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

941–950 of 1001 posts

Re: Claude 3.7 Sonnet and Claude Code

#941

Post by airstrike »

eschluntz wrote: post #140
light_triad wrote: post #109

Earlier quoted context omitted.

Thanks for this - exciting launch. Do you have examples of cool applications or demos that the HN crowd should check out?

hi! I've been working on demos where I let Claude Code run for hours at a time on a sandboxed project: https://x.com/ErikSchluntz/status/1894104265817284770 TLDR: asking claude to speed up my code once 1.8x'd perf, but putting it in a loop telling it to make it faster for 2 hours led to a 500x speedup!

dunno who else to tell this but my pet request for the next version of Claude is to have it say "ensure" and "You're absolutely right!" less often

Re: Claude 3.7 Sonnet and Claude Code

#942

Post by CamperBob2 »

AlfeG wrote: post #729

Ahha, recently my daugher come to me with 3rd grade math problem. "Without rearranging the digits 1 2 3 4 5, insert mathematical operation signs and, if necessary, parentheses between them so that the resulting expression equals 40 and 80. The key is that you can combine digits (like 12+3/45) but you cannot change their order from the original sequence 1,2,3,4,5" Grok3, Claude, Deepseek, Qwen all failed to solve this…

Geez. Who teaches this 3rd-grade class, Prof. Xavier?

Interestingly, the R1 1.58-bit dynamic quant model was able to sort of solve it. I changed the problem statement a bit to request only the solution for 40 and to tell it what operations it can use, both needed to keep from blowing out the limited context available on my machine (128MB RAM + 24MB GPU).

Took almost 3 hours and it wigged out a bit at the end, rambling about Lisp in Chinese, but it got an almost-valid answer: 1 * (2 + 3) * (4 + 5) - 5 (https://pastebin.com/ggL85RWJ) I didn't think it would get that far.

Re: Claude 3.7 Sonnet and Claude Code

#943

Post by baumy »

mattwad wrote: post #687
neal_ wrote: post #638

Earlier quoted context omitted.

Cursor has no models, they dont even have an editor its just vscode

And Typescript simply doesn't work for me. I have tried uninstalling extensions. It is always "Initializing". I reload windows, etc. It eventually might get there, I can't tell what's going on. At the moment, AI is not worth the trade-off of no Typescript support.

My entire company of 100+ engineers is using cursor on multiple large typescript repos with zero issues. Must be some kind of local setup issue on your end, it definitely works just fine. In fact I've seen consistently more useful / less junky results from using LLMs for code with typescript than any other language, particularly when cursor's "shadow workspace" option is enabled.

Re: Claude 3.7 Sonnet and Claude Code

#944

Post by edmundsauto »

wokwokwok wrote: post #753
M4v3R wrote: post #710

Earlier quoted context omitted.

I’m currently doing something very similar to what GP is doing - I’m building a hobby project that’s a desktop app with web frontend. It’s a map editor with a 3D view. My estimate is that 80-90% of the code was written by AI. Sure, I did have to intervene or write some more complex parts myself but it’s still exciting to me that in many cases it took just a single prompt to add a new feature to it or change existing…

> My estimate is that 80-90% of the code was written by AI Nice! It is entirely reasonable both to do that and to be excited about it. …buuut, if that’s what you’re doing, you should say so. Not: “no lines of code directly written, just directing the AI” Because those (gluing together AI code by hand and having the agent do everything) are different things, and one of them is much much MUCH harder to get right than t…

I don’t think this is a fair take. For self driving cars, you care about that because safety is involved and the reliability of the AI is the product itself.

For OP, the product is the product, how they got there is mostly irrelevant. We don’t really care what IDE they used (outside of being a tooling nerd).

Re: Claude 3.7 Sonnet and Claude Code

#945

Post by vintermann »

audiodude wrote: post #892
vintermann wrote: post #886

Earlier quoted context omitted.

I have another easy one which thinking models get wrong: "Anhentafel numbers start with you as 1. To find the Ahhentafel number of someone's father, double it. To find the Ahnentafel number of someone's mother, double it and add one. Men pass on X chromosome DNA to their daughters, but none to their sons. Women pass on X chromosome DNA to both their sons and daughters. List the Ahnentafel numbers of the closest 20 an…

Yeah I wouldn't call this easy...

You can just do it generation for generation. The only thing hard about it is that it's two explained concepts you need to combine. A model which aces math Olympiad problems shouldn't have any trouble with this whatsoever - unless it's overfitting on them somehow.

Re: Claude 3.7 Sonnet and Claude Code

#946

Post by weatherlite »

KoolKat23 wrote: post #903
weatherlite wrote: post #888

Earlier quoted context omitted.

The LLMs are quite widely distributed already, they're just not that impactful. My wife is an accountant at a big 4 and they're all using them (everyone on Microsoft Office is probably using them, which is a lot of people). It's just not the earth shattering tech change CEOS make it to be , at least not yet. We need order of mangitude improvements in things like reliability, factuality and memory for the real economi…

Not necessarily, workflows just need to be adapted to work with it rather than it working in existing workflows. It's something that happens during each industrial revolution. Originally electric generators merely replaced steam generators but had no additional productivity gains, this only changed when they changed the rest of the processes around it.

I don't get this. What workflow can have occasional catastrophic lapses of reasoning, non factuality, no memory and hallucinations etc? Even in things like customer support this is a no go imo. As long as these very major problems aren't improved (by a lot) the tools will remain very limited.

Re: Claude 3.7 Sonnet and Claude Code

#947

Post by habinero »

harshreality wrote: post #939
weatherlite wrote: post #798

Earlier quoted context omitted.

Like what economic changes? You can make a case people are 10% more productive in very specific fields (programming, perhaps consultancy etc). That's not really an earthquake, the internet/web was probably way more significant.

It's a force multiplier. Think of having a secretary, or ten. These secretaries are not as good as an average human at most tasks, but they're good enough for tasks that are easy to double check. You can give them an immense amount of drudgery that would burn out a human.

What drudgery, though? Secretaries don't do a lot of drudgery. And a good one will see tasks that need doing that you didn't specify.

If you're generating immense amounts of really basic make work, that seems like you're managing your time poorly.

Re: Claude 3.7 Sonnet and Claude Code

#948

Post by berkes »

CamperBob2 wrote: post #914
berkes wrote: post #756

Earlier quoted context omitted.

I think many of the "AI can do coding" narratives don't see what coding means in real situations. It's finding out why "jbdoe1337" added this large if/else around the entire function body back in 2016 - it seems important business logic, but the commit just says "updated code". And how the h*ll this interaction between the conf.ini files, the conf/something.json and the ENV vars works. Why sometimes the ENV var overr…

Systems built from scratch with AI won't have these limitations, because only the model will ever see the code. It will implement a spec that's written in English or another human language. When the business requirements change, the spec will change. When that happens, the system will either modify its previously-written code or regenerate it from the ground up. Which strategy it chooses won't be especially interesti…

> It will implement a spec that's written in English or another human language.

No, it won't. Because "human languages" lack the precision to describe such a spec. This is exactly why programming languages exist in the first place: a language that humans understand but that allow for precise and unambiguous specifications and/or instructions. Do note that a computer cannot execute "Python" or "C". We needs to translate it first (compiling). Edit: A programmer doens't just type curly brackets and semi-colons in the right place, she takes vague and ambigous specs and makes them precise enough so that machines can repeat them.

As a kid we had this joke (works better in Dutch).

John gets in an accident, looses both his arms. A doctor gives him futuristic voice-controlled prostethics.

John: "Pick up coffee-mug". "Bring to mouth to drink". woa! impressed he goes home.

John, all excited "unzip pants", "grab d#ck", "jerk off"

(in Dutch, trek af means both "rip off" and "w#ank")

Jokes aside, we do have such a language that's not a programming language in the common sense: executable specs - end to end tests. Gherkin being a famous one but certainly not the only one. BDD, where the B is described by humans, in a DSL and the DD is performed by AI. I could imagine this working. Not currently and not anywhere soon (current LLMs are great at making new stuff, horrible at changing existing stuff), but it might work.

We'd then end up with just another programming language, but one thats more accessible to more people, I guess. And the AI is "just a compiler" in that sense.

Re: Claude 3.7 Sonnet and Claude Code

#949

Post by Flenser »

gwd wrote: post #774
jonplackett wrote: post #752

Earlier quoted context omitted.

I like to make up my own tests, that way you know it is actually thinking. Tests that require thinking about the physical world are the most revealing. My new favourite is: You have 2 minutes to cool down a cup of coffee to the lowest temp you can. You have two options: 1. Add cold milk immediately, then let it sit for 2 mins. 2. Let it sit for 2 mins, then add cold milk. Which one cools the coffee to the lowest temp…

> Phrased this way without any help, all but the thinking models get it wrong I C&P'd it into Claude 3.7 with thinking, and it gave the correct answer (which I'm pretty sure is #2). Including the CoT, where it actually does math (which I haven't checked), and final response. # THINKING Let's analyze the two options. Option 1: Add cold milk immediately, then let it sit for 2 mins. Option 2: Let it sit for 2 mins, then…

I thought you were wrong so I went googling and someone did the experiment and you are correct. Option 2, adding the milk later, cools fastest: https://www.thenakedscientists.com/get-naked/experiments/whe...

The graph on that page explains it much better than any of the text explanations can.

Re: Claude 3.7 Sonnet and Claude Code

#950

Post by tecleandor »

billmalarky wrote: post #915
anotherpaulg wrote: post #331

Claude 3.7 Sonnet scored 60.4% on the aider polyglot leaderboard [0], WITHOUT USING THINKING. Tied for 3rd place with o3-mini-high. Sonnet 3.7 has the highest non-thinking score, taking that title from Sonnet 3.5. Aider 0.75.0 is out with support for 3.7 Sonnet [1]. Thinking support and thinking benchmark results coming soon. [0] https://aider.chat/docs/leaderboards/ [1] https://aider.chat/HISTORY.html#aider-v0750

Hi Paul, been following the aider project for about a year now to develop an understanding of how to build SWE agents. I was at the AI Engineering Summit in NYC last week and met an (extremely senior) staff ai engineer doing somewhat unbelievable things with aider. Shocking things tbh. Is there a good way to share stories about real-world aider projects like this with you directly (if I can get approval from him)? No…

Hope it gets to be public, I love to learn "weird" (or unusual) ways of using tools
Post reply on HN