[cmiles8]

Tl;dr is:

A temporary overvaluation can build enough real capital that the economy lands in a permanently higher-capital equilibrium, even after the inflated valuations correct. The future for AI companies may look rather iffy, but the whole economy may not be as screwed as some fear.

[Avicebron]

The author starts out with a quote from Keynes about the speculation/growth from 1925-1929, is the permanently high-capital equilibrium supposed to happen 10 years after the crash or after we win the world war that follows..?

[aureate]

Surely that outcome requires that the capital retains its value/usefulness. One advantage of crypto and AI is that they can utilise massively parallel computational resources (and don't have tight latency requirements, like gaming) in an age where we've hit the physical limits of sequential computation.

If the "higher capital" that results from an AI boom consists of massively parallel computational resources that currently can only be fully utilised by AI and crypto, and if those things turn out to be a bust, the "higher capital" only has value if we find something else to do with it.

Maybe we will...

[largbae]

Do you believe that machine learning or even specifically LLMs will "bust" out of existence?

The model in my head is more like DotCom telecom. The massive overbuild in fiber was eventually used and even used for the purpose that it was imagined for during the boom. It's just that the companies that built it mostly went under and new owners acquired it at a profit-supporting price.

[Retric]

Most of the cost in a fiber rollout is actual fiber in the ground which could be upgraded by simply swapping a few relatively cheap bits of equipment.

Data centers and electrical infrastructure has a similar long term value, but most of the AI investment is in compute/manufacturing capacity for current nodes which doesn’t age nearly as well.

[zer00eyz]

> but most of the AI investment is in compute

Some people thought that it was misguided when they extended the depreciation cycle of the current AI build out year(s).

In terms of raw performance, there is still some headroom (maybe) but those gains are going to be marginal when you look at the amount of compute per watt (if its more than 5 percent I will be shocked). And that push is going to create a whole other set of problems (cooling is going to be an issue, it already is).

It is fairly likely that this hardware buildout has more legs than one might suspect based on history.

[Retric]

20 year old fiber can have another 10-20 years of useful lifespan remaining, I think compute is going to be valuable for a while but even a 7 year cycle doesn’t change much.

Most compute isn’t going to be new at a potential crash and recovery takes time.

[altcognito]

> compute/manufacturing capacity for current nodes which doesn’t age nearly as well

I mean, compute depreciates, but I think there is zero chance that the value of inference or training is going to fall to zero. Market discovery will find the right price provided the market has the right degree of freedom. Given the type of market it is, I don't see how that won't be the case.

[jaggederest]

I'm a big fan stylistically of what https://taalas.com/ is doing, as far as models baked into silicon. If you haven't tried their chat it's absurdly fast (and also very very dumb)

That implies to me that in the future we'll have models as good or perhaps better than the state of the art at the moment, but on hardware chips that can be put in places where you can't currently locate a datacenter, and operating at hundreds of times better power efficiency, which sounds pretty great.

[lambdaone]

Algorthmic improvements in inference could make all that kit redundant very quickly - there are already moderately capable models that can be run on phones or laptops with specifications that are currently high-end but will be mainstream in another year or so.

This will lead to a superabundance of power-hungry compute power in the hyperscalers, and it's not entirely clear what can be done to consume it all and still run at a profit unless they manage to make ever greater gains for ever more compute-hungry models that cannot be run on consumer devices, unless they refresh their hardware at ever faster and more expensive rates.

The joke about data centers used to be that their core business was selling power at a loss; this may end up being true of the hyperscalers next.

[Retric]

That hardware costs GW of electricity at scale. So barring major disruption in R&D you hit a cost curve cliff where new hardware is simply more cost effective even if existing hardware is free.

Some workloads may make sense running for a few hours a day during cheap solar prices on outdated hardware, but in less than a decade the value is very much hitting zero.

[bluefirebrand]

> the economy lands in a permanently higher-capital equilibrium

Good for the economy, what about the value of the labor that it's currently screwing over?

I don't give a single damn if "the economy" grows if it means my skills become worthless and I become basically unemployable anywhere near my previous earning ability

Edit: even if the value of "the economy" does strongly in the future, is the value of "my labor" ever going to recover?

If no, then fuck it. Why should I care?

[Mistletoe]

The reality will be something more like this-

https://en.wikipedia.org/wiki/Minsky_moment

[chongli]

Doesn't that only apply if the capital is reusable? If we end up with a bunch of data centre GPUs after an AI bubble collapse, there's no guarantee those GPUs will find productive use for other things.

It's like the tulip bubble of the 17th century [1]. Having a bunch of money tied up in useless tulip bulbs didn't do anything productive after the collapse.

[1] https://en.wikipedia.org/wiki/Tulip_mania

[whimsicalism]

yeah how will we ever find a usecase for the machines that can run thinking software

[elefanten]

But it’s not just GPUs that result from it but also a lot of other infrastructure including, prominently, energy.

And beyond physical infrastructure there are the intangible assets: the learning and the process innovation across multiple fields.

The upfront price for all that may end up steep, or fair, or even cheap… the truth is no one knows yet

[chongli]

Energy infrastructure for powering a data centre isn't the same as energy infrastructure for powering a city. One is a simple point-to-point link (power plant to data centre), the other is a grid.

It's like comparing a railway line from a mine to a smelter with a city's road network.

[avianlyric]

Most data centres connect to the grid, they don’t connect to a single power station, except in scenarios where there’s a uniquely low cost power supply nearby, like a small Hydro Plant.

Utility scale power stations have outputs measured in GWs. Data centres are measured in MWs, although people are trying to build GW scale data centres at the moment. But even then a data centre will want a proper grid connection, otherwise they have a massive single point of failure in the form of the directly connected power station.

It’s also very unlikely that purpose built power station is capable of offering cheaper than grid power anyway, except in the very special situations like Hydro. So if you’re gonna build a datacentre, you will want a proper grid connection capable of providing all you needs. Even if you’re running on dirty gas turbines in car park initially while waiting the grid hardening happen. In the long term, that grid connection is always going to be the cheapest, most reliable source of power, ignoring it completely would be foolish.

[chongli]

The premise I was responding to was that the energy infrastructure would be built during the rise of the AI "bubble" and subsequently be repurposed after the collapse. If we're just connecting data centres to the grid then I don't see how that is providing any new infrastructure whatsoever. On the other hand, if we're building lots of new power plants to power the data centres (and connecting them to the grid) but then at some point the data centres are shut down, we end up with a huge glut of unneeded power.

Furthermore, to elaborate on my point above: building new power plants in tandem with data centres that demand GWs of additional power does nothing to address the needs of the grid itself. The grid is not built to handle all that new power, and the consumer electricity demand is not there anyway.

[avianlyric]

You can only get grid connections if the grid thinks they can support your demand. That’s why grid connections take so long, they’re not just plugging in a massive cable, they’re reinforcing other parts of the grid to make sure your massive cable won’t cause a blackout if use all the power it can provide.

Building new power plants immediately isn’t really required to serve data centre, they’re huge amounts of spare capacity in most grids that only exists for extreme situations. The hard part is making sure any existing capacity can actually be delivered to the new data centre. New power plants will get built to rebuild spare capacity over time, or more likely, solar and wind will get built to reinforce that spare capacity over time.

Datacentres are useful grid loads because they all have backup power on site capable of powering the entire site as an island. Which means grid operators can ask them to disconnect quickly if they need to extra power for short periods of time to handle power spikes, or local grid faults. So it’s possible to connect quite a lot of extra data centre type load before generation capacity becomes a serious issue.

[vmesel]

the guy wrote a paper on what a VC is, amazing