It's more likely that the demand for LLM inference will drop off only once AGI exists.
There are billions of active users already having it do everything from coming up with ideas for Christmas presents, to helping them write e-mails to clients.
That isn't going anywhere until there's a better option on the table. People would have already got bored and moved on if it wasn't doing anything useful for them.
I mean, we really don't have the data to prove this either way.
Meta's training of Llama3 405B model had a 1.34% failure rate for GPUs over the 54 days it ran, across 16387 gpus. It's not likely that all of those faults led to bricked hardware either, they could have just lost part of their performance or memory.
The real question is does that test scale to the long term, often with hardware like this there's a bathtub curve for failure. If those units used were brand new, many of the failures could have just been the initial wave of failures, and there could be a long period of relative stability that hadn't even been seen yet.
GPU based coin mining demonstrated that GPUs often had a lifespan over 5 years of constant use before failure on consumer cards in often less than ideal operating conditions.
You're pulling shit out of your ass at this point, there are some doom reports out of people suggesting that may be a problem, but there are also reports out of other companies(meta for example) with documentation saying the rate is much lower and the mean failure is 6+ years.
The other leftovers from the crash also won't have that problem. It's not just about GPUs. Datacenters and their infrastructure last a lot longer, and the electric generation/transportation networks will also potentially be useful for various alternative applications if the AI use case flops.
They're still somewhat functional for those workloads, and they can even use those low-precision components to emulate high-precision using libraries like cublas, though obviously not as fast as hardware that could do it natively.
It's not like they can't do it at all. It's just that Hopper was better at FP64 than Blackwell is but if Blackwell chips become effectively free due to an AI crash then you could likely still use them in that capacity.
The power costs are nothing compared to the hardware costs for those things.
Getting the massive power required is difficult for datacenters but the usage per unit of compute is actually quite low.
It's important to note that in some previous bubbles, the leftovers of the crash ended up spurring new beneficial growth after.
GPUlike computing power available at scape for essentially free after the ai crash could be used in all sorts of potential ways.
Maybe it makes rending movies with special effects super cheap, and available even to tiny indie studios. Maybe scientists grab it for running physics simulations or disease treatment computations.
I was working at a call center, and applied to become a supervisor, I didn't get the position.
My supervisor at the time gave me some offhand feedback about how I was approaching other people and coming off as too cocky because I was really really good at my current role, I was #1 out of over 1000 agents on my call metrics.
Not sure if it was just the right person, the right time, the right method of delivery, whatever, but it finally clicked for me. That one piece feedback changed my whole personality and has given me so much opportunity over the last 20 years.
I got the supervisor position the next time I applied.
Edit: Just wanted to add that teenagers are stupid, including myself.
Hockey has zero balls, 24 sharp pieces of metal moving around at 20-25mph, and a hard chunk of rubber moving at speeds of up to 100mph.
But I get your point.
That's actually a lot better than some of these. Take the win.
Baaaaaaaby shark do do do do do
That would be AI slop by most standards since the robot voice is "AI" in a way.
I honestly don't think either party is going to pick reduction in population beyond simply cancelling all immigration. And while that would lead to a reduction in the population over time, it's far too slow to handle the unemployment spiking like that.