A few things:

0) big companies already are very comfortable using contracts to trust other people with their data. Maybe if they're inflexible on the 30-day requirement for fable some orgs will opt out but by-and-large it's already happening and it's not a blocker.

1) Cost will be a blocker. The level of token spend is untenable and the pareto curve is flattening. Most orgs are going to default to either using a distilled model from China or a distilled model from the model companies (e.g, Sonnet 5). It'd behoove Claude/OpenAI to offer a model router before another vendor wins that area.

2) Karp is selling his book. No one knows or cares what an 'ontology' is. From what I can tell, company's product is a tool that helps governments bomb people.

The problem is nobody can build a real business on models they don't control. Cursor focused on making a great AI coding experience but they didn't control Claude and got destroyed once Anthropic started explicitly training Claude to be better at using Claude code. And then there's the cost front, a good model is soon going to be served at 90+% margin, meaning the model makers have a ton of flexibility in how they choose to subsidize their first party products. Cursor had to eat the full API price of Claude while Anthropic could sell Claude Code at a massive subsidy. If you're an AI product then the model product synthesis is your product, and you don't get any of that from a proprietary closed model.

Palantir has sold their flavor of "consulting" (forward-deployed whatevers) as an alternative to the old-school flavors of consulting (management consultants). Now, Alex Karp is upset because there's a new, new flavor of consulting (agents) that they now have to sell against.

Not only are your FDEs old news, but the new hotness is an order of magnitude cheaper. Tough sell.

Is there active research in improving the density of models i.e. higher performance at lower parameter counts? Is this something that's even theoretically possible to do with the architecture?

I feel like I can have a much better model for how I can expect the cost of LLMs to change over the next few years. "Build more datacenters" reeks of "just one more lane bro" and because of Jevons paradox I'm skeptical it will bring costs down. Could ASICs and custom processors (ala Groq) move the needle here at all if the model density can't be improved?

Voice of American *capitalists* used to the idea of investing capital to generate rent. Maybe AI can rebalance the market of capital and individual labor.

undefined

All I can say is, ‘fuck these people’.

No one is actually this stupid.

Archive is giving a 500 error. Anyone got another paywall bypass?

undefined