Pinned post
Seeing a lot of ai trading posts on the CT TL lately.
TLDR: Frontier LLMs are generalized intelligence models. Specialized models are trained on data sets in a specific domain.
What’s the difference?
Generalized intelligence is great at many things: it can draw your cat in anime style, create a landing page for you, and tell you the latest price of your favorite stock today.
But it is not incredible at one thing.
Specialized intelligence accrues compounding value when you relentlessly shape the models from the ground up on data sets in a specific domain.
Example: Anthropic paid $400m for Coefficient Bio because they built specialized biotech models.
They did not pay Coefficient Bio $400m because they used Anthropic’s models for their bio tech research. Make sense?
@usehotbot is not built on generalized intelligence. We are shipping specialized trading models built from the ground up on data sets in a specific domain.
The idea is simple, but it is not easy.
🔥 Cheers
TLDR: Frontier LLMs are generalized intelligence models. Specialized models are trained on data sets in a specific domain.
What’s the difference?
Generalized intelligence is great at many things: it can draw your cat in anime style, create a landing page for you, and tell you the latest price of your favorite stock today.
But it is not incredible at one thing.
Specialized intelligence accrues compounding value when you relentlessly shape the models from the ground up on data sets in a specific domain.
Example: Anthropic paid $400m for Coefficient Bio because they built specialized biotech models.
They did not pay Coefficient Bio $400m because they used Anthropic’s models for their bio tech research. Make sense?
@usehotbot is not built on generalized intelligence. We are shipping specialized trading models built from the ground up on data sets in a specific domain.
The idea is simple, but it is not easy.
🔥 Cheers
I just think you have 0 edge onchain if you don't have a frontier model pro subscription.View quoted post on X
And if you can't work out why you'd need one then I can already tell you have a negative pnl.