I asked Grok: “What do you think would happen if Bernie Sanders’ proposal for the government to own 50% of AI companies were to happen? What kinds of incentives or disincentives would this create? What would be the results...”

I asked Grok: “What do you think would happen if Bernie Sanders’ proposal for the government to own 50% of AI companies were to happen? What kinds of incentives or disincentives would this create? What would be the results of these incentives or disincentives?”

If implemented, Sanders’ proposal — a one-time 50% stock “tax” on major AI companies (those hitting ~$200M in annual AI revenue), creating a ~$7 trillion sovereign wealth fund managed by a 7-person presidentially appointed commission with voting shares and board influence — would represent a major shift toward state-directed capitalism in a high-stakes sector.

This isn’t full nationalization but gives the public sector effective veto/power-sharing rights (50% ownership) over core decisions in frontier AI, compute, data centers, robotics, etc. New entrants would face the same once they scale. Here’s a breakdown of likely incentives, disincentives, and results, grounded in economic principles, historical parallels (e.g., state-owned enterprises, sovereign wealth funds like Norway’s or Alaska’s), and innovation dynamics.Short-term effects (implementation shock)

Capital flight and legal battles: Founders, VCs, and employees holding equity would see immediate dilution. This is a massive uncompensated transfer — not a traditional tax on profits. Expect lawsuits (takings clause challenges under the 5th Amendment), lobbying, and attempts to restructure (e.g., spinning off non-AI units, relocating HQs, or accelerating foreign listings). Talent and capital could accelerate moves to friendlier jurisdictions (Singapore, UAE, Switzerland, or even China for some applications). Market reaction: AI-heavy stocks (Nvidia, Microsoft, Google, etc.) would tank on announcement due to uncertainty and dilution. Employee stock options lose value, harming retention. Valuation of the fund itself would be contested and volatile.

Reduced private innovation incentives: Disincentive: Founders and investors risk building something valuable only for the government to claim half upon success. This lowers expected returns dramatically. High-risk, high-reward frontier AI (e.g., AGI pursuits) relies on asymmetric upside for talent and capital. Halving it discourages moonshots. Result: Slower progress on capabilities. Companies shift to safer, incremental applications (e.g., enterprise tools over transformative systems). Less venture funding for startups, as the “exit” or scaling path includes mandatory government partnership. Historical analogy: Heavily regulated or partially state-owned sectors (e.g., European telecom/energy pre-liberalization, or modern critiques of SOEs in many countries) often lag in dynamism.

Political and bureaucratic incentives dominate: Incentive for commission: The 7-person body gains enormous power to “block harmful decisions” and “push helpful ones.” Definitions of “harm” (job displacement, bias, safety, military use, content moderation, pricing) become politicized. Future administrations could steer AI toward favored constituencies (e.g., union protections, green mandates, censorship, or national security priorities). Disincentive for companies: Executives prioritize regulatory compliance and political favor over pure R&D/market fit. Rent-seeking rises — lobbying the commission becomes as important as technical excellence. Result: Regulatory capture in reverse (government capture of firms), slower decision-making, and potential misalignment. AI safety could improve in some views (more oversight) or worsen (bureaucratic delays on genuine risks). Defense/military AI might advance if aligned with U.S. interests, but commercial/consumer innovation suffers. Parallels in state-influenced tech (e.g., certain European or Chinese models) show more caution but less breakthrough speed. Wealth distribution and fiscal incentives: Incentive: A 5% dividend could yield ~$1,000+ per American annually initially, funding UBI-like payments, healthcare, education, or housing. If AI booms, the fund grows and provides a broad revenue base (like Alaska’s oil fund or Norway’s). Disincentive: Governments face weak incentives for efficiency. Losses (if AI hype deflates or competition erodes margins) fall partly on the public balance sheet. Future dilution requirements (50% of new shares) complicate fundraising/employee comp. Politicians treat the fund as a piggy bank, leading to deficits elsewhere or mismanagement (common in sovereign funds under political pressure). Result: Short-term popular payouts, but potential long-term drag on growth. If AI drives massive productivity, total wealth rises — but the U.S. share of global AI leadership shrinks, reducing future tax base and fund value. Inequality might decrease via dividends, but opportunity and dynamism decline. Talent, competition, and global effects: Top researchers/engineers respond to upside and autonomy. Dilution + bureaucracy pushes talent abroad or to smaller, unregulated players. Open-source or decentralized AI efforts might gain relative advantage. Internationally: U.S. AI edge erodes vs. China (state-directed but without Western rule-of-law constraints) or agile allies. National security risks if control hampers rapid iteration.

Optimistic scenario (per proponents): AI advances under democratic guardrails. Public captures more upside, funding social programs and easing disruption (job loss from automation). Sovereign wealth stabilizes economy like resource funds. Pessimistic/base-case scenario (economics of incentives): Innovation slowdown. AI progress fragments and decelerates in the U.S., with breakthroughs happening elsewhere or in gray markets. Bureaucracy adds costs/delays (see historical national champions). Economic growth underperforms counterfactual private-led path. Dividends provide visible benefits but mask opportunity costs (fewer transformative products, lower wages/productivity overall). Precedents like partial state ownership in airlines, energy, or autos often yield inefficiency without eliminating private problems. Unintendeds: Evasion (redefining “AI” businesses), black markets in models, accelerated offshoring, or backlash leading to policy reversal. Employee ownership (as in SpaceX) already shares gains without government control — this overrides that.

In summary, this trades private incentives for rapid, decentralized innovation for centralized redistribution and control. AI’s transformative potential amplifies both upsides (shared gains) and downsides (stifled progress, politicization). Evidence from economics favors competitive markets with targeted regulation/taxation over equity seizures for sustaining technological leadership. Outcomes depend heavily on implementation, commission competence, and global competition — but incentives point toward less dynamism.

Commies have to commie. Ask Russia how that works for productivity and innovation.

State directed capitalism: Hahahahahahahahahaha!

Bureaucrats who never built anything, intelligently directing? Won’t happen.

If the government owned 50% of AI companies, that means it would share 50% in the profits, it would also share 50% in the loses.

IMO, the AI industry will have a day of reckoning when the bottom falls out and multiple companies go bust.

Bernie like most economic illiterates think companies only have profits; they don’t calculate what would happen if the company lost billions and went out of business.

A 5% dividend could yield ~$1,000+ per American annually initially, funding UBI-like payments, healthcare, education, or housing.

The AI firms usually don't pay a dividend. I guess we'd have to sell 5% of the shares per year to give these handouts.

Then we could just go back and tax them again. What could go wrong?

Ask Chinese citizens how they like the 24/7 surveillance state with the Social Credit System (developed for China by Google). That is what we would get with government control. We already have the NSA scooping up all electronic communication, Cell phones eavesdropping and tracking users, and cars logging driving habits and tracking vehicle locations. Government-owned AI centers would weaponize this data for the benefit of government power and control over citizens.

6 posted on 07/14/2026 11:03:39 AM PDT by MtnClimber (For photos of scenery, wildlife and climbing, click on my screen name for my FR home page.)

why do i care what some AI says? especially when i can just pull up a chat window and ask it myself?

7 posted on 07/14/2026 11:27:15 AM PDT by wafflehouse ("there was a third possibility that we hadn't even counted upon" -Alice's Restaurant Massacree)

Heh, were I able to watch the awfulness that would occur after a “man” who’s never had a real job in his life and has demonstrated a knowledge of history, economics, and polics that makes Occasional Cortex and Kamala appear to be Mensa types, it would be a most enjoyable year or so...before the little fugly turd would be hung on a lamp post.

Bernie like most economic illiterates think companies only have profits

Bernie, like most Commie Pinkos, loooves profits that end up in his wallet.

The Commies are afraid of AI. They are absolutely natural enemies.

They want to control AI because of the free flow of information is completely destructive to Socialists. They are basically afraid. Logic is another problem for socialists, they lie, and AI won’t be good at lying for them unless the Socialists can control the narrative.

High Tech and Semiconductor fabs are probably the toughest companies to understand and manage.
Bernie is so far out of his league he is practically on Pluto.
Handing them over to the gov would be handing the future of the USA to China.

AI only works if not a Federal Holidays, State Holidays, Woke Celebrations, Sick days. Also no working on Weekends, after 5 PM or before 9 AM. Start with 13 and move towards 26 vacation (annual leave) days per year.

Get day off to vote, get maturity leave and if AI can't find their socks, day off.

I know it maybe easier to list the days that AI would work........

12 posted on 07/14/2026 12:00:13 PM PDT by Lockbox (politicians, they all seemed like game show host to me.... Sting)

AI takes so many words when humans can express with much less words. You are screwed is the compact version.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.