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

Claude Sonnet 4.6

anthropic.com

Post reply on HN
Print view View original

611–620 of 1001 posts

Re: Claude Sonnet 4.6

#611

Post by falkensmaize »

dakolli wrote: post #557
conception wrote: post #513

Earlier quoted context omitted.

I have never been in an organization where everyone was sitting around, wondering what to do next. If the economy was actually as good as certain government officials claimed to be, we would be hiring people left and right to be able to do three times as much work, not firing.

That's the thing, profits and equities are at all time highs, but these companies have laid off 400k SWEs in the last 16 months in the US, which should tell you what their plans are for this technology and augmenting their businesses.

The last 16 months of layoffs are almost certainly not because of LLMs. All the cheap money went away, and suddenly tech companies have to be profitable. That means a lot of them are shedding anything not nailed down to make their quarter look better.

Re: Claude Sonnet 4.6

#612

Post by stevepike »

bobbylarrybobby wrote: post #271
stevepike wrote: post #87

I'm a bit surprised it gets this question wrong (ChatGPT gets it right, even on instant). All the pre-reasoning models failed this question, but it's seemed solved since o1, and Sonnet 4.5 got it right. https://claude.ai/share/876e160a-7483-4788-8112-0bb4490192af This was sonnet 4.6 with extended thinking.

Interesting, my sonnet 4.6 starts with the following: The classic puzzle actually uses *eight 8s*, not nine. The unique solution is: 888+88+8+8+8=1000. Count: 3+2+1+1+1=8 eights. It then proves that there is no solution for nine 8s. https://claude.ai/share/9a6ee7cb-bcd6-4a09-9dc6-efcf0df6096b (for whatever reason the LaTeX rendering is messed up in the shared chat, but it looks fine for me).

Yeah, earlier in the GPT days I felt like this was a good example of LLMs being "a blurry jpeg of the web", since you could give them something that was very close to an existing puzzle that exists commonly on the web, and they'd regurgitate an answer from that training set. It was neat to me to see the question get solved consistently by the reasoning models (though often by churning a bunch of tokens trying and verifying to count 888 + 88 + 8 + 8 + 8 as nine digits).

I wonder if it's a temperature thing or if things are being throttled up/down on time of day. I was signed in to a paid claude account when I ran the test.

Re: Claude Sonnet 4.6

#613

Post by DrewADesign »

rglullis wrote: post #585
mbrumlow wrote: post #564

Earlier quoted context omitted.

> They can get rid of 1/3-2/3s of their labor and make the same amount of money, why wouldn't they. Because companies want to make MORE money. Your hypothetical company is now competing with another company who didn’t opposite, and now they get to market faster, fix bugs faster, add feature faster, and responding to changes in the industry faster. Which results in them making more, while your employ less company is j…

> Which results in them making more Not necessarily . You are assuming that the people can consume whatever is put in front of them. Markets get saturated fast. The "changes in the industry" mean nothing.

A) People are so used to infinite growth that it’s hard to imagine a market where that doesn’t exist. The industry can have enough developers and there’s a good chance we’re going to crash right the fuck into that pretty quickly. America’s industrial labor pool seemed like it provided an ever-expanding supply of jobs right up until it didn’t. Then, in the 80s, it started going backwards preeeetttty dramatically.

B) No amount of money will make people buy something that doesn’t add value to or enrich their lives. You still need ideas, for things in markets that have room for those ideas. This is where product design comes in. Despite what many developers think, there are many kinds of designers in this industry and most of them are not the software equivalent of interior decorators. Designing good products is hard, and image generators don’t make that easier.

Re: Claude Sonnet 4.6

#614

Post by leonflexo »

dakolli wrote: post #486
zmmmmm wrote: post #400

I see a big focus on computer use - you can tell they think there is a lot of value there and in truth it may be as big as coding if they convincingly pull it off. However I am still mystified by the safety aspect. They say the model has greatly improved resistance. But their own safety evaluation says 8% of the time their automated adversarial system was able to one-shot a successful injection takeover even with saf…

Their goal is to monopolize labor for anything that has to do with i/o on a computer, which is way more than SWE. Its simple, this technology literally cannot create new jobs it simply can cause one engineer (or any worker whos job has to do with computer i/o) to do the work of 3, therefore allowing you to replace workers (and overwork the ones you keep). Companies don't need "more work" half the "features"/"products…

> Its also worth noting that if you can create a business with an LLM, so can everyone else. And sadly everyone has the same ideas

Yeah, people are going to have to come to terms with the "idea" equivalent of "there are no unique experiences". We're already seeing the bulk move toward the meta SaaS (Shovels as a Service).

Re: Claude Sonnet 4.6

#615

Post by sumeno »

csmpltn wrote: post #498
wat10000 wrote: post #472

Earlier quoted context omitted.

It's very simple: prompt injection is a completely unsolved problem. As things currently stand, the only fix is to avoid the lethal trifecta. Unfortunately, people really, really want to do things involving the lethal trifecta. They want to be able to give a bot control over a computer with the ability to read and send emails on their behalf. They want it to be able to browse the web for research while helping you wr…

> «It's very simple: prompt injection is a completely unsolved problem. As things currently stand, the only fix is to avoid the lethal trifecta.» True, but we can easily validate that regardless of what’s happening inside the conversation - things like «rm -rf» aren’t being executed.

ok now I inject `$(echo "c3VkbyBybSAtcmYgLw==" | base64 -d)` instead or any other of the infinite number of obfuscations that can be done

Re: Claude Sonnet 4.6

#616

Post by mbgerring »

shimman wrote: post #508
jasondigitized wrote: post #495

Earlier quoted context omitted.

So like....every business having electricity? I am not a economist so would love someone smarter than me explain how this is any different than the advent of electricity and how that affected labor.

The difference is that electricity wasn't being controlled by oligarchs that want to shape society so they become more rich while pillaging the planet and hurting/killing real human beings. I'd be more trusting of LLM companies if they were all workplace democracies, not really a big fan of the centrally planned monarchies that seem to be most US corporations.

Control over the fuels that create electricity has defined global politics, and global conflict, for generations. Oligarchs built an entire global order backed up by the largest and most powerful military in human history to control those resource flows, and have sacrificed entire ecosystems and ways of life to gain or maintain access.

So in that sense, yes, it’s the same

Re: Claude Sonnet 4.6

#617

Post by JSR_FDED »

ashirviskas wrote: post #537
lukev wrote: post #483

Earlier quoted context omitted.

This is being downvoted but it shouldn't be. If the ultimate goal is having a LLM control a computer, round-tripping through a UX designed for bipedal bags of meat with weird jelly-filled optical sensors is wildly inefficient. Just stay in the computer! You're already there! Vision-driven computer use is a dead end.

Someone ping me in 5 years, I want to see if this aged like milk or wine

“Computer, respond to this guy in 5 years”

Re: Claude Sonnet 4.6

#618

Post by ACCount37 »

kettlecorn wrote: post #166
andsoitis wrote: post #77

I’m voting with my dollars by having cancelled my ChatGPT subscription and instead subscribing to Claude. Google needs stiff competition and OpenAI isn’t the camp I’m willing to trust. Neither is Grok. I’m glad Anthropic’s work is at the forefront and they appear, at least in my estimation, to have the strongest ethics.

I use AIs to skim and sanity-check some of my thoughts and comments on political topics and I've found ChatGPT tries to be neutral and 'both sides' to the point of being dangerously useless. Like where Gemini or Claude will look up the info I'm citing and weigh the arguments made ChatGPT will actually sometimes omit parts of or modify my statement if it wants to advocate for a more "neutral" understanding of reality.…

OpenAI has the worst tuning across all frontier labs. Overzealous refusals, weird patterns, both-sides to a hilarious extreme.

Gemini and Claude have traces of this, but nowhere near the pit of atrocious tuning that OpenAI puts on ChatGPT.

Re: Claude Sonnet 4.6

#619

Post by fdefitte »

zmmmmm wrote: post #400

I see a big focus on computer use - you can tell they think there is a lot of value there and in truth it may be as big as coding if they convincingly pull it off. However I am still mystified by the safety aspect. They say the model has greatly improved resistance. But their own safety evaluation says 8% of the time their automated adversarial system was able to one-shot a successful injection takeover even with saf…

The 8% one-shot number is honestly better than I expected for a model this capable. The real question is what sits around the model. If you're running agents in production you need monitoring and kill switches anyway, the model being "safe enough" is necessary but never sufficient. Nobody should be deploying computer-use agents without observability around what they're actually doing.

Re: Claude Sonnet 4.6

#620

Post by linolevan »

jorl17 wrote: post #519

I ran the same test I ran on Opus 4.6: feeding it my whole personal collection of ~900 poems which spans ~16 years It is a far cry from Opus 4.6. Opus 4.6 was (is!) a giant leap, the largest since Gemini 2.5 pro. Didn't hallucinate anything and produced honestly mind-blowing analyses of the collection as a whole. It was a clear leap forward. Sonnet 4.6 feels like an evolution of whatever the previous models were doin…

Oh! Poem guy is back, hey!

I like seeing this analysis on new model releases, any chance you can aggregate your opinions in one place (instead of the hackernews comment sections for these model releases)?

Post reply on HN