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

Claude 4

anthropic.com

Post reply on HN
Print view View original

471–480 of 1001 posts

Re: Claude 4

#471

Post by sunaookami »

codyvoda wrote: post #324
datpuz wrote: post #311

Earlier quoted context omitted.

DeepSeek never mopped the floor with anyone... DeepSeek was remarkable because it is claimed that they spent a lot less training it, and without Nvidia GPUs, and because they had the best open weight model for a while. The only area they mopped the floor in was open source models, which had been stagnating for a while. But qwen3 mopped the floor with DeepSeek R1.

counterpoint: influencers said they wiped the floor with everyone so it must have happened

Who cares about what random influencers say?

Re: Claude 4

#473

Post by tasuki »

jonny_eh wrote: post #295
vFunct wrote: post #120

Earlier quoted context omitted.

Yah Claude tends to output 1200+ line architectural specification documents while Gemini tends to output ~600 line. (I just had to write 100+ architectural spec documents for 100+ different apps) Not sure why Claude is more thorough and complete than the other models, but it's my go-to model for large projects. The OpenAI model outputs are always the smallest - 500 lines or so. Not very good at larger projects, but p…

Who's reading these docs?

Another LLM, which distills it back into a couple of sentences for human consumption.

Re: Claude 4

#474

Post by Workaccount2 »

bbor wrote: post #398
vinhphm wrote: post #93

Earlier quoted context omitted.

The option just shown up in Copilot settings page for me

Turns out Opus 4 starts at their $40/mo ("Pro+") plan which is sad, and they serve o4-mini and Gemini as well so it's a bit less exclusive than this announcement implies. That said, I have a random question for any Anthropic-heads out there: GitHub says "Claude Opus 4 is hosted by Anthropic PBC. Claude Sonnet 4 is hosted by Anthropic 1P."[1]. What's Anthropic 1P ? Based on the only Kagi result being a deployment tuto…

Google launched Jules two days ago, which is the gemini coding agent[1]. I was pretty quickly accepted into the beta and you get 5 free tasks a day.

So far I have found it pretty powerful, its also the first time an LLM has ever stopped while working to ask me a question or for clarification.

[1]https://jules.google/

Re: Claude 4

#476

Post by jpau »

low_tech_punk wrote: post #213

Can anyone help me understand why they changed the model naming convention? BEFORE: claude-3-7-sonnet AFTER: claude-sonnet-4

Seems to be a nod to each size being treated as their own product.

Claude 3 arrived as a family (Haiku, Sonnet, Opus), but no release since has included all three sizes.

A release of "claude-3-7-sonnet" alone seems incomplete without Haiku/Opus, when perhaps Sonnet is has its own development roadmap (claude-sonnet-*).

Re: Claude 4

#477

Post by saaaaaam »

ldjkfkdsjnv wrote: post #362
calebm wrote: post #360

Earlier quoted context omitted.

It's interesting that we are in a world state in which "HATE the new personality it's got" is applied to AIs. We're living in the future ya'll :)

"His work was good, I just would never want to get a beer with him" this is literally how you know were approaching agi

Problem is the work isn’t very good either now. It’s like being sold to by someone who read a “sales techniques of the 1980s” or something.

If I’m asking it to help me analyse legal filings (for example) I don’t want breathless enthusiasm about my supposed genius on spotting inconsistencies. I want it to note that and find more. It’s exhausting having it be like this and it makes me feel disgusted.

Re: Claude 4

#478

Post by NitpickLawyer »

hsn915 wrote: post #416

I can't be the only one who thinks this version is no better than the previous one, and that LLMs have basically reached a plateau, and all the new releases "feature" are more or less just gimmicks.

> and that LLMs have basically reached a plateau

This is the new stochastic parrots meme. Just a few hours ago there was a story on the front page where an LLM based "agent" was given 3 tools to search e-mails and the simple task "find my brother's kid's name", and it was able to systematically work the problem, search, refine the search, and infer the correct name from an e-mail not mentioning anything other than "X's favourite foods" with a link to a youtube video. Come on!

That's not to mention things like alphaevolve, microsoft's agentic test demo w/ copilot running a browser, exploring functionality and writing playright tests, and all the advances in coding.

Re: Claude 4

#479

Post by Phelinofist »

minimaxir wrote: post #118

An important note not mentioned in this announcement is that Claude 4's training cutoff date is March 2025, which is the latest of any recent model. (Gemini 2.5 has a cutoff of January 2025) https://docs.anthropic.com/en/docs/about-claude/models/overv...

Why can't it be trained "continuously"?

Re: Claude 4

#480

Post by charlie0 »

runjake wrote: post #81
thimabi wrote: post #67

It’s been hard to keep up with the evolution in LLMs. SOTA models basically change every other week, and each of them has its own quirks. Differences in features, personality, output formatting, UI, safety filters… make it nearly impossible to migrate workflows between distinct LLMs. Even models of the same family exhibit strikingly different behaviors in response to the same prompt. Still, having to find each model’…

My advice: don't jump around between LLMs for a given project. The AI space is progressing too rapidly right now. Save yourself the sanity.

A man with one compass knows where he's going; a man with two compasses is never sure.
Post reply on HN