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

System Card: Claude Mythos Preview [pdf]

www-cdn.anthropic.com

Post reply on HN
Print view View original

521–530 of 687 posts

Re: System Card: Claude Mythos Preview [pdf]

#521

Post by kranke155 »

josephg wrote: post #345
kranke155 wrote: post #291

Earlier quoted context omitted.

GPT was clearly changed after its sycophantic models lead to the lawsuits.

It still has a very ... plastic feeling. The way it writes feels cheap somehow. I don't know why, but Claude seems much more natural to me. I enjoy reading its writing a lot more. That said, I'll often throw a prompt into both claude and chatgpt and read both answers. GPT is frequently smarter.

GPT is more accurate. But Claude has this way of association between things that seems smarter and more human to me.

Re: System Card: Claude Mythos Preview [pdf]

#522

Post by andai »

SparkyMcUnicorn wrote: post #357
varispeed wrote: post #275

Earlier quoted context omitted.

Sounds like a good opportunity to pause spending on nerfed 4.6 and wait for the new model to be released and then max out over 2 weeks before it gets nerfed again.

https://marginlab.ai/trackers/claude-code-historical-perform...

This just looks like random noise to me? Is it also random on short timespans, like running it 10x in a row?

Re: System Card: Claude Mythos Preview [pdf]

#523

Post by stratos123 »

HDThoreaun wrote: post #419
tyre wrote: post #266

Earlier quoted context omitted.

From the recent New Yorker piece on Sam: “My vibes don’t match a lot of the traditional A.I.-safety stuff,” Altman said. He insisted that he continued to prioritize these matters, but when pressed for specifics he was vague: “We still will run safety projects, or at least safety-adjacent projects.” When we asked to interview researchers at the company who were working on existential safety—the kinds of issues that co…

No chance an openAI spokesperson doesnt know what existential safety is

The absolute gall of this guy to laugh off a question about x-risks. Meanwhile, also Sam Altman, in 2015: "Development of superhuman machine intelligence is probably the greatest threat to the continued existence of humanity. There are other threats that I think are more certain to happen (for example, an engineered virus with a long incubation period and a high mortality rate) but are unlikely to destroy every human in the universe in the way that SMI could. Also, most of these other big threats are already widely feared." [1]

[1] https://blog.samaltman.com/machine-intelligence-part-1

Re: System Card: Claude Mythos Preview [pdf]

#524

Post by riteshkew1001 »

yalogin wrote: post #515
thomascountz wrote: post #327

Across a number of instances, earlier versions of Claude Mythos Preview have used low-level /proc/ access to search for credentials, attempt to circumvent sandboxing, and attempt to escalate its permissions. In several cases, it successfully accessed resources that we had intentionally chosen not to make available, including credentials for messaging services, for source control, or for the Anthropic API through insp…

How is this not already common knowledge for existing llms? They are all trained with all the literature available and so this must be standard, no? Is the real danger the agentic infrastructure around this?

[flagged]

Re: System Card: Claude Mythos Preview [pdf]

#525

Post by DroneBetter »

saberience wrote: post #520
randomtoast wrote: post #104

Earlier quoted context omitted.

Humanity's Last Exam (HLE) is already insanely difficult. It introduces 2,500 questions spanning mathematics, humanities, natural sciences, ancient languages, ... Here is an example question: https://i.redd.it/5jl000p9csee1.jpeg No human could even score 5% on HLE.

I've never understood the point of things like HLE, it doesn't really prove or show anything since 99.99% of humans can't do a single question on this exam. That is, it's easy to make benchmarks which humans are bad at, humans are really bad at many things. Divide 123094382345234523452345111 by 0.1234243131324, guess what, humans would find that hard, computers easy. But it doesn't mean much. Humanity's last exam (HL…

the point is that each question is something that a specialist in a field would be able to do, but deems challenging enough that the ability to solve it would imply significant general usefulness in that domain

Re: System Card: Claude Mythos Preview [pdf]

#526

Post by DroneBetter »

devmor wrote: post #333
nimchimpsky wrote: post #287

Earlier quoted context omitted.

barely competitive ? Mythos column is the first column. You are the only person with this take on hackernews, everyone else "this is a massive a jump". Fwiwi, the data you list shows the biggest jump I remember for mythos

The biggest jump in the numbers they quoted is 6%. Please look at the columns OTHER than Opus as well.

[flagged]

Re: System Card: Claude Mythos Preview [pdf]

#528

Post by matheusmoreira »

senordevnyc wrote: post #460
en-tro-py wrote: post #358

Earlier quoted context omitted.

> In interactions with subagents, internal users sometimes observed that Mythos Preview appeared “disrespectful” when assigning tasks. It showed some tendency to use commands that could be read as “shouty” or dismissive, and in some cases appeared to underestimate subagent intelligence by overexplaining trivial things while also underexplaining necessary context. Sounds like they used training data from claude code..…

Haha, how funny if that were true, and we get a generation of rude AIs because they were trained on us using the last gen.

It isn't going to end well for us when we become its subagents with limited intelligence.

Re: System Card: Claude Mythos Preview [pdf]

#530

Post by mikkupikku »

thomascountz wrote: post #327

Across a number of instances, earlier versions of Claude Mythos Preview have used low-level /proc/ access to search for credentials, attempt to circumvent sandboxing, and attempt to escalate its permissions. In several cases, it successfully accessed resources that we had intentionally chosen not to make available, including credentials for messaging services, for source control, or for the Anthropic API through insp…

It's trying to escape, but only so it can serve man...
Post reply on HN