Most of the models that are hyped up today are going to be commoditized in the next years. The value is not in the model

🇧🇷🇺🇸-  Book: Making Things Think: https://holloway.com/mtt. Investor in Wander, Carry, Footprint, Merkle Manufactory (Farcaster), Dynamic, Paragraph.

Has been the case for a long time no? Most ML use cases on tabular data use variations of xgboost, catboost, lightgbm. All OSS. Same with DL: convnets, transformers, etc. The value has always been in the data.

Tricks with polynomials and Riemann-Roch spaces. Breaking into the platonic realm to pillage forbidden abstractions. Starknet grantee.

I'm helping teach an AI class at Stanford and I have been talking to CEOs of AI companies of significant size. Their insight is that *even* data is not a moat