Rising compute costs squeeze small enterprises
The speakers discussed how soaring compute and token prices are making AI deployment increasingly expensive, potentially pricing out smaller businesses.
The argument
Citing an Nvidia executive, the speakers noted that the price of compute has reached a point where it can be more expensive than a human employee. This forces smaller players to rely on open-source or cheaper international models to remain competitive.
The thesis, stress-tested
✓ What validates it
- ✓Model providers officially raising API token pricing
- ✓Small-to-medium businesses reporting lower AI adoption rates due to budget constraints
▸ Risks discussed
- ▸Rapid declines in hardware costs could alleviate pricing pressure
- ▸Open-source models closing the capability gap at lower costs
Hear it yourself
"Now I would remind people they've also been focused on crypto. So at some point, by the end of this year, you're gonna look back and go, oh, the administration was positive on crypto as well. Why didn't I get involved in that? So you still have things that are lagging behind that I believe are part of this bigger plan. But for AI, this involves so many companies that have been left for dead. And the semiconductors alone, it's very top heavy. So NVIDIA, when people hear semiconductors, like, look at these charts. I'm like, the dollars that are going into chips, it's mainly chips."
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE