AI needs $6T in annual revenue to justify data centre boom

It's been almost half a year since we got powerful enough models, and I'm wondering, where are all those great things that were created with them? Maybe it's just me, but nothing has changed in my life so far. No improvements, but definitely, I need to be more careful when reading, listening, watching, and using things, as bad quality slowly creeps everywhere.

They talk a bit about where the AI companies are supposed to get their revenue from. I’d like to see these analyses go a step further and talk about how those AI customers are supposed to come up with that money as well.

For example, the article suggests $1T from enterprise productivity tooling. A quick Google suggests the global enterprise productivity tooling market is only worth about $0.1T right now.

Do we really expect all the world’s corporations to suddenly up their annual spend by 1,000% on average? Why?

I think what's lost on the general public who mainly thinks AI is ChatGPT, is that these companies are gunning for knowledge worker salaries. That by itself is ~$30T/year globally. Nothing else justifies the infra spend. Data center investment is a bet that they will be able to virtualize knowledge workers in the next 5 years.

Warehouse and factory labor is the next largest market for AI/robotics, at about $7T/year globally.

The bubble question is really about the target year. Will these companies start on a trajectory where AI begins eating these massive markets before investors lose faith?

To me it looks like one of these is true:

(1) this is indeed a bubble, and AI will turn into a useful tool integrated into everything but fall short of expectations. We'll see some significant multiple compression in the share prices of these core AI companies, but they'll remain good investments in the long term.

(2) this is a real transformation, the investment was worth it, and these companies will earn massive revenue while causing Depression-era levels of unemployment (24% if half of knowledge workers and factory/warehouse workers can get jobs).

Basically, the revenue required to justify the infra spend requires that they start eating existing industries, because that's certainly not going to come from ChatGPT/Claude subscriptions.

To put this into perspective, healthcare, real estate and banking each earn ~$1.5T in revenue annually. This is REVENUE, not profit which is obviously lower than the revenue in every industry. I fail to understand how AI can achieve $6T in annual revenue when the TAM for healthcare and banking is a quarter of that in the US.

I think the statement is the wrong one.

I think their claim they can make, assuming their evidence is true, is

"AI needs $6T in annual revenue for every investment in AI to be profitable"

I think some people will make a lot of money and more people will lose a lot of money.

Most of those investing know this, they are taking the risk of losing a lot to attempt to be one of the ones that make a lot.

A lot of early internet tech companies went bust, a few became extremely successful. At the time I would see high valuations and ask "Yes, but how do they make money". Most of those with an unclear answer to that question are not around now. The ones who had a tangible plan to make money, even if they were making a loss at the time, tended to survive.

There is a lot of Dumb money going into AI, when that is the dumb money of wealthy individuals, I am more than happy to let them receive no return while paying for infrastructure. There's a problem when the dumb money comes from dumb individuals who are routing money from others who do not have a say in how it gets invested.

That is not an AI specific problem, that is a problem where people can make decisions where others carry the risk without their consent. That is the problem that needs to be met head on.

To avoid this becoming a “tail wags the dog” scenario, folks who have money would need to spend it. We told people for 20 years to “learn to code”. Technology is, outside of healthcare, the leading profession for upward economic mobility.

What happens when the adoption of AI destroys the middle 60% of the bell curve of programmers and technologists? How will the middle class of people spend money on products/services if they don’t have a job?

We can’t both create wealth and destroy jobs with nothing to replace them.

America has little to no manufacturing base.

Technology and finance has been its differentiator economically speaking for job creation. Both are at risk from AI adoption. Businesses have to sell to someone, and at the moment it’s not clear — if the stated goals of revenue are to be believed, how that can happen.

Then again, investments in AI are 1-2 % of worldwide GDP I think, if you think that this technology can probably increase efficiency at least a couple of times in anything that has to do with information processing it doesn't seem so ridiculous. The electric grid, railroads and other large-scale infrastructure required similar amount of investments, though they took much longer. I guess the main risk here is the speed at which everyone races to capture this market as unlike electricity or railroads this revolution is entirely digital so competition happens nearly on a global scale and everything is much faster. That is probably the main risk, going this fast will definitely cause overshoots in one or the other direction.

This is like sitting in 1997, confidently assuming all future Internet services would run on one of Sun, Compaq/HP, DEC, or SGI, and projecting costs accordingly.

  > forecasts that annual spending on AI infrastructure might hit $1.5 trillion by 2031. 
  > ...
  > sustaining this level of investment would require an AI market approaching $6 trillion annually

That segment would include innovations in search, advertising, autonomy and physical AI, analysts at Boston-based Bain said.

Consumer-focused services, which includes subscriptions and advertising revenue, is seen to contribute $200 billion to $400 billion

So not $6T of ARR subscription sales, but $6T of "economic enablement". By 2031, the global economy is estimated to add another $40T of economic activity to GDP. So AI needs to account for 15% of that.

Enjoy the bredth of choice you have with LLMs. Most models will fail and disappear, and those that don't will become a lot more expensive.

I have an obvious but different perspective on the article...

Currently, hyperscalers are growing revenues at an impressive annual rate (~20%), which comfortably covers current annual CapEx. While traditional cloud services still drive a large share of this growth, AI services are becoming an increasingly material contributor. So far, so good.

However, the real question is whether a market actually exists to sustain this CapEx spending at scale relative to revenue growth. The current CapEx growth rate (~50%+ year-over-year) has already outstripped revenue growth, and no one truly knows where the trillions in required future revenue will come from.

If this projected revenue growth fails to materialize, tech equities face severe re-rating risk, given that current stock valuations have already priced in these best-case expectations. A market crash unless a real market is discovered.

In summary, nothing new—perhaps just another red flag.

I can't think of any use case for AI that is a necessity for myself (or people I know), even if coding agents (which I use every day) disappeared tomorrow it would only be a minor inconvenience that I had to write my own code again. Now lets contrast that with electricity or internet, if either of those went offline I would be on a call trying to get them restored asap, now as far as I know neither of those industries bring in $6T in revenue.

Active working population: north america: 195 millions, europe: 366 millions (says ChatGPT).

$6 trillion a year / 195+366m = $900 a month, from the guy flipping burgers to the hedge fund manager. Average salary is in the region of $5000 a month, pre-tax.

I love AI, spend my evening clauding. But still. Call me a sceptic on paying $1000 a month.

Such an interesting dilemma. Arguably an incredibly transformative new industry for which the market cannot fundamentally bear the costs. (or IDK, does it actually work now?) How will it play out??

I don't care how good AI is at something like software development or whatever task. If customers were paying real prices for usage would any of us be using it?

Any data out there on how far off enterprise API pricing is from actual compute COGS for these models right now?

It's gonna have to be so good that it creates entire industries that didn't exist before. That would be awesome, but I'm not seeing it so far.

What's total US annual payrolls? I don't think it's beyond the order of $20 trillion. So they're going to have to replace something on the order of magnitude of half of all jobs OR "create" the difference, in terms of jobs going to AI.

Zitron has been talking about this for some months now. The numbers fluctuate, I think $6T is the highest I've seen but low-end is about $1T that doesn't exist today.

His latest newsletter (from today) also talks about this: https://www.wheresyoured.at/dead-money/

Global electricity provider revenue is $7T, so $6T for a whole new horizontal category of "thing" isn't unreasonable.

The $6T / yr seems to be to justify future data centers much larger than we have now.

I guess either we get human level AGI so it can do human like jobs, or they'll have to trim those plans a good bit?

The sooner it pops the better

Not going to happen, but also not every company will "win" so all of this revenue doesn't need to be recouped, just the revenue of the market leader/leaders.

it does seem a lot like the whole AI thing is the market staying irrational a lot longer than the naysayers can afford not to go along with it, but still, nobody can stay irrational forever.

is their goal to value tokens as some pseudo-currency and create some wacko new age economy like crypto tried and failed to do?

How can investors position to protect themselves?

The high end sectors using these servers are paying way more than joe consumer and his 30 bux a month. We are already seeing justification for the boom, and I'd expect even more advancements to come in multiple sectors. That number is missing from the article.

It's clearly a bubble. I'm not sure why more people don't describe it that way. It's a gold rush. People are hoping that everyone else drops out before they do. Nobody thinks all the participants can win. At best one of them can take a small fraction of $6T.

Feels right and sustainable, ballpark.

as ai becomes an actual replacement for workers, costs will rise to what you would pay a worker, and $6T will be easy

Them AI must make more jobs than the opposite, nobody unemployed will spend cash with shit

[dead]

not really, stop these BS