Depends what you mean by "very few"

https://areweanticheatyet.com/

Check the list, see what your collection looks like.

It cannot do so. Neither "AI" nor a path to achieving it exist. For now you get LLMs.

That aside, the trend over most of the last century now has been that as productivity enhancement tools have continue to improve and workers produce more value, that value keeps getting directed to the most wealthy overall and real wages tend to remain flat. This trend is not going to change without major societal shakeup.

If a large chunk of that $20B is stock ... Groq might do well to unload a chunk before the LLM bubble pops. (Again, I keep saying it, we don't have AI and we don't have a path to it... just LLMs for now). Exact timing may not be "immediately", but...

Your completely shit attitude does nothing to change Windows itself *is* malware anymore.

A global switch would be better than the current state of having to dig into about:config and find all the toggles.
Still, at this point, it's less effort for me to start playing with various forks that have better default settings.
As always, re-evaluate regularly.

Remind me to check how Proxmox's new datacenter management tools are coming along.

Replacing unreliable, poorly informed middle managers that the rank and file have to work around to get their jobs done with unreliable hallucinating LLMs that the rank and file have to work around to get their jobs done...
(And yet wouldn't take much to be better than half the people I dealt with when last at a Fortune X company!)
(How do Fortune X companies function, you ask? Usually despite themselves!)
(For those not initiated, let me give it to you straight: many megacorps actually manage to look dumber from the inside than the news even begins to describe.)

Article mentions gallium arsenide and very briefly issues with using it.

Anyway, research on alternatives to Silicon-based semiconductors comes up in the news every few years or so, and then vanishes again. Research may get some real funding at some point though, as eventually we're gonna run out of ways to get more mileage from Silicon base. Eventually. Probably. Just that there's so much invested into current that trying to catch anything else up for performance seems like a real risk with no guaranteed payout.

Article is a bit light on details. Might be nice to see the kinds of things that were mentioned in articles like this in The Times Before, such as comparisons of electron tunneling voltages, achievable clock speeds, feature sizes, lithography variants in use, feasibility of going 3d and stacking wafers compared to with current processes and so on and so on and so on and so ... ack.

... showing how much every single cellular provider in existence sucks rocks.

Departmental level bodge jobs done by someone in management who at a *stretch* might try to migrate to Access instead of a real DB when pressured to create a better solution because the current one is choked and falling down.

"I want to pick up that thing and take a bite out of it"

Out of touch and taking the good drugs while at work, yup, true denizen of the C-suites.

TSIA, really.
Still, this won't move any kind of needle for the cultists.

Now convince the beancounters who only ever learned Excel and Access that their legacy analytics plugins can be replaced.

... and risks!

Not sure they've thought their cunning plan all the way through.