($) Dot-Com “Dumb Money”
How I look for the AI bubble bursting
Today’s post is for our premium members only. If you are reading this in full, thank you for being an Interconnected Premium member! If you aren’t, I hope you become one by scrolling down and tear down that paywall! 😎
I have been on the road the last few weeks, touching both sides of the Atlantic Ocean. In between flights, shuttles, and long drives with a baby in the backseat, the world of AI seemed to have taken another step function higher. As I catch up on the “everything everywhere all at once” of AI news – Nvidia investing in OpenAI, Alibaba increasing capex and leaning into physical AI, DeepSeek v3.2 release, OpenAI’s multiple product releases and DevDay, the OpenAI-AMD GPU deal (did I miss anything?) – the “AI bubble” discussion has also reached a higher pitch.
In moments like this, the natural reaction is to reach for historical analogs. The most apt analogy, in my view, is the railroad boom and bust during the 1800s in Great Britain and the US. But none of us were alive then, so we resort to accounts like “Railroaded” and other treaties of history to help us make sense of the present. There are other finance books that analyze speculative “bubbles” as a general phenomenon and a repeated symptom of the human condition, like “Devil Take the Hindmost”. (I read this book early this year, when I was already feeling the bubble vibe back then.) But the most common analog people have been reaching for is, unsurprisingly, the dot-com boom and bust, if only because it is the easiest to remember, even if the analogy is weak.
What’s interesting is that many of the movers and shakers of the dot-com era are also the most relevant players driving all the AI headlines today. Larry Ellison of Oracle had a biography written about him in 1997 examining the difference between him and God, before witnessing his net worth drop 62% as the bubble bursted. (In case you are wondering, the difference was: God didn’t think he was Larry Ellison.) Not to be outdone, Masayoshi Son of Softbank saw his own paper wealth fall by 99% in the aftermath of the dot-com crash. Both are now central figures in the consortium that is bankrolling OpenAI’s Stargate datacenters. These behemoth projects, with each data center facility the size of 600 (American) football fields, have become the physical embodiment of the AI boom or bubble depending on who you talk to. They will be filled with many of Jensen Huang’s NVIDIA GPUs; NVIDIA went public in January 1999 and barely survived the crash. These datacenters are being connected by Ciena, a lesser-known darling of the fiber cable construction boom, which saw its stock price peak at more than $1,000 before crashing to less than $20. Its CEO at the time, Gary Smith, is still leading the company today.
If you are afraid of the AI bubble bursting, having the same players driving the train again may give you some comfort. They felt the pain before, so there is no way they would let history repeat itself, right? Then again, arguably the most relevant conductor of the AI train, Sam Altman, was a high school student during the dot-com era. As precocious as he might have been, did he have enough experiences, memories or scars to exercise better judgment now?
Thinking along this thread took me down memory lane. I’m a couple of years younger than Altman, so of the same generation. My first interaction with the Internet happened via dial-up, while living in rural Canada. Altman was living in the midwest of the US at the same time. St. Louis, where he grew up, was of course a bigger concern than the mountain town of 70,000 people where I connected to the interweb for the first time. Perhaps his family was rich enough to give him high speed DSL internet. Mine wasn’t. Nevertheless, neither of us were in the epicenter of tech per se. Would there be any shared experiences of what the bubble looked like back then to me that could be instructive now?
Reaching further into the depth of my hippocampus, there was one glaring experience for me that looked utterly stupid and bubble-licious now: getting paid by AltaVista as a teenager to show ad banners on my PC while pretending to be online.