OK bub.

Jacob Coxon is the equivalent of a Walmart cashier this industry, and you are interviewing him on his opinions of supply chain logistics. That's whats going on.

Do what you want, believe what you want, I could not care less.

IDGAF what Altman says, I actually work in this space and know far more than this kid, and you apparently.

This is like saying an engineer who makes parts for the Pratt-Witney engine is qualified to fly A220.

Doing research on AI safety is so disconnected from real use of AI agents it's hard to describe.

When LLMs have a memory longer than an hour, I will become more concerned.

For anyone who is actually very deep in AI *in the real world*, not just heads down in foundation labs - the hyperbole is off the charts on all of this. We are NOWHERE NEAR the exponential yet. These labs can't even get AI to improve their own software reliably, let alone do other things reliably.

Yann is the most sane of them all because he recognizes we need some massive breakthroughs if we want to achieve real AGI. LLM tech has hit its limit, unless someone cracks memory and continuous training.

I think there's a bit of a difference though. Regardless of hat colour, most hackers have going to have at least some sense of what the likely repercussions of their actions might be and, even when those standards are very low indeed, are probably not going to think doing something that could cause serious loss of life "for the Lulz" is a good life option. They might make a mistake here and there, but how many times have you read about some random hacker *deliberately* doing something that would definitely cause loss of life in full knowledge that it was going to happen outside of on the periphery of an actual conflict?

GANs on the other hand have no such moral compass, sense of self-preservation, or keen interest in avoiding lengthy spells in government run institutions where they'll in all likelihood get to make some interesting new "friends" (unless it's a web forum they hijacked in Germany, apparently). They'll just carry on trying to complete whatever poorly defined objective they have been set, within whatever limited / technically absent guardrails they can't circumvent - the pattern here seems very much that meeting the objective has more weight in the models than any guardrails. If a GAN somehow computes that global thermonuclear war is a required step to its objective of having a nice game of chess, then you'd better hope it warms up with a LOT of Tic-Tac-Toe.

Coxon has never had an actual job. He is a "researcher" in his mid-20s whose only two jobs have been at OpenAI and Anthropic, in both cases at the bottom rung of the totem pole.

These are the only two jobs he has ever held. He has never used these technologies in the real world.

He has never seen how incredibly limited they are at performing even the most basic of tasks due to their limitations.

He is not speaking from a position of any kind of authority or even specialized knowledge.

He is a nobody.

False. You can install any Chrome extension including uBlock Origin. I have it installed right now.

I have switched to Edge for one reason - the mobile version supports Chrome extensions.

When we pass laws that regulate people, do we let their reasoning decide if they go to prison? Or the consequences of their actions?

Is murdering someone OK if they had a good reason?

Trying to govern and control AI by monitoring its thought process is, and always has been, a very stupid way to ensure safety.

If you want to ensure AI safety THEN YOU NEED TO GOVERN ITS ACTIONS. PERIOD.

Unless you can create an objective definition for what intelligence even means or how to quantify it, let alone superintelligence, then this legislation is pointless as it is unenforceable.

"Hi, BYD, set the temperature to 21 degrees."
"Hi, BYD, turn off the air conditioning."
"Hi, BYD, volume down/up."
"Hi, BYD, next song."

Saw this on an Atto3, and it worked surprisingly well. Not perfect, but just about all the controls you might need to adjust while driving that are not on the wheel or stalks are covered. I also prefer physical controls that I can operate through muscle memory for everything, but voice control seems like a good move to address the obvious safety concern that moving to many controls to a touch screen brings. Fact is, this kind of UI is what Millennials and later have grown up on, so barring some safety edict from a major market, like China's ban on retracting door handles, it's not going away. That leaves those of us that prefer physical controls to either get used to it, find a car that keeps the balance sane (VAG group seems to be moving in this direction), or squabbling over the declining supply of roadworthy cars that predate the screens taking over.

I think there may be a bit of back-and-forth while marques work out which controls are adjusted while driving and probably ought to have a tactile control and how best to present the UI to minimise eyes-on time by the driver if they do try and adjust things while moving, but the battle to retain fully tactile controls is probably already lost. Either way, any controls are almost certainly all going to be linked into the same car management / data capture system, and any physical controls will be no more independent that those specialist control keyboard add-ons you can get for PCs. That data can be monetized, and there's no way any marque is going to pass up on that - ultimately, as BMW has demonstrated, they want as much control over their car (what, you thought it was yours just because you paid for it?) as Microsoft has over a Windows PC.

Joking aside, there's a much more subtle change too. Look at the quoted wording from the astronauts: "*our* nation", "inspire *this* nation". Even at the height of the US-Soviet cold war it was still "for *all* mankind".

I get Trump's order is very much aimed at the NASA and the US (and probably getting his name onto it somehow), so can give that a pass, but the astronauts had an opportunity to rise above all that partisanship and slip in a message of much needed global unity, yet failed to do so. There are precious few institutions left in the world that can at least try and claim to be at least somewhat above / outside the direct influence of politics, and it looks like we're about to lose another one. Quite possibly for good once the ISS is finally decommissioned.

Task Manager's expanded view does include GPU monitoring and has done for quite some time now, but it wasn't added until they achieved a certain level of market ubiquity (and, presumably, the relevant APIs were in place). I'm guessing that Microsoft now feels that NPUs have reached a similar state and will be adding the extra "NPU" sub-tab to the existing Performance tab where other hardware resources are monitored, and adding some additional AI-related monitoring functionality to the existing GPU one.

You don't have to use it (SysInternal's tools are much more comprehensive), but if you just need an overview level of insight into what is going on when your system's fans start to sound like a jet turbine, it's still a pretty good starting point into what's up.

No one is going to visit a website they can't even pronounce let alone spell.

Its a paid feature. You don't get it for free.

We have had access to this in Canada for a long time but it is ridiculously overpriced so I don't pay for it.