The a16z Show

Podbit · The a16z Show

Beyond P(doom): Marc Andreessen - Betting on America

Explore episode Jun 29, 2026
Technology
Taiwan Is Almost Too Important for Its Own Good

Beyond P(doom): Marc Andreessen - Betting on America · Jun 29, 2026 Technology

The US is completely dependent on Taiwanese semiconductor fabs for advanced AI chips. But that dependency paradoxically makes Taiwan more dangerous, not safer — it turns the island into an irresistible prize for China. Andreessen argues Taiwan is 'almost too important for its own good,' and this is the clearest national security argument for US chip reindustrialization.

Where this was said

Infrastructure Bottlenecks: Every Layer of the AI Stack Is Constrained

At 23:38 · chapter starts 18:30

Shifting to the infrastructure dimension, Andreessen offers a detailed supply-chain taxonomy of AI's physical constraints. At the bottom: energy production, where permitting and generation capacity are already binding. Above that: physical data centers, where turbines are sold out four years in advance and transformers are unavailable — one major hyperscaler is apparently milling its own turbine blades. Cooling systems, large HVAC units, and water cooling are similarly constrained. Inside the data center: NVIDIA GPUs are in tight supply, memory chips face acute shortages with prices spiking and stocks following. And at the raw materials layer: rare earth minerals for advanced semiconductors are becoming bottlenecks. The practical consequence, Andreessen says, is that consumers and businesses are accessing dumber versions of AI than would exist if the supply chain were liberated — models are less capable than they could be because companies don't have the chips and power to train better ones. Meanwhile, the 5-year trend of rapidly declining AI costs is running up against these physical limits, and Andreessen suggests the price of intelligence may actually start rising.

Technology
Supply Chain Bottlenecks at Every Layer of the AI Stack

Beyond P(doom): Marc Andreessen - Betting on America · Jun 29, 2026 Technology

Every layer of the AI infrastructure stack is constrained: energy production, physical data centers, turbines (sold out 4 years), transformers, HVAC cooling, GPUs, memory chips, and rare earth materials. The consequence is that the AI products available today are dumber than they need to be — not because of algorithm limits but because there aren't enough chips and power to train better models.

Similar podbits