Crypto won the political war through retail ownership; AI completely locked retail out
Retail investors held crypto from the start, enabling them to organize against tightening regulation; but private markets lock ordinary people out of AI unicorn funding rounds. They bear the data-center noise and unemployment risk, yet capture zero capital gains.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
AI labs that monopolize discovery will lose their license to operate
Vlad imagines a future he views as dangerous: AI labs monopolizing discovery by owning data centers, using employee-managed models for in-house research, validation, and publication—essentially ‘making and selling the drug’ like pharma companies. He expects this outcome unlikely, not primarily for technical or market reasons, but because the public won't accept it. He illustrates with an example: a proof of the Navier-Stokes equations or Riemann hypothesis discovered inside an AI company provokes public hostility, but the same proof from an ordinary person using GPT-6 would be celebrated. His conclusion: pursue concentration too far, and AI labs forfeit their license to operate.
— Vlad TenevTokenized stocks spawn unexpected developer moves the moment they hit chain
Robinhood's stock tokenization follows the stablecoin model: assets enter a custody ‘box,’ tokens are minted and burned proportionally, and those tokens flow freely through DeFi to be combined, split, and embedded in lending pools. Robinhood Chain is now fully DeFi-compatible, and the company claims developer activity stands ‘first’ among public chains. Vlad admits developers have already created combinations he never anticipated: meme tokens bundled with stock tokens, stock tokens converted to NFTs. These ‘financial primitives’ are just beginning, and will evolve into foundational infrastructure for developer platforms.
— Vlad TenevKPI contracts let you bet directly on earnings numbers, not stock price
In the past, if you correctly predicted an earnings beat, you could express that thesis only through stock or options—imperfect proxies where the stock often falls anyway due to forward guidance or other unrelated factors, leaving your correct call unrewarded. Robinhood's forthcoming KPI contracts allow you to trade the actual metrics directly: EPS, revenue, and similar figures, bypassing stock price as the imprecise middle variable. Vlad situates these products within the broader prediction market landscape: prediction markets are ‘evolved news,’ turning financial and sports outcomes into odds, and many users open the app simply to see what happens rather than to wager.
— Vlad TenevAgentic trading has achieved product-market fit
Robinhood's agentic trading offering, launched this year on MCP, has demonstrated product-market fit by acquiring over 150,000 new users in just a few months. Tool invocations follow smooth exponential growth, currently averaging approximately 30 million calls per day. At Hood Summit weeks ago, Robinhood unveiled Robinhood Agents, moving the entire trading interface into a harness similar to Claude Code, where users configure daily rebalancing, satellite data integration, ‘skills,’ and ‘Agent Apps.’ Vlad defines the product goal as: placing capabilities that only hedge funds and high-frequency traders could previously afford directly into the hands of ordinary retail investors.
— Vlad TenevPrivate markets lock retail out of AI unicorn returns
When Microsoft and Amazon went public, they were valued at only hundreds of millions; early retail buyers at IPO stage could realize 1000-fold returns. Today an AI lab lists at $1 trillion to $2 trillion; to achieve the same multiple return, the company would need to reach ‘quadrillions’—essentially impossible. Consequently, gains from OpenAI, Anthropic, and similar ventures flow almost entirely to a small cluster of VCs and institutions, leaving ordinary people without any path in. Vlad draws a sharp contrast: crypto, broadly held by retail since inception, allowed the 2024 crypto community to still organize and influence elections and resist regulation even after the 2022–2023 bear market. In contrast, the AI wave leaves ordinary people exposed to data center noise and unemployment threats while yielding them zero capital upside.
— Vlad TenevRobinhood made VC funds tradeable like stocks
To address retail's systemic exclusion from private rounds, Robinhood established two funds. Fund One draws retail capital through Robinhood's own IPO underwriting channels, then invests in late-stage unicorns and companies such as OpenAI, 11 Labs, and Crusoe; the fund itself trades on NYSE like ordinary stock. Fund Two, managed by Rich Aberman, partners with Y Combinator to access seed and Series A rounds—highest risk, highest potential return. Vlad's ultimate vision: enable private companies to trade like stocks or crypto in fractional units 24/7, such that anyone could buy small amounts of Anthropic any time and anywhere.
— Vlad TenevEvery U.S. newborn gets an automatic S&P 500 stake
Robinhood was selected as the sole initial trustee and broker for Trump accounts, working alongside BNY, the Treasury, and the IRS. Beginning January 2025, every child born will receive $1,000 from the Treasury over three years to invest in a low-cost S&P 500 ETF. Michael and Susan Dell have committed over $6 billion in foundation donations directed to children in low-income zipcodes; Gwynne Shotwell has pledged SpaceX shares. Vlad frames the project's ambition as: pushing U.S. household stock ownership from 65% to 95%—what he terms the best product government has ever made.
— Vlad TenevVlad deliberately engineers failure for his children as an experiment
When asked how he maintains such steadiness, Vlad traces it to childhood: his family emigrated from Bulgaria to the United States during hyperinflation, and from age five, he came home alone with his own key and did his homework unsupervised. In comparison, business challenges and even 5% inflation discussions seem minor. His own children enjoy security and resources—tutors, museums—that he lacked, so he deliberately attempts to create similar struggle environments for them, though he admits he ‘hasn't figured out how yet.’ He offers other leaders a reflection inspired by Bezos: developing thick skin is the one thing you can actually control.
— Vlad TenevIn their own words · checked verbatim
It's too centralized, right? It's like AI companies with their big data centers and their employees using secret models to sort of do all this work.
Vlad Tenev14:41
If all of that is happening and it's owned by the AI labs themselves, I think that's a bit of like a dystopian outcome and unlikely to happen for a couple of reasons.
Vlad Tenev16:46
So KPI contracts now allow you to actually trade the, the EPS print, the revenue print, or any of the other KPIs that, that might matter to a company.
Vlad Tenev27:11
we're seeing really strong adoption, like over 150,000, uh, customers have opened agentic trading accounts with us in, in the few months since we've launched it.
Vlad Tenev29:16
if you wanted to invest in a large language model, a couple of years ago, and you wanted exposure to open AI or Anthropic, you're basically out of luck as a as an individual investor.
Vlad Tenev39:38
everyone born in January 2025 onward for the next three years, they get a thousand dollars ceded by Treasury. And Michael and Susan Dell, through their foundation, added another over six billion to children in lower income zip codes.
Vlad Tenev49:16
The best approach to all of these challenges is just grow thick skin. You can control that about yourself, but you can't control the outside world.
Vlad Tenev52:32
Figures
| Agent trading daily tool calls | ~30 million/day | 29:16 |
| Dell Foundation additional donation | over $6 billion | 49:16 |
| Robinhood target household stock ownership | 65% to 95% | 50:17 |
| Bulgaria 1991 inflation | ~170% | 5:10 |
| Bulgaria peak annual inflation | ~2000% | 7:14 |
Glossary
- MCP
- Model Context Protocol; interface enabling AI assistants to call external tools and accounts to execute operations. Robinhood uses it for agentic trading.
- KPI contracts
- Contracts that directly trade specific company financial metrics like EPS or revenue, rather than indirectly betting on stock price movement.
- license to operate
- The implicit social and regulatory assent that allows a company to continue functioning; its loss triggers sharp backlash.
- accredited investor
- U.S. regulatory status that qualifies individuals to invest in private equity and other high-risk assets.
How to listen
Founders and VCs tracking private market access rules, crypto asset policy, and AI commercialization paths. Product and policy people watching how retail financial infrastructure evolves.
44:52–47:05 discusses accredited investor certification exam details that can be skipped; the core argument is already clear in the private funds section.