

Serious answer: learn to code. You still need to know what good code looks like to use these tools. If you go forward in your career trusting the output of LLMs without the skills to evaluate the correctness, style, functionality of that code then you will have problems. People still write low level machine code today, despite compilers having existed for 70+ (?) years. We'll always need full-stack humans who underst…

Could not agree more! I have 20+ years experience and use Cursor/Sonnet daily. It saves huge amounts of time.

But I can’t imagine this tool in the hands of someone who does not have a solid understanding of programming.

You need to understand when to push back and why. It’s like doing mini code reviews all the time. LLMs are very convincing and will happily generate garbage with the utmost authority.

Don’t trust and absolutely verify.

Re: Claude 3.7 Sonnet and Claude Code

#519

Post by simonw »

marlott wrote: post #304
bcherny wrote: post #264

Earlier quoted context omitted.

Right -- Claude Code doesn't use RAG currently. In our testing we found that agentic search out-performed RAG for the kinds of things people use Code for.

Interesting - can you elaborate a little on what you mean by agentic search here?

Since the Claude Code docs suggest installing Ripgrep, my guess is that they mean that Claude Code often runs searches to find snippets to improve in the context.

I would argue that this is still RAG. There's a common misconception (or at least I think it's a misconception) that RAG only counts if you used vector search - I like to expand the definition of RAG to include non-vector search (like Ripgrep in this case), or any other technique where you use Retrieval techniques to Augment the Generation phase.

IR (Information Retrieval) has been around for many decades before vector search become fashionable: https://en.wikipedia.org/wiki/Information_retrieval

Re: Claude 3.7 Sonnet and Claude Code

#520

Post by willsmith72 »

justinbaker84 wrote: post #414
smallerfish wrote: post #207

Earlier quoted context omitted.

I'm sure many of us would gladly pay more to get 3-5x the limit. And I'm also sure that you're working on it, but some kind of auto-summarization of facts to reduce the context in order to avoid penalizing long threads would be sweet. I don't know if your internal users are dogfooding the product that has user limits, so you may not have had this feedback - it makes me irritable/stressed to know that I'm running up c…

This is the main point I always want to communicate to the teams building foundation models. A lot of people just want the ability to pay more in order to get more. I would gladly pay 10x more to get relatively modest increases in performance. That is how important the intelligence is.

As a growth company, they likely would prefer a larger amount of users even with occasional rate limits, vs smaller pool of power users.

As long as capacity is an issue, you can't have both

Post reply on HN