How AI labs are solving the power problem • Hacker News Board

Re: How AI labs are solving the power problem

#31

Post by beepbooptheory » Wed, Dec 31, 2025, 2:43 PM UTC

FergusArgyll wrote: post #26

pdpi wrote: post #7

Part of what bothers me with AI energy consumption isn't just how wasteful it might be from an ecological perspective, it's how brutally inefficient it is compared to the biological "state of the art" — 2000kcal = 8,368 kJ. 8,368 kJ / 86,400 s = 96.9 W. So the benchmark is achieving human-like intelligence on a 100W budget. I'd be very curious to see what can be achieved by AI targeting that power budget.

You can't compare a training run that produces a file which can be run forever after to a human day

Inference itself is also very costly!

But either way, how many human lives are spent making that file?

Re: How AI labs are solving the power problem

#32

Post by dannersy » Wed, Dec 31, 2025, 2:44 PM UTC

Aurornis wrote: post #17

Apreche wrote: post #6

> An AI cloud can generate revenue of $10-12 billion dollars per gigawatt, annually. Citation needed.

This is coming from a group that does analysis on the semiconductor and cloud industries and provided very expensive access to their models and info. They are the citation.

I mean, if so then they are lying through their teeth.

Re: How AI labs are solving the power problem

#33

Post by einrealist » Wed, Dec 31, 2025, 2:45 PM UTC

goda90 wrote: post #3

Power problem: solved Natural Gas supply problem: worsened Carbon in the atmosphere problem: worsened

The word 'pollution' appears exactly one time in this entire thing, the word 'community' or 'communities' never.

Re: How AI labs are solving the power problem

#34

Post by lukan » Wed, Dec 31, 2025, 2:45 PM UTC

imglorp wrote: post #11

goda90 wrote: post #3

Power problem: solved Natural Gas supply problem: worsened Carbon in the atmosphere problem: worsened

And the air quality around these plants is poor, leading to health problems for the neighbors. This short term, destructive, thinking should be criminalized. I think it's time to discuss changing the incentives around ai deployment, specifically paying into a ubi fund whenever human jobs are replaced by ai. Musk himself raised the idea. https://www.indexbox.io/blog/tech-leaders-push-for-universal...

I assumed gas plants are pretty good in terms of air quality?

Coal plants are bad.

Re: How AI labs are solving the power problem

#35

Post by codingdave » Wed, Dec 31, 2025, 2:46 PM UTC

This is a really long way of saying "We need to burn fossil fuels to make more money."

It didn't make long-term sense for our world before AI. It makes no more sense with AI.

Re: How AI labs are solving the power problem

#36

Post by quikoa » Wed, Dec 31, 2025, 2:47 PM UTC

einrealist wrote: post #33

goda90 wrote: post #3

Power problem: solved Natural Gas supply problem: worsened Carbon in the atmosphere problem: worsened

The word 'pollution' appears exactly one time in this entire thing, the word 'community' or 'communities' never.

This website appears to be very AI heavy in articles. I think it's fair to say these articles are biased because of that.

Re: How AI labs are solving the power problem

#37

Post by teknopaul » Wed, Dec 31, 2025, 2:47 PM UTC

Economic * need dwarfs problems like an overloaded electric grid.

*greed.

We are well past the point that any economic growth at all is anything but a distribution of income problem.

Re: How AI labs are solving the power problem

#38

Post by condensedcrab » Wed, Dec 31, 2025, 2:47 PM UTC

lukan wrote: post #34

imglorp wrote: post #11

Earlier quoted context omitted.

And the air quality around these plants is poor, leading to health problems for the neighbors. This short term, destructive, thinking should be criminalized. I think it's time to discuss changing the incentives around ai deployment, specifically paying into a ubi fund whenever human jobs are replaced by ai. Musk himself raised the idea. https://www.indexbox.io/blog/tech-leaders-push-for-universal...

I assumed gas plants are pretty good in terms of air quality? Coal plants are bad.

Gas plants are not bad… but imagine 400 MW of gas plants in a concentrated area. You’ll always have NOx and SOx by products whenever you’re burning gas.

Re: How AI labs are solving the power problem

#39

Post by airstrike » Wed, Dec 31, 2025, 2:48 PM UTC

Apreche wrote: post #24

Aurornis wrote: post #17

Earlier quoted context omitted.

This is coming from a group that does analysis on the semiconductor and cloud industries and provided very expensive access to their models and info. They are the citation.

So I guess it’s not a bubble then since these companies are raking in the big revenues? Or maybe they are counting all those circular investments as revenues somehow?

Revenue isn't profit, and the presence or absence of either, separately or jointly, isn't sufficient to determine there is a bubble.

Re: How AI labs are solving the power problem

#40

Post by exitb » Wed, Dec 31, 2025, 2:49 PM UTC

pdpi wrote: post #7

Part of what bothers me with AI energy consumption isn't just how wasteful it might be from an ecological perspective, it's how brutally inefficient it is compared to the biological "state of the art" — 2000kcal = 8,368 kJ. 8,368 kJ / 86,400 s = 96.9 W. So the benchmark is achieving human-like intelligence on a 100W budget. I'd be very curious to see what can be achieved by AI targeting that power budget.

How so? A human needs the entire civilisation to be productive at that level. If you take a just the entire US electricity consumption and divide it by its population, you'll get a result that's an order of magnitude higher. And that's just electricity. And that's just domestic consumption, even though US Americans consume tons of foreign-made goods.