https://scispace.com/pdf/planning-data-center-waste-heat-re-...

And again this is a pretty idealized case-- very nearby housing, optimistic estimated costs.

Other projects analyzed in your paper had "raw" paybacks of 15 years and 17 years-- AKA never in actuality.

When we are talking about protecting the environment, efficiency and return on capital is important. You'd be better dumping those projects into generating more green energy instead of trying to reclaim the DC energy. This is why:

> > Even much hotter low grade heat usually goes unused. You usually need a perfect confluence of a warmer source, very close need for building heat, and a willingness to pay more for environmental friendliness for it all to work out.

If token prices fall a bunch then it may not even be worth leaving on, depending on your facility’s relative power, cooling costs.

If we push far into oversupply eventually a bunch of firms building this infrastructure are going to lose out.

A related argument is speed of intelligence vs capability at that speed. You can think of a three way trade off between latency, cost, and capability that is unlikely to be linear in any dimension and that changes in steps as technology or biology evolves.

Ultimately relating to the properties of the computing substrate and almost certainly bounded by some kind of thermodynamic limits that present systems do not approach.

Then using a heat pump on it makes it even worse.

Low grade heat source is not a term I made up- it refers to heat sources under 100C.

Even much hotter low grade heat usually goes unused. You usually need a perfect confluence of a warmer source, very close need for building heat, and a willingness to pay more for environmental friendliness for it all to work out.

But ripping off pension funds so you can tax me more, as a side effect of enriching people that engage in these shennanigans, doesn’t seem like a great move.

If SpaceX is such a great business, it should be able to win legitimately.

Only about 35 percent is “associated gas” production from oil production.

I would not voluntarily take that bet, but we have regulatory and financial engineering nearly forcing me to do so.

If training models gets way cheaper, I would expect the diminishing returns to get steeper too.

The same argument you’ve made would work for tulip bulbs, dotcom prices, or whatever. Prices go up until they don’t. Exponentials don’t last forever and the intrinsics of technology assets depreciate: things wear out and are also replaced with better things.

Further, space is just reaching the point it can really grow.

Space is strategically important but perhaps not as directly economically valuable as Microsoft yet, and SpaceX is just a fraction of that value.

The problem is, when there is so much overinvestment, everything gets wrecked. In the aftermath of the dotcom boom there was at least a bedrock of fiber and still useful equipment to build upon amid the rubble. This time we are going so much further; also many of the durable assets are misplaced bets and the depreciating ones will depreciate more steeply.

I didn’t read it that way. I see a lot of value in it.

I just don’t see us justifying the amount of infrastructure being built or current valuations. Or in the unlikely event that we do, the societal upheaval is going to take away the ability to monetize it meaningfully.

OpenAI and Anthropic may make it through. But that is different from saying valuations are justified or that all this infrastructure will pay off.