Asianometry: The Coming AI Startup Bust

Obvious bubble is obvious.

Completely ignoring the wow factor of the underlying technology, this piece focuses on current revenue, potential future revenue, and P/E ratios. The Tulip Mania vibe is strong, looking at fundamentals brings things back down to Earth.

The company I am watching the most in all of this is … Helion Energy. We’ve been pissing money away on Tokamak experiments since the early Cold War, while these guys are on track to be producing power for Microsoft within a couple years. If AI’s need for power forces the fusion market to hurry up, perhaps with a little AI assisted physics along the way, that will be a HUGE win for our species, even if every single AI startup craters hard along the way.

A baseload capable electricity generating method that does not require cooling … the value of that can not be overstated. Right now baseload is fossil fuel, hydro, or nuclear. The first and third require copious cooling water, while hydro uses water directly, and hydro here in the U.S. west is becoming erratic due to drought. The only sources we have that don’t require cooling are solar PV and wind.

If fusion can be made workable there will be a race between it and perovskite solar panels, with the panels taking a share of the market because their production matches air conditioning loads. Couple this with the graphene based advances in desalination and we might just pull off a Houdini with regards to climate change.