8 Comments

This is a good argument for social networks to not offer APIs, encourage paywalls, and block scrapers.

Not too radical of an idea, given that we do this with books

Expand full comment
author

Or to offer gated APIs that block scrapers! Check out https://kudurru.ai/. Building something to do this for every piece of content on the internet at Story - turning not just websites but every piece of content programmable.

Expand full comment

Very cool

Expand full comment
Aug 12Liked by Jason Zhao

The complexity of the data labelling process may steadily increase as easier mental constructions continue to be labelled (abstracted) and labels are distributed. Maybe intelligence can be faked with sufficient data and large enough neural networks to the point where we can't say it is fake intelligence.

Expand full comment
author

Yet to be seen whether synthetically generated/labeled data can rival human data. Evidence is mixed, but a crucial question for scaling laws.

Expand full comment

Evidence is mixed.

Expand full comment
Aug 12Liked by Jason Zhao

(3) seems like it could be solved with a shift into creator industries built in Web3 where the companies themselves are decentralized.. bc I think so much of this whole scale-ification is born from 2020+ capitalism where the top is extended out of reach for the lower-level folks. Even on Substack, we have the company itself making money from the users, while the writers/readers do the same but also build dependency on the platform. That currency gets lost in translation bc the idea of 1000 true fans forgot to talk about how platforms yield dependency and creative monetization can't live without mainstream capitalism. Lots to think about here - thanks for sharing!

Expand full comment
author

Totally! Exactly what I was alluding to and what we're doing at Story is using blockchain to power an internet native IP system.

Expand full comment