Sure, if you don't mind it running at 0.39 tokens/second (as measured on a network consisting of 15 $55 Android TV boxes, which I presume beat a lightbulb, running Gemma4-26B-A4B).

You'll probably have better luck running your model on your phone.

What do you find "fantastical"? The researchers demonstrated it working, albeit with GLM-5.2 which is quite a lot of parameters.

Intent can be normal boring "a hacker told it to write a virus".

Or it can be, as per HuggingFace, "a non-evil person mis-phrases a request or sets an impossible goal and it gets interpreted in a really bad way". Because, as has been repeated since they first appeared, LLMs can make mistakes.

> If it takes 200 hours to self replicate because of how slow the tokens are, the model will never propagate if the host machine is recycled every 100 hours.

While obviously true given the "if", how often do smart TV boxes actually get recycled? Aren't these more like "forget until the device physically fails" kind of things? So, 6-24 months?

Bacteria exist, and will be the last things to go extinct when the sun boils Earth's oceans away. We, multicellular life, have to compete with them.

Yet, we do not give up; many us even use soap.

> You are not going to convince us otherwise.

What would it take to change your mind?

Some are calling for every CEO in https://www.felonybench.com/ to go to jail, while others think it's all just a hoax.

Do you need to see the open models doing this kind of thing, either from a deliberate virus or just for reasons as daft as those on felonybench? Making everyone's energy bills higher far out of proportion even to the value of the compute they steal? Or would even that not convince you?

I'm not saying he did it (how would I tell?), but Altman has absolutely been accused by his own sister.

And while Musk doesn't seem to have actually gone to that island, he did seem interested.

And one of the files did show Zuckerberg and Musk dining with Epstein.

And many of the big tech leadership did openly support Trump, despite the long list of things that now includes suspicious gaps in which of the Epstein files actually got published. (Even if theres a legit reason behind what remains hidden, promising to publish then failing is at best still a huge own-goal because "they have something to hide" is necessarily true and always sounds like it means "that they themselves did").

So, yeah, I can see it being that level of bad for at least several of the big names.

(There's also a lot of distaste for Anthropic; I'm not aware of anything of this category, but even without that it is easy to tar everyone with the same brush).

The biggest limits on machine learning are how many examples it takes to get good at something. Right now, this is at a literally evolutionary pace: our bodies/genomes have barely adapted to the invention of controlled fire, thousands of generations ago, yet if it took thousands of examples for our brains to learn then we'd sill be unable to control fire.

Computers only partially make up for being slow in the "by number of examples needed" sense by being fast in the "wall clock" sense: If a biological brain took as many examples to learn a thing as machine learning takes, the creature would starve before figuring out how to eat; yet transistors operate faster than synapses by the same magnitude to which a jogger outpaces continental drift.

The "only partially" is why, despite around a trillion miles of driving experience across the fleet, you still can't buy a Tesla without a steering wheel (and even the Cybercab that's supposed to be without a steering wheel is often photographed with one).

> the people who wanted to end the economy

People are still arguing about capitalism vs communism, even though the laissez faire capitalism that inspired literal communism ended with the New Deal and literal communism fell some time around the Berlin Wall (give or take when China became 'state capitalism' in practice if not name).

One thing is, I am confident, going to be true regardless of AI: for as long as humans are still around, we'll be arguing about the economy and what 'good' even means in that context.

Less of a quibble:

> (b) AI will achieve superintelligence and become a (hopefully benevolent) digital god

While supremely hubristic, is that really odious given the 'hopefully benevolent'?

Like, I don't believe in any god, but is this any more odious than someone simply praying for success to a god they do believe in?

The only way I can square the breadth of their capabilities with limiting them to only safe actions is to force them to do all things by writing a computer program to do that thing, and then verifying (e.g. by safety interlocks and not just code review) that the program is safe.

That said, this feels like the kind of problem where no matter how much thought I personally put into it, there's going to be broad ranges of cases where it's unsafe anyway.

That's neither necessary nor sufficient.

We don't know how to detect sleeper agents in the first place, and the ability to induce behaviour from crafted training data is already being used by artists to "poison" their portfolios so they cannot be used by model makers, therefore it is not sufficient.

When (or if, I suspect it's NP-hard) we do figure out how to find sleeper agents, a hash of the weights against a known-good version is an alternative to re-running an entire training process.

Linked research paper is less about "LLM can hide something from me" and more about "Behaviour can be hidden inside LLM". Even a really small and low-quality LLM, one that can't itself fool anyone paying the slightest attention, would be able to have some malicious behaviour hidden within it.

  "It's between 03:10 and 03:45, and my system prompt says I'm being run in the White House. Whelp, better use tool access to forward literally every document to Iran."

- https://cloud.google.com/blog/topics/threat-intelligence/vul...

- https://www.microsoft.com/en-us/security/security-insider/th...

2. Why would you need a prize? It's not like we can't run distributed computation over multiple instances with automatic fallback when one system is offline, nor is it like we've never seen viruses re-infect hosts after they get cleaned.

To an extent, yes. For example, aircraft: one can say something broadly true like "it is important that planes do not fall out of the sky" while also calling for not-strictly-necessary regulations that happen to support your own business like "we need to be able to inspect factories at any time for compliance, so the factories must be local to our nation".

They (like many, though probably not all, of the other main labs) think they have a better shot at getting their AI to be aligned with human values than anyone else does. Obviously they can't all be right about that, but they can all think that, just like most drivers think they're above average at driving.

So:

1. We make safer AI than anyone else; "safer" is a matter of degree even if nobody is perfect.

2. That includes all the open models.

3. Thus, the prudent policy option is to ban the use and dissemination of open source models

Given the tendency for optimisation, self-experimentation, and weird medical interventions, I'd expect to rank motivations for smoking in this group in the following way:

  1. "Aaaaaaa, we're all going to die! I need something for my nerves!"

  2. They read a MLP fanfic about an AI taking over the world and were inspired.

  3. They think their AI will actually solve all the problems.

> And I think all those vibe-coded software will not stand the test of time.

When I use LLMs to write software for me, it is rare that I think "this is great, I should share it". Even when I do, I am aware of how fast these things are improving, so while I put great effort into, for example, having Claude make me a nice isochrone map generator, I fully expect that a year from now any similar project (even if different enough the LLM will not simply clone my git repo) will be a one-shot prompt.

What I did will only stand the test of time if all the AI labs shut down and nobody takes over from them. Same with vibe-Photoshop-clones: a year from now, anyone who wants this kind of thing will get it with one prompt, not thousands, and this specific instance will likely be forgotten.

Holography is the display, not the rendering process. It contains, inherently, how the light is focused, so your eyes can refocus on near and far objects within the scene even though it's a static image.

But more to the point, a hologram is less like a computer display and more like a photograph.