rando1234

This is a game changer for local inference IMO, where you will basically never need to do this. On the other hand in a serverless/cloud setting it may be more problematic if you don't want to strand GPUs with locally attached HBF.

What will be the durability/lifetime properties of this technology in comparison to traditional DRAM?

What are you wittering on about. Get a grip.

Amazing looking machine! $380 million a pop. Must surely be one of the most sophisticated machines humanity has ever built? A great achievement of European science and technology (ducks).

Yeah, I feel like Rust has killed Cheri.

Could the registers potentially be used to store encryption keys?

It's funny, I found exactly the same thing when I asked about P=NP. The models outright refused to attempt to solve it, claiming it was too hard. I had to really battle to get it to suggest some promising suggestions.

A related Rust model checking tool more focused on detecting concurrency bugs: https://pure.royalholloway.ac.uk/ws/portalfiles/portal/71989...

Yes, if US tech companies exploitative behaviour is a net negative for children in our society then banning it doesn't seem unreasonable.

I love the 'How we made the...' series on the guardian. They also do it for songs, movies etc. Really interesting to hear the back stories.

So it will be possible to jailbreak it and upload my own files still?

See here for a recent example (albeit not fully autonomous: https://arxiv.org/abs/2605.10402)

The point in the article that 'It's not about optimisations' really got my attention. I've previously done some work where we wrote an analysis pass under the assumption that it executed last in the transformation pipeline and this was needed for correctness. The assumption was that since no further optimisations happened it was safe. Now I'm not so sure...

I recall seeing some pretty damning reports from a security pentester that was able to escape from a container on Azure and found the management controller for the service was years old with known critical unpatched vulnerabilities. Always been a bit sceptical of them since then

Where did the 4.8% number come from? Is it based on the validator stake? How does that compare to the number required to fork Bitcoin as a function of it's supply?

Where we've had some success is with heterogeneous agents with some cheap quantised/local models performing certain tasks extremely cheaply that are then overseen or managed by a more expensive model.

Struggling to find anything interesting or non-obvious about this article. You give a bunch of LLMs various parallelizable task and some models manage to do it well but others don't. No insights as to why. As someone with a distributed systems background the supposed 'insights' from distributed computing are almost trivial.

Interesting perspective re CXL synchronous API. Wouldn't things like OOO execution and speculation help with that? And anyway the latency is supposed to be comparable to NUMA latency, is that really such a deal breaker?

On a related question, I'm in the market to buy a new laptop for development and want to get something with good support for local models. What is a good recommendation in terms of GPU support etc? I currently have a Dell XPS 13. Should I just get a MacBook? Or are there good non-Mac options?

I actually disagree, thought it read reasonably well and didn't feel LLMy at all.

Sure - I was just referring to the use of a Vagrantfile to configure the VM. The start of the article seemed to be pushing the Dockerfile itself as a big innovation.

Didn't Vagrant/Vagrantfiles precede Docker? Unclear why that would be the key to its success if so.

You should read them as publicity to convince stupid politicians to continue to fund basic research when they are more inclined to go for tax cuts for billionaires. Annoying, but a necessary evil.

Has this been peer-reviewed?

You can pay proton anonymously according to some other comments here...

Bombs for peace.

My point is that any Americans claiming moral legitimacy for these actions due to human rights considerations should give us all a break.

And are you really claiming the CIA was not involved in instigating a coup to bring in the Shah?

Don't know why you are being down voted. I mean Iran had a democracy that was toppled by the CIA when they tried to nationalise their resources in favour of a puppet dictator. If the US cared so much about human rights why not go invade Saudi Arabia.

So I suppose you'll be attacking Saudi Arabia after this if you're so worried about humanitarian conditions?