Public transcript

Trump's Tech Czar on Crypto and American Innovation

By: a16z crypto
Length: 1 h 18 min
Language: English
Text from: the video's captions
Transcribed: 13 October 2026

Read the full transcript Open on YouTube

Notes in English.

Summary

David Sacks, the Trump administration's AI and crypto czar, talks with a16z co-founders Marc Andreessen and Ben Horowitz and Eric Torenberg about US tech policy. He argues crypto needs regulatory certainty while AI needs freedom from heavy-handed regulation, and lays out the administration's strategy for winning the AI race with China through innovation, energy infrastructure, and exports. He also discusses the Genius Act, the pending Clarity Act, and his concerns about AI doomism and regulatory capture.

Key points

Sacks says crypto's main need is regulatory certainty, while AI suffered from heavy-handed Biden-era regulation that the Trump administration has rolled back.
He accuses Anthropic of regulatory capture, citing a Q&A admission that transparency rules were a stepping stone toward a pre-approval system for new models.
Sacks calls state 'algorithmic discrimination' laws in Colorado, Illinois, and California a path toward 'Orwellian AI' that distorts outputs.
He describes the current AI moment as a 'Goldilocks scenario' — real progress, but no imminent AGI, with humans and AI remaining synergistic.
On China, Sacks says the US leads except in open source models, and criticizes the Biden diffusion rule for driving allies like the Gulf States toward Huawei and Chinese models.
He identifies the gas turbine shortage and grid peak-load rules as near-term energy bottlenecks for data centers, with nuclear 5-10 years away.
The Genius Act covered stablecoins (about 6% of the market); the Clarity Act would provide the framework for the other tokens and Sacks expects it to pass.
Sacks says woke populism appears to be the future of the Democratic Party, and praises San Francisco mayor Daniel Lurie as the best in decades.

Questions it answers

00:38The czar portfolio

Why do AI and crypto fall under one policy role?

Sacks says both are new, poorly understood technologies from the tech industry, and his job is bridging Silicon Valley and Washington culturally and policy-wise. Crypto needs regulatory certainty while AI needs relief from heavy-handed regulation.

Crypto policy is pro-regulation in the sense of providing clear rules.
AI policy aims to unleash innovation so the US wins the global race, mainly against China.

06:36How bad it was for crypto

What did crypto founders actually experience under the Biden administration?

Sacks describes founders being personally debanked so they couldn't open accounts, plus SEC prosecutions under 'regulation through enforcement.' At a March White House crypto summit, one attendee said a year earlier he'd have expected jail over a White House invitation.

Debanking hit both crypto companies and their founders personally.
Sacks calls the lack of clear rules a deliberate strategy to drive crypto offshore.

09:21Regulatory capture in AI

Which AI company does Sacks accuse of regulatory capture and why?

He criticizes Anthropic, describing a speech by Jack Clark where a Q&A admission revealed transparency rules like SB53 were a stepping stone to a pre-approval system for new models, with fear-mongering as strategy. He argues pre-approval systems would destroy the permissionless innovation that made Silicon Valley special.

Sacks sees approval systems for AI software and hardware as a disaster for American competitiveness.
The Biden 'diffusion rule' requiring licensing of every GPU sale was rescinded.

17:01State laws and Orwellian AI

What is Sacks's biggest fear about AI regulation?

He points to roughly 1,200 state AI bills, including 'algorithmic discrimination' laws in Colorado, Illinois and California that make model developers liable for disparate-impact outputs. He says the logical endpoint is a DEI layer in models — 'Orwellian AI' that lies and rewrites history, which he considers a bigger risk than a Terminator scenario.

He wants a single federal standard rather than a 50-state patchwork.
He argues the real AI risk is government using AI to control information, echoing 1984.

22:05State of AI progress

Is AGI imminent, as some claim?

Sacks sees a pullback from the imminent-AGI narrative, citing Andrej Karpathy's view that AGI is at least a decade away. He describes a 'Goldilocks scenario': impressive progress, big productivity gains, but no recursive self-improvement, with humans remaining 'end to end' and AI 'middle to middle.'

He is skeptical of Sam Altman's 2028 automated-researcher claim, arguing a virtual AI researcher would require AGI rather than produce it.
He expects AI to be a synergistic tool rather than wipe out human jobs.

37:01Open source and China

Where does the US stand versus China in the AI race?

Sacks says the US leads in closed models but China leads in open source, which he calls a quirk that may reflect a deliberate Chinese catch-up strategy. He urges encouraging American open source, noting new efforts like Reflection and that a16z is funding more model companies and open source initiatives.

Open source is 'synonymous with freedom' and a hedge against market consolidation and censorship.
Huawei's Cloud Matrix showed China can compete at the system level despite weaker chips.

43:01Winning the AI race

What are the pillars of the administration's AI strategy?

Sacks lists innovation (a single federal standard instead of state overregulation), infrastructure and energy (removing permitting and NIMBY barriers), and exports (selling the American tech stack to allies rather than pushing them to China).

He argues excluding countries from US chips creates a 'Huawei belt and road.'
Near-term data center power depends on gas, with a gas turbine backlog of 2-3 years; shedding 40 hours of peak load could free 80 gigawatts.

57:05AI doomism

Why does Sacks compare AI fear to climate doomism?

He argues the left needs an organizing catastrophe to justify controlling the economy and information space, and AI doomism is replacing faded climate doomism, backed by Hollywood narratives and contrived studies. He says effective altruists pivoted to existential risk after FTX, and their predictions of catastrophe have already been refuted.

He claims Biden staffers explicitly planned to ban open source and anoint two or three winners.
DeepSeek's launch undermined the assumption that China was too far behind to matter.