Having been out of work for a while, people ask me if I’ve been displaced by AI. My reply is “Not directly.” Indirectly, I think the factor is that just about all venture capital funds are throwing money at AI-related companies, meaning non-AI startups that might need technical writers aren’t being funded.
Having lived through the dotcom bust, I have to wonder how bad the fallout from the AI bubble bursting is going to be. The dotcom bubble wasn’t all beenz and pets.com…
…and it fueled a whole lot of subsidiary bubbles: PC and server manufacturers to run the software, Microsoft to run the PCs, semiconductor manufacturers to provide chips for the PCs and servers, semiconductor equipment manufacturers to build those same chips, network gear providers to connect the data centers, etc. And that only scratches the surface. Cisco, Dell, Compaq, Netscape, Yahoo, AOL, Oracle, Sun, HP, Intel, AMD, Applied Materials (where I worked 1997-2001), LAM Research, KLA-Tencor, all had huge growth spurts during the dotcom era as their customers spent big money to get “on the web.” Even dinosaurs like IBM, Motorola and DEC enjoyed business boosts from the era. All suffered in the wake of the dotcom bust, some being bought up or disappearing into other companies.
The same is true of today’s multi-trillion dollar AI boom. Companies like OpenAI may get the most ink, but a whole lot of other companies are getting boosted as well. Some of the names are even the same as the dotcom bubble: Microsoft, Oracle, AMD. Applied Material stock has gone through the roof now that I don’t own any. Cisco is just getting back to the level of their record stock highs during the dotcom era.
Data centers are supposedly planned or going up all around the country, and so many are buying Nvidia’s AI chips that they now boast a breathtaking $4.88 trillion market cap.
Someone is supposedly going to build a $165 billion data center in New Mexico near El Paso. That number is kind of insane, as you could build 5-10 cutting edge fabs for that kind of money. I don’t see how you get any sort of ROI on such a big upfront investment.
Nuclear power is also seems to be enjoying a long-overdue renaissance due to AI, as a lot of companies think it’s just the thing to power those AI data centers. Google plans to restart an Iowa nuclear plant. Fermi America just announced “plans to build a ‘first-of-its-kind behind-the-meter HyperGrid campus‘ back in July, and now it has signed deals to begin the engineering of four nuclear reactors” in Amarillo. (Former Texas Governor and Energy Secretary Rick Perry is also involved.) And the Trump Administration just announced a contract to support Westinghouse’s nuclear power initiatives, though the “aggregate investment value of at least $80 billion” is not the same as some of the “Trump is subsidizing nuclear power with $80 billion” headlines.
When the AI bubble busts (not if, when), a whole lot of these projects will likely come a cropper. A lot of people will have made a lot of money, AI will probably revolutionize a few industries and prove mostly hype in others, and retail investors and bondholders will be left holding the bag. Like the doctom bust, a lot of new companies will rise from the wreckage and start the cycle all over again.
And companies that can best take advantage of idle data centers and newly abundant nuclear power (assuming the boom even lasts that long) will be the ones poised to help build the next tech boom…
Tags: AI, Amarillo, AMD, Applied Materials, Cisco, Economics, El Paso, Fermi America, Google, Microsoft, New Mexico, nuclear power, Nvidia, OpenAI, Oracle, Rick Perry, technology, Texas, Westinghouse
I haven’t seen anybody talking about this, but this is at least the 3rd “AI” boom we’ve had. Though the previous ones probably only barely surfaced above the noise.
During the second one (late 80’s, early 90’s, IIRC) it was expert systems that were going to revolutionize the capture of institutional knowledge. A lot of packages to capture the rules (Level 5 is the only one I remember) and a lot of LISP based stuff (Gensym G2) with inference engines.
At the chemical company I was working at, we did a lot of work trying to connect those things to the SCADA systems in our plants, but not a lot of capturing of that knowledge that was retriring.
Ya, I’ve been making a good return on my investments being mostly in the S&P500 for the last few years but I’m looking at trying to retire around 2030 so I’m trying to figure out where to spread my investments so I don’t get hit too hard when this whole thing comes crashing down.
AI is sufficiently notorious that it forms the measurement standard for bogosity. Like the Farad, the base unit is too large, only microlenats for everyday usage.
http://catb.org/jargon/html/M/microLenat.html
One of my most dour professors in my MBA program in 1996 was adamant about the coming bust of the IT sector. Didn’t take long to come to fruition.
As a nuclear engineer, I certainly welcome the construction of more nuclear power plants but the intent to dedicated them as “behind the meter” (BTM) raises a lot of questions.
First, what is the load factor of a data center? Best answer is that it is about 90% which implies they run full on except for changing cooling needs and internal maintenance.
Second, will the nukes have connections to export some or all of their output to the grid? They will need a big switchyard and detailed studies of what that does or doesn’t do to the public transmission system.
Third, The Atomic Energy Act says a licensed reactor can be commandeered by the government in an emergency. Does that impact question 2?
Fourth, how much backup power can the data center expect from the grid? These are huge loads and supplying them on/off will have huge impacts on the rest of the customers.
I could go on….
Let me know how that all works out when the debt becomes unsustainable within just a few years. The existential problem almost no one’s speaking about because there’s no solution.
Whitehall above is very prescient I think the word is. Best set of AI Questions I’ve seen. Reminds me of the Interurban Era ca. 1900 or so.
AI is still immature as in dumb. ask AI whether castor oil is used to oil casters. Time flies va. fruit flies and arrows and bananas sends AI into fits.
So far, it looks like most capital costs are being funded from equity. I do recall one $3.8B debt offering.
So long as these absurd valuations are liquid, new stock buyers will be at risk. Of course, the premise is, like all bubbles, that the valuations are not likely to meet expectations from revenue.
A deeper problem is that there is just too much stupid money in the system. I’ll give Biden and the Democrat Congress the credit for that.
Nukes are getting the nod because there’s no room to build the data centers near existing base load hydroelectric dams, and that power is already obligated anyway to power cities, so they will need to be built in the middles of their respective nowhere’s where land is cheapest.
But the grid in its current form cannot carry the loads that these massive new data centers will draw to these respective middles of nowhere.
This means either building new natgas plants running as datacenter base load on site, with requirements for pipelines to supply them continuously, which in turn don’t currently exist, or new nuclear plants sited next to the new day centers. Doing the math on a gazillion miles of pipeline right of ways and construction or one modern nuke, the nuke wins, especially if arms are twisted to reduce the regulatory obstructions that have been put in place since TMI.
Hopefully the bubble remains unburst long enough to get those reg changes done, so the post bubble cost math on the new modular designs can compete fairly with other baseload options, at least in states that don’t end in “rnia”.