7 hours ago
AI will solve everything is the mantra.
What I'd like to see is autonomous vehicles and autonomous solar deployments: do vertical panels on high rise and rooftop without labor. Autonomous vehicles will have millions of lives and tens of millions of accidents per year. Solar is already dirt cheap, flattening the labor cost will be the biggest gamechanger.
Take this challenge AI folks before making claims of AI can solve everything!
8 hours ago
Some decent commentary here on the unit economics of AI, in the sixth section "will AI get switched off after a crash?"
7 hours ago
To take as stab at summarizing (for the folks who prefer text), if there's a crash then:
1. The "GPUs" in these datacenters won't be worth much. (Because they can't easily be used for anything else, not even gaming.)
2. Employers might like a cyber-employee that (versus a human) is 0.5x as good at 0.1x the price, but they won't accept 0.5x at 1.2x the price.
3. They attracted customers with ~1% costs and everyone was super-eager, but recently going towards ~5% everybody recoiled and started rationing.
My only critique is that if we're asking "what rises after the crash and bankruptcy" sense, then #3 shouldn't be using inference+R&D costs, but closer to inference alone because the original company and its R&D debts might be gone.
6 hours ago
I think the counterargument to your critique would be that the R&D costs are dwarfed by running costs, but I'm not sure how true that is.
The one interesting prospect is that of cheap GPUs flooding the market and making pervasive personal AI a realistic possibility.
5 hours ago