Peter H. Diamandis
August 25, 2026
The Coming Crisis: Universal Right to AI
TL;DR
Within one year, hardware scarcity will make AI access concentrated among a few companies, forcing society to demand universal basic rights to artificial intelligence.
Key Takeaways
- 1HBM memory and GPU shortages will limit AI access to only Anthropic, OpenAI, and a few others by next year
- 2Next-generation models with 10-20 trillion parameters require hardware that won't be available to the general population
- 3Open source AI models will exist but lack accessible hardware infrastructure to run them
- 4Public demand for universal AI rights will emerge as a major policy issue within 12 months
Notable Quotes
“Nobody right now that I bump into on the street talks about a universal right to AI, but one year from today everybody will be.”
— Dave
“You won't find any place to run it when you start talking about the next generation which are 10 and 20 trillion parameter models.”
— Dave
Chapters
1. Current Indifference to AI Access
Today, few people discuss universal rights to AI, but this will change dramatically within a year
2. Hardware Scarcity Driving Consolidation
HBM memory and GPU shortages will concentrate AI power among Anthropic, OpenAI, and Chinese providers
3. The 10-20 Trillion Parameter Problem
Next-generation models requiring massive computational resources won't be accessible to the general public
4. The Inevitable Demand for Universal AI Rights
Public pressure will force society to address universal basic rights to artificial intelligence as a critical issue
Key People & Entities
- Anthropic
- AI company expected to control access to frontier models
- OpenAI
- AI company expected to control access to frontier models
Glossary
- HBM memory
- High-bandwidth memory used in GPU systems to accelerate AI model processing
- GPU
- Graphics Processing Unit; the primary hardware used to train and run large AI models
- Frontier model labs
- Organizations developing cutting-edge, state-of-the-art AI models
- Trillion parameter models
- Large language models with trillions of learned weights, requiring significant computational resources to operate