Why blockchains are the antidote to internet consolidation

Executive overview

The internet promised decentralisation but delivered a handful of gatekeepers capturing most of the money and power. Today five tech companies account for 95%+ of internet traffic and revenue, and AI will accelerate that trend.

Blockchains introduce a missing capability — digital ownership — that can shift money and power back to users.

The consolidation problem

• Web1 was read-only; Web2 added read-write but concentrated control in platforms
• Top five tech companies now represent ~50% of Nasdaq 100 market cap
• Platforms set opaque algorithms, control access, and take 50–100% of revenue flows
• AI rewards incumbents with capital, data, and compute — worsening consolidation
• Creators and users have no ownership of followers, data, or in-network assets

What blockchains actually enable

• Blockchains are a new class of cloud computer that enables digital ownership
• Bitcoin was the first proof: the private-key holder owns the asset, not a platform
• NFTs generalised that concept — ownable art, game objects, social handles
• In a blockchain model, users own their handle, followers, and data — portably
• Low take rates replace platform extraction; services are governed by users
• Email newsletters are an analogy: you own your subscriber list and can leave Substack

Two cultures in crypto

• The casino: speculation, meme coins, tokens bought and sold for short-term gain
• The computer: building internet services with user ownership and low take rates
• Casino behaviour has dominated perception and attracted harmful regulatory focus
• Smart policy would suppress the casino and enable the computer — current US policy does the opposite

The AI crisis and a new content model

• AI is destroying existing creator business models: why pay an illustrator when Midjourney exists?
• Without a new model, the internet risks five services and millions of disenfranchised creators
• Story Protocol (cited as an example) lets creators set terms on IP objects tracked on-chain
• Each blockchain object links to a legal agreement enforceable through existing copyright law
• AI systems can participate as long as they comply with the economic terms set by creators
• This introduces an internet-scale system where anyone can contribute and get paid

Technology cycles and conviction

• AI, the internet, and crypto all followed long incubation periods with ups and downs
• Neural networks date to 1943; AI had a speculative boom in the 1980s before flatlining
• Bad applications of a technology are not evidence the technology itself is bad
• The best opportunities come when a technology is declared dead — crowds arrive only at the top