# Steven Sinofsky: AI Doesn't Need New Rules Yet > Source: > Published: 2026-07-27 10:00:00+00:00 Steven Sinofsky says AI companies lobbying for regulation are just trying to kill open-source competition — and every AI risk scenario people fear is already illegal under existing law. Jul 27, 202629:16 Difficulty: Intermediate Played The a16z Show Steven Sinofsky: AI Doesn't Need New Rules Yet Steven Sinofsky says AI companies lobbying for regulation are just trying to kill open-source competition — and every AI risk scenario people fear is already illegal under existing law. Jul 27, 202629:16 Difficulty: Intermediate Played TL;DR Steven Sinofsky argues that AI regulation is premature and counterproductive, drawing on decades of tech history to show that regulators consistently move too fast before understanding the technology they're trying to govern[1]— Steven Sinofsky"Every major technology — cars, planes, the internet — took decades before the right regulatory moment arrived. Regulating AI now, before we…"03:30. He makes the case that virtually every AI risk scenario people cite is already covered by existing law[2]— Steven Sinofsky"New AI-specific laws are almost entirely redundant. Non-consensual nudity is already illegal. Discriminatory lending is already illegal. Da…"16:40, and that AI companies lobbying for early regulation are essentially engaging in regulatory capture to kneecap open-source competitors[3]— Steven Sinofsky"Two years ago, AI company executives went to Congress with fear in their eyes, begging to be regulated. That wasn't safety-consciousness — …"13:40. The sharpest takeaway: don't regulate what you don't yet understand. #AI regulation#open source AI#precautionary principle#regulatory capture#U.S.-China AI competition#tech industry lobbying#innovation policy#existing law sufficiency#central planning failure#AI liability#open weight models#chip export controls#tech history parallels#open source#Steven Sinofsky#Microsoft#Anthropic#U.S.-China#innovation#tech history#existing law#central planning#AI safety#protectionism Steven Sinofsky joins Theo Jaffee and Sofia Puccini to argue that AI regulation is moving too fast, open source should remain free, and virtually every AI risk scenario is already covered by existing law. Drawing on tech history from cars to the internet, Sinofsky warns that premature regulation will kneecap innovation and serve corporate interests more than the public. Chapter list Before the formal introduction even begins, Steven Sinofsky states the episode's central argument with striking clarity: the entire conversation about regulating AI is backwards because it starts before anyone understands what's actually being regulated. He challenges the predictive logic behind calls for early intervention, noting that the history of being right about technological futures is poor, and that those making confident predictions about AI's trajectory are almost always advancing self-serving interests. It's a tight, combative opening that sets the argumentative tone for everything that follows. The show's narrator provides a clean framing of the episode: how should governments regulate AI when the technology is still evolving? Theo Jaffee and Sofia Puccini are introduced alongside guest Steven Sinofsky, described as a tech analyst, author of Hardcore Software, and longtime Microsoft leader. The hosts express enthusiasm at his return, and Sinofsky signals his intention to say things that haven't been said yet — a modest promise that turns out to undersell what follows. Sinofsky begins by tracing how every major 20th-century technology evolved for decades before meaningful regulation arrived — not because regulators were asleep, but because the technology itself wasn't yet legible[1]— Steven Sinofsky"Cars took 60 years to get safety rules: It took roughly 60 years of automotive evolution before seatbelts and airbags became upstream desig…"04:16. Cars are his sharpest example: it took roughly 60 years before seatbelts and airbags became standard design criteria, and demanding them from Henry Ford while he was still trying to make engines run would have been absurd. The same pattern holds for antitrust law — vertical integration rules in oil, film, and broadcasting only became coherent once those industries matured enough to understand what monopolistic harm actually looked like. Sinofsky then turns this lens on AI: today's systems are simultaneously being called the most intelligent things ever built and unreliable hallucination machines, and those claims can't both be true at the same time. Social media went through the same contradictory discourse — first a toy, then a government-toppling force, then a mind-control device. The lesson is not that regulation is wrong, but that the story of regulation is always told in hindsight, projecting false clarity onto a messy, unpredictable process of technological emergence. A technical glitch briefly interrupts the flow but Sinofsky resumes without missing a beat. The precautionary principle, Sinofsky explains, isn't just a philosophical error — it's a structural one[1]— Steven Sinofsky"The precautionary principle sounds sensible but leads to one outcome: you freeze the technology at whatever stage the government is most co…"10:12. Regulators don't get hired to not regulate; they come to work to craft rules, and the precautionary principle gives them a mandate to act before anything bad happens. The problem is that early regulation doesn't prevent harm — it constrains the solution set, locking the technology at whatever stage the government is most comfortable with. Sinofsky's illustration is the internet: if precautionary regulation had been applied in the mid-1990s, the government would have blessed two companies — Yahoo and Excite — and frozen web development at Web 1.0, with no video, no audio, no commerce. The companies that were already dominant would have captured the regulatory process, pushing it in their direction. He notes this is a consistent historical pattern: whatever company is ascendant when regulation begins tends to define what regulation looks like. Technology incumbents have historically said 'hands off,' but AI companies have done the opposite — a reversal Sinofsky finds genuinely shocking given his Microsoft background. In one of the episode's most pointed segments, Sinofsky describes watching AI company leaders appear before Congress two years ago and plead for regulation as a genuinely shocking moment[1]— Steven Sinofsky"Two years ago, AI company executives went to Congress with fear in their eyes, begging to be regulated. That wasn't safety-consciousness — …"13:40. Having lived through Microsoft's government investigation beginning around 1992 — triggered by antitrust concerns that most people today wouldn't even recognize — he knows intimately how government scrutiny of tech works. What stunned him about the AI moment was that executives actively invited it. The government had been waiting for this opening for decades, having missed the opportunity to regulate the mainframe, the PC, and the internet. When AI leaders rolled out the red carpet, government walked right in. Sinofsky is careful to distinguish this from naivety: both sides knew what was happening. Government knew companies were trying to shape regulation to their advantage — regulatory capture — and companies knew the government knew. Yet both parties went along because it served their respective interests. The result was a deal that looks like policy but functions like corporate entrenchment, a pattern Sinofsky traces through AT&T, Standard Oil, JP Morgan, Warner Brothers, and the FCC's relationship with radio and television under FDR. The logical conclusion of Sinofsky's argument arrives here, sharp and direct: not 90%, not most — 100% of the risk scenarios people use to justify new AI regulation are already illegal under existing law[1]— Steven Sinofsky"New AI-specific laws are almost entirely redundant. Non-consensual nudity is already illegal. Discriminatory lending is already illegal. Da…"16:40. Senator Warner's four-point AI plan addresses non-consensual nudity? Already illegal — you can't show nude pictures of children and you can't show them to children. States passing laws saying AI can't be registered as an attorney? Already impossible — AI can't take the bar exam, sign a license, or register for college under any existing framework. The laws don't need to be rewritten; they need to be applied. Sinofsky acknowledges that this framing is uncomfortable for legislators, because their entire professional identity is built around creating rules. But his point isn't that government should be idle — it's that the first move should be verification, not legislation. Do the existing statutes actually cover these scenarios? And if they don't, is that because of a real gap in the law, or just imprecise wording that needs updating? Given the episode's anti-regulation thrust, it's refreshing that Sinofsky offers a concrete alternative rather than just critique[1]— Steven Sinofsky"Before writing a single new AI law, governments should audit the 2 million laws already on the books and ask: do they apply to AI? This is …"19:45. The first responsible step, he argues, is auditing existing law: with roughly 2 million statutes covering nearly every conceivable scenario, the question isn't whether to regulate AI but whether current laws already do. He draws a direct parallel to the EV industry, which had to work through every car safety rule designed around combustion engines and verify it still made sense for battery-powered vehicles — including previously non-existent features like front trunks. The same exercise is needed for AI. His most colorful illustration is personal: getting Microsoft Word accepted by the legal system required adding a footnote-continuation feature because legal briefs have strict page-fraction limits for footnotes — a constraint that was invisible with typewriters but became a barrier for word processors. Nobody rewrote legal procedure; the technology adapted. Sinofsky also uses a TV show example (The Pit) to make a parallel point about liability: when a medical resident uses AI to write patient notes and an error occurs, there's no ambiguity — it's her note, her license, her accountability. The law already handles this perfectly without any new AI provisions. The conversation pivots to the geopolitical dimension: both the U.S. and China appear to be moving toward restricting open-source AI models, and Sofia Puccini asks how this plays out[1]— Steven Sinofsky"The U.S. and China aren't in a Cold War or even a classic trade war — they're too economically intertwined for that. What they're fighting …"23:00. Sinofsky's framing is precise: this is an innovation leadership war, not a Cold War (no ideological confrontation) and not a classic trade war (the two economies are too mutually reliant). Government has two tools — use existing regulations as a weapon against international competitors, or fund and nationalize domestic companies to compete. China is clearly doing the latter, pouring national money into AI companies. The U.S. is doing both. But the tools governments actually reach for are indirect: chip export controls, tariffs, import restrictions. Sinofsky calls these 'bank shots' — hitting a side target to affect the main one — and notes they're popular precisely because they're less diplomatically rude than direct action. His Japan parallel is instructive: Japan's Ministry of International Trade and Industry set up an enormous program to dominate global memory chips in the 1980s and partially succeeded — before ultimately losing market leadership. Central planning for technology, Sinofsky argues flatly, has no track record of sustained success. The episode's most provocative moment arrives when Sinofsky names Anthropic directly[1]— Steven Sinofsky"When Anthropic argues that restricting open source is the best way to hurt China, the actual translation is: restricting open source is the…"25:40. When Anthropic argues that restricting open source is the best way to hurt China, Sinofsky translates: what they actually mean is that restricting open source is the best way to help Anthropic by eliminating a competitor — whether that competitor is Chinese or domestic. He calls this un-American and un-tech, noting that AI companies owe their existence to the open academic research community whose papers and code were published freely. Pulling up the ladder now is not just hypocritical; it's a betrayal of the ecosystem that made them. He contrasts AI's behavior unfavorably with Detroit in the 1970s — at least Detroit had the excuse of genuine market distress from fuel-efficient Japanese imports. Even then, the protectionist strategy failed spectacularly: Japan simply built factories in Alabama, Mississippi, and the Carolinas, neutralized the political argument, and Detroit still doesn't effectively make cars anymore. The implication for AI is clear — restricting open source to protect incumbents is a losing strategy that history has already graded. Theo Jaffee wraps up with warm thanks to Sinofsky for what he calls a super interesting conversation. The narrator then delivers the standard a16z Podcast outro, inviting listeners to like, comment, subscribe, and share, and pointing them to the show's YouTube channel, Apple Podcasts and Spotify presence, X account at @a16z, and Substack at a16z.substack.com. The segment closes with the required disclosure that the content is for educational purposes only, is not investment advice, and that some episodes may include paid promotional content from parties unaffiliated with a16z. Precautionary principle A policy approach that calls for regulatory action before harm is proven, on the grounds that potential risk justifies early restriction; Sinofsky argues this freezes technology development at an immature stage. Regulatory capture A phenomenon where the industries being regulated gain disproportionate influence over the regulators, effectively steering rules to serve their own interests rather than the public's. Open weight models AI models whose trained parameters (weights) are publicly released, allowing anyone to download, run, and modify them — distinct from closed commercial models accessible only via API. ASI Artificial Superintelligence — a hypothetical AI system that surpasses human intelligence across all domains; referenced in the debate over whether AI's potential future capabilities justify early regulation. AGI Artificial General Intelligence — an AI system with human-level cognitive ability across a broad range of tasks; often cited by AI labs as a near-term development horizon requiring special governance. Dumping In international trade law, selling goods in a foreign market below their cost of production to undercut competitors; Sinofsky notes this has a specific legal definition and cannot simply be asserted. Universal access A telecommunications policy goal ensuring all citizens have access to essential communication services; AT&T used a universal access promise to secure its government-sanctioned monopoly status. Bank shot A billiards metaphor for an indirect strategy — hitting one target to affect another; Sinofsky uses it to describe trade measures like chip bans that aim to slow AI rivals without directly targeting AI. CSAM Child Sexual Abuse Material — illegal content involving minors; Sinofsky cites it as an example of an AI risk scenario already fully covered by existing criminal law. Kneecap Informal verb meaning to deliberately cripple or weaken a competitor or industry; used repeatedly by Sinofsky to describe the real motive behind some AI regulation proposals. Hallucination In AI, when a language model confidently generates false, fabricated, or nonsensical information — a key limitation used both to argue AI is dangerous and to argue it isn't capable enough to need heavy regulation. Walled garden A closed digital ecosystem controlled by a single company, where users can only access approved content or services; used to describe AOL's model as the kind of internet early regulators would have preferred. Obnoxious Extremely unpleasant or objectionable; Sinofsky uses it to describe AI companies' opposition to open source as morally inconsistent given how much they benefited from open academic research. Beholden Owing a duty or obligation to another party; Sinofsky uses it to describe how private broadcasters and telecoms became effectively arms of the federal government through regulatory deals. Chapter 3 · 02:15 The Regulatory History of Technology: Cars, Antitrust, and the Danger of Getting Ahead of Innovation Sinofsky begins by tracing how every major 20th-century technology evolved for decades before meaningful regulation arrived — not because regulators were asleep, but because the technology itself wasn't yet legible[1]— Steven Sinofsky"Cars took 60 years to get safety rules: It took roughly 60 years of automotive evolution before seatbelts and airbags became upstream desig…"04:16. Cars are his sharpest example: it took roughly 60 years before seatbelts and airbags became standard design criteria, and demanding them from Henry Ford while he was still trying to make engines run would have been absurd. The same pattern holds for antitrust law — vertical integration rules in oil, film, and broadcasting only became coherent once those industries matured enough to understand what monopolistic harm actually looked like. Sinofsky then turns this lens on AI: today's systems are simultaneously being called the most intelligent things ever built and unreliable hallucination machines, and those claims can't both be true at the same time. Social media went through the same contradictory discourse — first a toy, then a government-toppling force, then a mind-control device. The lesson is not that regulation is wrong, but that the story of regulation is always told in hindsight, projecting false clarity onto a messy, unpredictable process of technological emergence. A technical glitch briefly interrupts the flow but Sinofsky resumes without missing a beat. Every major technology — cars, planes, the internet — took decades before the right regulatory moment arrived. Regulating AI now, before we even understand what it is, risks locking in the wrong answers at the worst possible time. It took roughly 60 years of automotive evolution before seatbelts and airbags became upstream design criteria — illustrating why premature regulation stifles innovation. The precautionary principle sounds sensible but leads to one outcome: you freeze the technology at whatever stage the government is most comfortable with. If applied to the internet in the 1990s, we'd all still be using Yahoo and AOL with no video, no audio, no commerce. 10:12 13:40 Chapter 4 · 10:15 The Precautionary Principle and What It Would Have Done to the Internet The precautionary principle, Sinofsky explains, isn't just a philosophical error — it's a structural one[1]— Steven Sinofsky"The precautionary principle sounds sensible but leads to one outcome: you freeze the technology at whatever stage the government is most co…"10:12. Regulators don't get hired to not regulate; they come to work to craft rules, and the precautionary principle gives them a mandate to act before anything bad happens. The problem is that early regulation doesn't prevent harm — it constrains the solution set, locking the technology at whatever stage the government is most comfortable with. Sinofsky's illustration is the internet: if precautionary regulation had been applied in the mid-1990s, the government would have blessed two companies — Yahoo and Excite — and frozen web development at Web 1.0, with no video, no audio, no commerce. The companies that were already dominant would have captured the regulatory process, pushing it in their direction. He notes this is a consistent historical pattern: whatever company is ascendant when regulation begins tends to define what regulation looks like. Technology incumbents have historically said 'hands off,' but AI companies have done the opposite — a reversal Sinofsky finds genuinely shocking given his Microsoft background. Regulating before harm occurs doesn't prevent it — it constrains the available solutions, potentially freezing innovation at an immature stage, like locking the internet at AOL Instant Messenger. Two years ago, AI company executives went to Congress with fear in their eyes, begging to be regulated. That wasn't safety-consciousness — it was regulatory capture. They handed government the opening it had been waiting for since it missed regulating the PC, the internet, and the mainframe. 13:40 16:40 Chapter 5 · 13:50 AI Companies Begging for Regulation: Regulatory Capture in the Open In one of the episode's most pointed segments, Sinofsky describes watching AI company leaders appear before Congress two years ago and plead for regulation as a genuinely shocking moment[1]— Steven Sinofsky"Two years ago, AI company executives went to Congress with fear in their eyes, begging to be regulated. That wasn't safety-consciousness — …"13:40. Having lived through Microsoft's government investigation beginning around 1992 — triggered by antitrust concerns that most people today wouldn't even recognize — he knows intimately how government scrutiny of tech works. What stunned him about the AI moment was that executives actively invited it. The government had been waiting for this opening for decades, having missed the opportunity to regulate the mainframe, the PC, and the internet. When AI leaders rolled out the red carpet, government walked right in. Sinofsky is careful to distinguish this from naivety: both sides knew what was happening. Government knew companies were trying to shape regulation to their advantage — regulatory capture — and companies knew the government knew. Yet both parties went along because it served their respective interests. The result was a deal that looks like policy but functions like corporate entrenchment, a pattern Sinofsky traces through AT&T, Standard Oil, JP Morgan, Warner Brothers, and the FCC's relationship with radio and television under FDR. Sinofsky noted that he had only been at Microsoft for about two years before the government began investigating the company in 1992, illustrating how quickly government scrutiny can arrive. The government missed regulating the mainframe. It missed regulating the PC. It missed regulating the internet. AI regulation is its chance for a do-over — and AI company executives handed them the opening by showing up to Congress and begging for it. That's a once-in-a-generation political gift. Just two years ago, AI company leaders went to Congress and literally pleaded to be regulated — a move Sinofsky called a 'crazy notion' that handed government the leverage it had been seeking over tech. Every AI Risk Scenario Is Already Illegal — So Why New Laws? The logical conclusion of Sinofsky's argument arrives here, sharp and direct: not 90%, not most — 100% of the risk scenarios people use to justify new AI regulation are already illegal under existing law[1]— Steven Sinofsky"New AI-specific laws are almost entirely redundant. Non-consensual nudity is already illegal. Discriminatory lending is already illegal. Da…"16:40. Senator Warner's four-point AI plan addresses non-consensual nudity? Already illegal — you can't show nude pictures of children and you can't show them to children. States passing laws saying AI can't be registered as an attorney? Already impossible — AI can't take the bar exam, sign a license, or register for college under any existing framework. The laws don't need to be rewritten; they need to be applied. Sinofsky acknowledges that this framing is uncomfortable for legislators, because their entire professional identity is built around creating rules. But his point isn't that government should be idle — it's that the first move should be verification, not legislation. Do the existing statutes actually cover these scenarios? And if they don't, is that because of a real gap in the law, or just imprecise wording that needs updating? New AI-specific laws are almost entirely redundant. Non-consensual nudity is already illegal. Discriminatory lending is already illegal. Dangerous drugs are already illegal. Even AI practicing medicine without a license is already impossible under existing licensing law. The only thing new regulation does is give companies a way to kneecap competitors. Sinofsky claims that every scenario critics identify as an AI risk — non-consensual nudity, discriminatory lending, dangerous drugs — already has a law against it. One state passed a law saying AI can't be registered as an attorney — but this is already impossible under existing licensing law, illustrating how redundant new AI rules often are. AT&T, a private company, became effectively the United States national phone company by promising universal telephone access to every address in America in exchange for a government-sanctioned monopoly. Chapter 7 · 19:45 The Responsible Path: Audit Existing Laws, Then Iterate Given the episode's anti-regulation thrust, it's refreshing that Sinofsky offers a concrete alternative rather than just critique[1]— Steven Sinofsky"Before writing a single new AI law, governments should audit the 2 million laws already on the books and ask: do they apply to AI? This is …"19:45. The first responsible step, he argues, is auditing existing law: with roughly 2 million statutes covering nearly every conceivable scenario, the question isn't whether to regulate AI but whether current laws already do. He draws a direct parallel to the EV industry, which had to work through every car safety rule designed around combustion engines and verify it still made sense for battery-powered vehicles — including previously non-existent features like front trunks. The same exercise is needed for AI. His most colorful illustration is personal: getting Microsoft Word accepted by the legal system required adding a footnote-continuation feature because legal briefs have strict page-fraction limits for footnotes — a constraint that was invisible with typewriters but became a barrier for word processors. Nobody rewrote legal procedure; the technology adapted. Sinofsky also uses a TV show example (The Pit) to make a parallel point about liability: when a medical resident uses AI to write patient notes and an error occurs, there's no ambiguity — it's her note, her license, her accountability. The law already handles this perfectly without any new AI provisions. Before writing a single new AI law, governments should audit the 2 million laws already on the books and ask: do they apply to AI? This is what happened with EVs — engineers had to verify that car safety rules designed for combustion engines still made sense for batteries. Do the same for AI, and you'll know what's actually missing. Before writing new AI laws, governments should first audit whether the roughly 2 million laws already on the books apply to AI use cases — covering CSAM, spam, surveillance, and more. If you receive a government grant for computer science research, you are required to release your software as open source — making government opposition to open-source AI internally contradictory. A doctor who lets AI write a patient note and gets it wrong is fully liable — it's her medical license, not the AI's. Accountability doesn't disappear just because AI was involved. This clarity already exists in professional licensing law; we don't need new rules to establish it. The U.S.-China AI Innovation War and the Limits of 'Bank Shot' Policy The conversation pivots to the geopolitical dimension: both the U.S. and China appear to be moving toward restricting open-source AI models, and Sofia Puccini asks how this plays out[1]— Steven Sinofsky"The U.S. and China aren't in a Cold War or even a classic trade war — they're too economically intertwined for that. What they're fighting …"23:00. Sinofsky's framing is precise: this is an innovation leadership war, not a Cold War (no ideological confrontation) and not a classic trade war (the two economies are too mutually reliant). Government has two tools — use existing regulations as a weapon against international competitors, or fund and nationalize domestic companies to compete. China is clearly doing the latter, pouring national money into AI companies. The U.S. is doing both. But the tools governments actually reach for are indirect: chip export controls, tariffs, import restrictions. Sinofsky calls these 'bank shots' — hitting a side target to affect the main one — and notes they're popular precisely because they're less diplomatically rude than direct action. His Japan parallel is instructive: Japan's Ministry of International Trade and Industry set up an enormous program to dominate global memory chips in the 1980s and partially succeeded — before ultimately losing market leadership. Central planning for technology, Sinofsky argues flatly, has no track record of sustained success. The U.S. and China aren't in a Cold War or even a classic trade war — they're too economically intertwined for that. What they're fighting is an innovation leadership war, and the only tools governments have are indirect: ban chips, apply tariffs, nationalize companies. These 'bank shots' rarely hit the target. Sinofsky frames U.S.-China AI competition as an 'innovation leadership war' rather than a Cold War or trade war, noting mutual trade reliance complicates direct economic confrontation. Japan set up a huge national ministry in the 1980s to dominate the memory chip industry — and ultimately failed, showing that centralized government planning rarely wins technology races. When Anthropic argues that restricting open source is the best way to hurt China, the actual translation is: restricting open source is the best way to help Anthropic by eliminating a competitor. Sinofsky calls it un-American and un-tech — an industry that was built on open academic research trying to pull up the ladder behind it. Calling Out Anthropic: Open-Source Restriction Is Competitive Strategy, Not National Security The episode's most provocative moment arrives when Sinofsky names Anthropic directly[1]— Steven Sinofsky"When Anthropic argues that restricting open source is the best way to hurt China, the actual translation is: restricting open source is the…"25:40. When Anthropic argues that restricting open source is the best way to hurt China, Sinofsky translates: what they actually mean is that restricting open source is the best way to help Anthropic by eliminating a competitor — whether that competitor is Chinese or domestic. He calls this un-American and un-tech, noting that AI companies owe their existence to the open academic research community whose papers and code were published freely. Pulling up the ladder now is not just hypocritical; it's a betrayal of the ecosystem that made them. He contrasts AI's behavior unfavorably with Detroit in the 1970s — at least Detroit had the excuse of genuine market distress from fuel-efficient Japanese imports. Even then, the protectionist strategy failed spectacularly: Japan simply built factories in Alabama, Mississippi, and the Carolinas, neutralized the political argument, and Detroit still doesn't effectively make cars anymore. The implication for AI is clear — restricting open source to protect incumbents is a losing strategy that history has already graded. Detroit lobbied the U.S. government to stop Japanese car imports by claiming dumping — Japan responded by building factories in the U.S., but Detroit still lost and no longer makes cars. Detroit convinced the government that Japan was dumping cars below cost. Japan responded by building factories in the American South. Detroit still collapsed. The same playbook is being run today against Chinese AI companies — and historical precedent suggests it won't work either. Two years ago, AI company executives went to Congress with fear in their eyes, begging to be regulated. That wasn't safety-consciousness — it was regulatory capture. They handed government the opening it had been waiting for since it missed regulating the PC, the internet, and the mainframe. When Anthropic argues that restricting open source is the best way to hurt China, the actual translation is: restricting open source is the best way to help Anthropic by eliminating a competitor. Sinofsky calls it un-American and un-tech — an industry that was built on open academic research trying to pull up the ladder behind it. 25:40 27:20 Snapshots () Key Quotes () This episode Claims & Sources 0 / 12 cited (0%) Factual claims made this episode, and whether a source was named. ⚠ It took roughly 60 years of automotive evolution before society understood cars were dangerous enough to mandate safety features like seatbelts and airbags. Steven Sinofskyno source cited ⚠ The U.S. government began investigating Microsoft approximately two years after Steven Sinofsky joined the company, around 1992. Steven Sinofskyno source cited ⚠ About two years ago, leaders of major AI companies appeared before Congress and explicitly requested to be regulated. Steven Sinofskyno source cited ⚠ 100% of the risk scenarios people use to justify AI regulation are already covered by existing laws. Steven Sinofskyno source cited ⚠ If you receive a U.S. government grant for computer science research, you are required to release your software as open source. Steven Sinofskyno source cited ⚠ Japan set up a large national ministry in the 1980s designed to help Japan dominate the global memory chip industry, and ultimately failed to maintain that dominance. Steven Sinofskyno source cited ⚠ AT&T, a private company, became effectively the United States national phone company after promising the government it would deliver telephone service to every address in America. Steven Sinofskyno source cited ⚠ FDR used the Federal Communications Commission to control the messaging of the New Deal by threatening the broadcast licenses of radio stations. Steven Sinofskyno source cited ⚠ Detroit claimed Japanese automakers were selling cars in the U.S. below their cost of production (dumping), and Japan responded by building manufacturing plants in the American South. Steven Sinofskyno source cited ⚠ Detroit no longer effectively manufactures cars, despite years of lobbying for protectionist trade measures against Japanese automakers. Steven Sinofskyno source cited ⚠ Japan perfected high-quality automobile manufacturing through the 1960s after having had a prior reputation for low quality, and also developed small cars suited to its dense urban environment that proved popular in the U.S. Steven Sinofskyno source cited ⚠ There are approximately 2 million laws on the books in the United States covering a wide range of scenarios. Steven Sinofskyno source cited Sign up free to see the claims & sources Create an account to keep exploring every factual claim and its cited source. Invoked as an example of why demanding safety features like airbags at the dawn of the automobile era would have been absurd and counterproductive. Cited as a U.S. government figure who indicated interest in examining IP theft related to AI and open-source models. Cited as an example of a legislator proposing new AI rules — such as addressing non-consensual nudity — that Sinofsky argues are already covered by existing law. Referenced for publicly encouraging open-source AI development in China, contrasting with other regulatory pressures on open-source models. Sinofsky's longtime employer, cited as an example of a closed-source tech company that nonetheless coexisted with open-source alternatives and faced government antitrust investigation. Used as a historical case study of a private company that became a de facto government-sanctioned national monopoly by trading universal access guarantees for regulatory protection. Sinofsky accused Anthropic of framing opposition to open-source AI as national security policy when the real motive is eliminating a business competitor. Used as an example of the type of early internet company that regulators would likely have blessed if the precautionary principle had been applied to the web. Cited as an early regulatory body created under FDR's New Deal to control broadcast licensing and, by extension, media messaging. Cited alongside AT&T and Standard Oil as examples of private companies that effectively became national institutions through government-sanctioned monopoly arrangements. Mentioned as a historical example of a private company that became effectively a national industry arm, parallel to AT&T in telecommunications. Mentioned alongside AT&T and Standard Oil as an example of a private company that became a de facto national institution through regulatory arrangement. Used as a metaphor for the stage at which precautionary internet regulation might have frozen online communication development. Central to the discussion of AI competition, open-source model regulation, export controls, and the framing of a U.S.-China innovation leadership war. Cited as a historical example of both failed central planning for tech dominance (1980s memory chip ministry) and successful competitive strategy (building U.S. auto factories). Stats Episode stats Insight Overview insights chapters Insight distribution Sub-Categories Speaker breakdown Talk Time Sign up free to see the full analytics Cast, category and speaker breakdowns — free with an account.