Ben Horowitz on American Dynamism and the Future of AI | The a16z Show
Summary
A16Z’s more-than-$15 billion new set of funds turns scale into an explicit national-technology obligation. Ulevitch describes it as the largest set of funds in a16z history and later frames it as the largest bet in American history on America winning the next century of technology. Invoking Andy Grove’s warning that an industry leader inherits responsibility for the industry’s ethics and direction, Horowitz defines the firm’s role as helping America win the AI revolution through investment, government integration and citizen access, and allied integration.
Horowitz’s American-dynamism thesis rests on durable rights and the credible promise that builders get “a chance.” America’s opportunity is imperfect, but entrepreneurs can bet their lives on an idea without expecting arbitrary confiscation; in his reading of “self-evident” rights, freedoms come from God or a higher source rather than government, so government cannot take them away.
The U.S. started far behind China specifically in integrating AI with government, even as ChatGPT challenged the broader conventional wisdom that China led overall. Horowitz says American entrepreneurs and a government willing to change the rules are now catching up quickly, a development he finds heartening and sees reflected in current conflicts.
Horowitz reads the Anthropic–Department of War breakdown as a commercial exit, not a genuine philosophical impasse. Because Anthropic was already deployed and held extraordinary leverage as the U.S. was about to go to war, he argues it could have secured anything reasonable it requested, probably more than it should have; walking away and not returning Emil’s calls instead suggests it “wanted out of the deal,” though he does not know why.
His advice to founders starts from the premise that war and crime persist. Employees should not assume they are ethically or geopolitically wiser than the State Department, the Department of War, or intelligence agencies, or decide that personnel risking their lives should not receive the best technology.
The allied industrial thesis is capability-specific: Mexico for high-quality manufacturing and Japan for robotics, supply chains, and defense alignment. Mexico’s expertise can support reindustrialization but requires greater security across a “thin border”; Japan can address America’s robot-supply-chain deficit while its defense spending moves from 0% to 3% of GDP and its interests align with America’s on China.
Venture is becoming a barbell of scaled multi-category platforms and narrow specialists, squeezing firms in the middle. When roughly 15 companies per year could ever reach $100 million in annual revenue, small partnerships made sense; now that “every interesting company that gets started is a technology company,” coverage requires scale, shared services, and centralized authority capable of reorganizing.
Media strategy has flipped from minimizing mistakes to earning attention across unlimited channels. The winning founder or official must be interesting while remaining anchored to a consistent message; if something goes wrong, Horowitz’s prescription is to “do 10 podcasts tomorrow and flood the zone.”
Horowitz’s biggest worry is American pessimism about technology rather than a shortage of technical ambition. He tentatively recalls a poll showing over 70% optimism about AI in China versus less than 30% in America. He relays that someone in Japan said its startup ecosystem is restarting around AI, then makes categorical upside claims: “we’re going to end traffic deaths,” cure cancer, and end poverty as currently understood. Like fire, AI can cause severe harm, but its dangers can be managed without surrendering its benefits.
Deep dive
1. Scale turns technological leadership into an obligation
Ulevitch says the firm has raised over $15 billion in a new set of funds, the largest set of funds in a16z history, and later frames it as the largest bet in American history on the U.S. winning the next century of technology. Horowitz answers the resulting responsibility question with Andy Grove’s warning: for an industry leader, “the ethics of it, the morality of it kind of depends on you.”
Horowitz’s chain runs from opportunity to national power: giving people “a chance” unlocks contribution, America’s superior technology won the Industrial Revolution, and whether it wins the AI revolution remains “an open question.”
Ulevitch’s challenge—why should American technological dominance benefit everyone?—draws Horowitz to “we hold these truths to be self-evident.” In his interpretation, rights originating with God or a higher source rather than presidents or Congress are unusually difficult for government to revoke.
2. Government catch-up exposes the limits of corporate geopolitics
Ulevitch describes American Dynamism as being in its second fund after more than four years, investing at the intersection of advanced software and hardware to rebuild the defense industrial base, manufacturing, and related precursors.
Before ChatGPT, conventional wisdom held that China had a major overall AI lead; Horowitz says ChatGPT suggested America might now have the lead, while China’s lead in integrating AI into military, bureaucratic, and other government systems was real. The surprise has been the speed of U.S. catch-up, driven by willing entrepreneurs and a government prepared to change the rules. Horowitz says current conflicts make the progress visible and give him hope.
Horowitz’s reading of the Anthropic dispute is blunt: “that deal did not fall apart because of philosophical differences.” Already deployed and holding extraordinary leverage as the U.S. was about to go to war, Anthropic could have won anything reasonable it requested, probably beyond what it should have; finding an issue to walk away on and not returning Emil’s calls instead indicates it “wanted out of the deal.”
He preserves uncertainty about the motive—“whether it was an employee thing or Dario’s thing or what have you”—while rejecting the reported ethical-conflict framing. The Department of War has extensive rules, and he says any breach would have a “100% chance” of leaking because “there are no secrets in Washington.”
Horowitz starts his advice to founders from the reality that war and crime persist. He argues that employees should not substitute their judgment for that of the State Department, the Department of War, or intelligence agencies, or deny personnel risking their lives the best technology. Ulevitch calls the opposing impulse “dime-store morality” and “vibe geopolitics.”
3. Allied industrial policy starts with complementary capabilities
Horowitz’s honest hedge is that A16Z has not “entirely figured it out.” Startup culture requires confidence that government will not arbitrarily take away an entrepreneur’s life’s work—a condition he finds globally rare. He names Sweden and Israel as places with strong technology entrepreneurship.
Mexico supplies high-quality manufacturing expertise; Horowitz thinks some American cars built there are better than Chinese manufacturing “on many fronts.” He says the partnership also means helping secure Mexico, a country connected to the U.S. by a “thin border.”
Japan combines enthusiasm for robots with manufacturing strength that can address America’s robot-supply-chain deficit. Horowitz says its defense spending has moved from 0% to 3% of GDP, and its interests concerning China are closely aligned with America’s, creating openings for A16Z, its portfolio companies, and the U.S.
4. Venture capital is splitting into platform giants and specialists
At A16Z’s founding, a study suggested that only about 15 companies started in any year would ever reach $100 million in annual revenue. Then “software is eating the world,” followed by AI, expanded the investable universe until nearly every interesting company became a technology company.
Organizational scale requires reorganization, and “reorgs are essentially redistributions of power.” Shared-control partnerships struggle because people whose authority is being redistributed can block the change; Horowitz credits centralized control with allowing A16Z to keep adapting.
His market-structure forecast leaves large firms covering every technology category and specialists concentrating on AI infrastructure, bio, crypto, or games. Firms in the middle are getting squeezed out. Ulevitch describes A16Z’s 600-plus people as small teams using shared platform services under an effective “benevolent dictatorship.”
Horowitz calls “power” a feature offered to entrepreneurs who lack access to Congress, the right regulatory meetings, and major CEOs. Its cultural counterweight is “first-class business and only in a first-class way”: punctuality, responsiveness, honesty, and respect rather than strong-versus-weak behavior.
5. Media abundance rewards interest, while AI pessimism threatens adoption
Old media strategy emphasized defense because channels were scarce, formats restrictive, and gaffes effectively permanent. With unlimited channels and formats, Horowitz says the priority is no longer avoiding every mistake but “being interesting.”
Alex Karp exemplifies the model: entertaining and hard to predict, yet consistently pro-America. A mistake can be displaced by doing “10 podcasts tomorrow and flood the zone”; the underlying message still must remain coherent.
Horowitz’s biggest worry is the American perception of technology. He recalls a poll that, he thinks, showed over 70% AI optimism in China versus less than 30% in America. He also relays that someone in Japan said the country’s startup ecosystem is restarting around AI.
His counterweight to fear is categorical: “we’re going to end traffic deaths,” cure cancer, and end poverty as we know it. His analogy is fire: it can burn down a village, but it also heats homes and cooks food; technology’s negative aspects can be managed alongside its benefits.