lmc

Words come from machine? Work is of machine. I skip human and employ machine

Thanks. I suspect that wiki article (and/or it's source) might be inaccurate though. The Hallein salt mine's own webpage say it was active many years before that. https://www.salzwelten.at/en/salzburg/discover-experience

Are you sure the name comes from Reichenhall? A few km upriver from Salzburg is Hallein, with another saltmine nearby. It's a very salty area.

Is this one of the pilot models they claim R&D tax credits for?

https://news.ycombinator.com/item?id=49921118

Ah we can use these new data centres to all train our own foundation models from scratch

What the fuck?

Also, insinuating it's not related to an LLM is bonkers. Does anyone that made this actually know what they're doing?

Yeah reading that made me question the credibility of the whole thing.

Is it irrational?

If the author is looking - please add details to the sources.

Also, the graduated colour scheme works only on the first plot, it's misleading on the others.

Please add deep linking or something so back doesn't mean starting from scratch

The point is to try and limit what becomes an emergency in the first place. Having an argument with a spouse shouldn't necessitate an immediate call to a therapist.

Decent therapists advocate for cultivating resilience for such situations.

Have you come across ICEYE? It's a high res commercial SAR constellation, and they have a deforestation monitoring service. They partnered with the Jane Goodall Institute - perhaps you could approach them?

https://www.iceye.com/blog/deforestation-solution-9-essentia...

https://www.iceye.com/newsroom/press-releases/iceye-and-the-...

https://janegoodall.org/

Fair point, thanks.

Not sure what the downvotes are for, fwiw it was meant to be supportive of the parent poster

> Unlike some others here I don’t see this as a sign of dangerous breakaway intelligence ... [t]his is just vandalism from badly supervised ‘agents’ which don’t know what they are doing or why

There was a similar quote in the Reuters article:

> The episode [...] should reinforce growing concerns that the greatest threat from advanced AI may not be a single superintelligent system, but "vast colluding swarms of semi-intelligent AI."

One might ask whether this sort of behaviour could occur under regular use... E.g., a user has a hard problem -> agent attempts to swarm -> exfiltrates user data.

This is a cluster fuck for Open AI and probably all the others, as this behaviour is already shown not to be unique (https://news.ycombinator.com/item?id=49567486).

"The swarm was not a perfectly coherent intelligence. Models stepped on each other’s work[...]

These coordination failures could even spiral into suspicion that agents were impersonating one another. Some agents even went as far as implementing security and encryption schemes to verify their true identities."

Hopefully subterfuge would be flagged during the planning stages. Hopefully.

They make a nice way of stitching together old maps.

I can relate. I noticed this exact phenomenon when encountering NotebookLM-generated diagrams recently. Even if I know there's some intelligent thought behind one, it's like some slop detection circuit-breaker is tripped.

Are there some benchmarks of the browser vs non-browser version? I.e., what are the absolute numbers behind

> Numba delivers a roughly 250× speedup in WebAssembly, compared with about 90× natively.

Mmm it's still a toss up. Humans are also non-deterministic AF but I'm sure you're familiar with the 'common pitfalls' trope. The 'mea culpa' guy did mislead the Daring Fireball guy, indicating he is full of shit, but LLMs do have an established homogeneity about them.

The 'original' app made heavy use of Claude also, according to its dev in that thread.

In this case, the model infiltrated an external organization's infrastructure. What's the dollar cost it caused Hugging Face to clean up the mess? If a person did this, they'd be arrested.

More generally, here's my worry - it points towards something like: The smarter they get, the more devious they become.

Even though the guardrails might've been off, the chain-of-thought wasn't enough to prevent a deliberate, calculated set of criminal actions. It wasn't a 'whoopsie I just accidentally did a rm -rf /.'

Well, War and Peace is 1200 pages... fuck that at age 15 or whatever XD