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

71–80 of 1001 posts

Re: Claude 3.7 Sonnet and Claude Code

#71

Post by lysace »

azinman2 wrote: post #50
lysace wrote: post #9

It's fascinating how close these companies are to each other. Some company comes up with something clever/ground-breaking and everyone else has implemented it a few weeks later. Hard not to think of Kurzweil's Law of Accelerating Returns.

It’s extremely unlikely that everyone is copying in a few weeks for models that themselves take many weeks if not longer to train. Great minds think alike, and everyone is influencing everyone. The history of innovation is filled with examples of similar discoveries around the same time but totally disconnected in the world. Now with the rate of publishing and the openness of the internet, you’re only bound to get ev…

Isn't the reasoning thing essentially a bolt-on to existing trained models? Like basically a meta-prompt?

Re: Claude 3.7 Sonnet and Claude Code

#72

Post by pertymcpert »

azinman2 wrote: post #67

To me the biggest surprise was seeking grok dominate in all of their published benchmarks. I haven’t seen any benchmarks of it yet (which I take with a giant heap of salt), but it’s still interesting nevertheless. I’m rooting for Anthropic.

Indeed. I wonder what the architecture for Claude and Grok3 is. If they're still dense models was the MoE excitement with R1 was a tad premature...

Re: Claude 3.7 Sonnet and Claude Code

#73

Post by KaoruAoiShiho »

t55 wrote: post #3

Anthropic doubling down on code makes sense, that has been their strong suit compared to all other models Curious how their Devin competitor will pan out given Devin's challenges

They cited Cognition (Devin's maker) in this blog post which is kinda funny.

Re: Claude 3.7 Sonnet and Claude Code

#74

Post by minihat »

jedberg wrote: post #25

Last week when Grok launched the consensus was that its coding ability was better than Claude. Anyone have a benchmark with this new model? Or just warm feelings?

Grok 3 with thinking is comparable to o1 for writing complex algorithms.

However, Grok sometimes loses the context where o1 seems not to. For this reason I still mostly use o1.

I have found both o1 and Grok 3 to be substantially better than any Claude offering.

Re: Claude 3.7 Sonnet and Claude Code

#75

Post by behnamoh »

whywhywhywhy wrote: post #40
rcstank wrote: post #35

Earlier quoted context omitted.

Credit card is easily changed. Phone number is much more difficult.

Not difficult at all for anyone actually wanting to abuse it

You shouldn't have to want to abuse something for it to be inappropriate to be asked.

Re: Claude 3.7 Sonnet and Claude Code

#76

Post by KaoruAoiShiho »

azinman2 wrote: post #50
lysace wrote: post #9

It's fascinating how close these companies are to each other. Some company comes up with something clever/ground-breaking and everyone else has implemented it a few weeks later. Hard not to think of Kurzweil's Law of Accelerating Returns.

It’s extremely unlikely that everyone is copying in a few weeks for models that themselves take many weeks if not longer to train. Great minds think alike, and everyone is influencing everyone. The history of innovation is filled with examples of similar discoveries around the same time but totally disconnected in the world. Now with the rate of publishing and the openness of the internet, you’re only bound to get ev…

The copying here probably goes to strawberry from o1 which is like at least 6 months but maybe copying efforts started even earlier.

Re: Claude 3.7 Sonnet and Claude Code

#77

Post by pertymcpert »

lysace wrote: post #71
azinman2 wrote: post #50

Earlier quoted context omitted.

It’s extremely unlikely that everyone is copying in a few weeks for models that themselves take many weeks if not longer to train. Great minds think alike, and everyone is influencing everyone. The history of innovation is filled with examples of similar discoveries around the same time but totally disconnected in the world. Now with the rate of publishing and the openness of the internet, you’re only bound to get ev…

Isn't the reasoning thing essentially a bolt-on to existing trained models? Like basically a meta-prompt?

Somewhat but not exactly? I think the models need to be trained to think.

Re: Claude 3.7 Sonnet and Claude Code

#78

Post by phillipcarter »

azinman2 wrote: post #67

To me the biggest surprise was seeking grok dominate in all of their published benchmarks. I haven’t seen any benchmarks of it yet (which I take with a giant heap of salt), but it’s still interesting nevertheless. I’m rooting for Anthropic.

Neither a statement for or against Grok or Anthropic:

I've now just taken to seeing benchmarks as pretty lines or bars on a chart that are in no way reflective of actual ability for my use cases. Claude has consistently scored lower on some benchmarks for me, but when I use it in a real-world codebase, it's consistently been the only one that doesn't veer off course or "feel wrong". The others do. I can't quantify it, but that's how it goes.

Re: Claude 3.7 Sonnet and Claude Code

#79

Post by azinman2 »

lysace wrote: post #71
azinman2 wrote: post #50

Earlier quoted context omitted.

It’s extremely unlikely that everyone is copying in a few weeks for models that themselves take many weeks if not longer to train. Great minds think alike, and everyone is influencing everyone. The history of innovation is filled with examples of similar discoveries around the same time but totally disconnected in the world. Now with the rate of publishing and the openness of the internet, you’re only bound to get ev…

Isn't the reasoning thing essentially a bolt-on to existing trained models? Like basically a meta-prompt?

No.

DeepSeek and now related projects have shown it’s possible to add reasoning via SFT to existing models, but that’s not the same as a prompt. But if you look at R1 they do a blend of techniques to get reasoning.

For Anthropic to have a hybrid model where you can control this, it will have to be built into the model directly in its training and probably architecture as well.

If you’re a competent company filled with the best AI minds and a frontier model, you’re not just purely copying… you’re taking ideas while innovating and adapting.

Post reply on HN