The only logical way.

Also, when you find that the first five episodes have lousy sound quality, you skip those.

After a few months, I’d get laid off and have to look for other data center work.

New York has quite a bit going for it. I think we can stand up for ourselves. I think Jersey, Connecticut, and Vermont would join us right out of the gate. I’d certainly support secession.

Additionally, NY plus CA seceding would put way too much pressure on the remainder for the rest of the states to manage the federal government. If Texas secedes for the opposite reasons, that’s the end of it.

Well, I work at an AI hyperscaler. I can tell you how much my facility uses, and how much each rack uses, but don’t have any way to determine what the customer is doing on that server. Or even which servers a given customer is using. Is it being used heavily for queries? How many? Of what kind? We don’t know. Only what the rack/row/pod/hall is consuming.

Also, does the network gear overhead count? How do you apportion that?

We have no visibility into the customer workload. Some of our customers use our systems for scientific research. Drugs, etc. How do you tally that?

I’m not saying that it is impossible, just that if the customer won’t pay for that report, we’re not going to spend money to build the systems to produce it.

Do I agree? No. But I’m just a grunt.

A problem is that the information is not in the hands of the company selling the AI. The actual hardware is often owned by service providers and independent data centers.

I received a bunch of these texts

Thanks

I’ve never heard of it.

Ouch.

I’d like to know if there is are general differences between the branches. Like, is the Air Force more liberal?

Red and green go well together. Perhaps a natural foresty color.

Colors. Society has been getting more monochrome for years. And now black and white houses are all over.

https://qz.com/ai-data-centers-microsoft-amazon-electric-grid-land-1851782495

I would guess it’s diesel as that’s what most MW scale generators for data centers use.

https://en.m.wikipedia.org/wiki/Boltzmann_brain

US

NVL72 will be enormously impactful on high end performance.

So, a lot of our AI customers have no real use for LLM. It’s pharmaceutical and genetics companies looking for the treatments and cures for things like pancreatic cancer and Parkinson’s.

It is a big problem to paint all generative AI with the “stealing IP” brush.

It seems likely to me that an AI may be the only controller that can handle all of the rapidly changing parameters needed to maintain a safe fusion process. Yes it needs safeties. But it needs research, too.

I urge much more consideration of the specific uses of this new technology. I agree that IP theft is bad. Let’s target the bad parts carefully.

New construction work is also much easier than repairs. Commercial is better than residential.

HVAC is a bit of both.