upvote

points

by jmyeet17 hours ago |
comments
upvoteby JumpCrisscross17 hours ago|
next
[-]
> What do you do if you've invested $1 trillion that will be worth $100 billion or less in 5 years?

I think the aim would be to generate at least $900bn of cash flow from those assets.

reply
upvoteby paulbgd16 hours ago|
prev|
next
[-]
As I was reading the start of your argument, I thought you were gonna call the models a depreciating asset! Totally agree about GPUs too, but literally everything they’re spending money on has to be rebuilt to stay competitive. They have to go for the moonshot of training a full new model when better tech comes, they have to upgrade GPUs to keep their data centers efficient.
reply
upvoteby jmyeet16 hours ago|
parent|
next
[-]
Technically, the model is a depreciating asset too. Just consider the difference between a model you need a B200 cluster to run vs one you can run on a Raspberry Pi. One's going to have a moat around it that gives it value and the other isn't. It's a hyperbolic argument to be sure but the nature of "enthusiast" hardware is that we're currently running, say, ~27B parameter models on hardware for a few thousand. What's that going to look like in 2 years?

Anthropic/OpenAI really need to train ever-bigger models to keep their moat. But that assumes there isn't a law of diminishing returns and also that a compressed model isn't sufficient for what many people need.

You mihgt say that the training is a barrier. And it is, kind of. Notice how it's Chinese companies coming out with open-source models like DeepSeek and Qwen? That's no accident. As soon as DeepSeek came out I knew what was going on: China is going to make sure no single Western company "owns" AI. It's in their national interest for that not to happen.

I wouldn't be surprised if the rush-to-IPO is motivated, at least in part, by getting ahead of Chinese AI commoditization.

reply
upvoteby gen22013 hours ago|
parent|
prev|
[-]
How many failed foundation model training run cycles do you think these companies can tank before the bubble pops and deepseek/etc. catch up to frontier quality?

If Ant, OAI, etc. aren't able to make 20-30% improvements on Opus 4.6 in 2026, does the music stop playing altogether? It seems like they'd lose their ability to charge >10% gross margin on inference in a span of 3-6 months.

reply
upvoteby qaq17 hours ago|
prev|
next
[-]
"OpenAI and Anthropic have invested trillions in a rapdily depreciating asset". Anthropic raised a bit over 100B and has 47B ARR. Where are you getting trillions from ?
reply
upvoteby redox999 hours ago|
prev|
next
[-]
Now that Moores law is dead I think GPUs will depreciate a lot slower. I mean there's already a lot of hardware that has gotten more expensive in the last 5 years.
reply
upvoteby flumpcakes8 hours ago|
parent|
[-]
> I mean there's already a lot of hardware that has gotten more expensive in the last 5 years.

The vast majority of the price rise is mainly due to AI companies sucking all the air out of the room and everyone investing in "AI" regardless.

If China gets their process down to match US/Korea/Taiwan and they decide to flood the market to drown out competitors then hardware is going to be an order of magnitude (or two) cheaper than it is today.

reply
upvoteby redox9912 minutes ago|
parent|
[-]
You're partly right, but we can't ignore that EUV is inherently a lot more expensive, and progress towards new nodes is very slow (current fast hardware will stay fast much longer).

Back in the day the PS3 came with a 90nm GPU and was cost reduced all the way to 65nm, 40nm and finally 28nm. That won't happen again.

reply
upvoteby aurareturn17 hours ago|
prev|
next
[-]
Source on the trillions invested?
reply
upvoteby Rekindle809017 hours ago|
prev|
next
[-]
[dead]
reply
upvoteby tristanj17 hours ago|
prev|
[-]
Starlink Mobile (i.e. Starlink direct-to-cellphone without modifications) is already happening, and fast. Phones that have the recently announced Qualcomm X105 modem will support Starlink Mobile 5G at speeds up to 150Mbps, direct from satellite. The Qualcomm X105 modem will be in most Android flagship phones coming later this year, and by 2027 most new phones will support Starlink direct-to-cell. The next iPhone that supports the 3GPP Rel-19 standard will too.

The rollout relies on Starlink V3 sats, which can only be launched Starship, but Starship progress is going well and is already able to deploy satellites from orbit. SpaceX is capable of launching Starlink V3 on the current iteration of Starship, but they want more testing. We'll probably see Starlink V3 launching late this year or early next year.

reply