Gavin Baker

New York’s Democratic-socialist sweep showed how disciplined activists can capture safe seats and pull an entire party left. Polymarket put the three-candidate sweep at just 26% before election day, yet Mamdani-backed challengers won in NY-7, NY-10, and NY-13, including two defeats of established incumbents. David Sacks’s key investor-relevant point was organizational leverage: the DSA views Democrats as “a ballot access vehicle,” so even members who survive may alter votes and rhetoric to avoid primaries.

The panel attributed socialism’s momentum less to working-class demand than to downward mobility, institutional failure, and exceptional political communication. Gavin Baker described the emerging base as relatively affluent, highly educated white liberals whose outcomes differed sharply from peers in industry, while Chamath Palihapitiya emphasized housing costs, college debt, healthcare dysfunction, and inflation as proof to younger voters that “the system’s rigged.” Baker’s changed view was explicit: he once considered AOC the Democrats’ standout communicator, but now calls Zohran Mamdani “one of the most talented politicians I’ve ever seen.”

The proposed cure for youth radicalization exposed a fundamental split between child protection and anonymous speech. Chamath Palihapitiya argued that under-16 social-media bans could reduce early exposure and produce “a far less radicalized youth”; Travis Kalanick agreed that social media is “brain rot for real” but said age verification is really a mechanism to deanonymize adults and construct “a full-scale censorship regime.” Baker supported the child-safety objective but judged Kalanick’s free-speech cost “really high.”

Israel has become a primary-election fault line whose intensity is increasingly explained by age rather than party alone. Sacks presented Brad Lander’s defeat of pro-Israel incumbent Dan Goldman as close to a single-issue test, citing 80% Democratic disapproval of Israel and 57% disapproval among Republicans under 50. Chamath insisted critics must separate Jews, Israelis, the state of Israel, and Benjamin Netanyahu: collapsing them into one object of blame is “insane.”

GLM 5.2 turns China’s open-model progress from a strategic warning into an immediate commercial constraint. The 744-billion-parameter, one-million-token-context MIT-licensed model scored 51 on Artificial Analysis, beat GPT 5.5 on Frontier SWE, trailed Claude Opus 4.8 by under one point, and was described as 85% cheaper than GPT 5.5 at comparable performance. With Opus 4.8 rolled back and GPT 5.6 navigating approvals, Sacks warned that America is “on a shot clock”: stopping domestic deployment will not stop China.

Open weights may compress frontier-model margins while expanding the infrastructure opportunity. Baker expects enterprises to run a “council of LLMs,” routing perhaps 85% of work through customized open models and escalating only difficult tasks to frontier systems; his rough current split was open models processing 80%-plus of tokens while frontier tokens capture about 90% of economic value. His conclusion was not that frontier labs disappear, but that composability shifts value toward chips, memory, hosting, and routing.

Micron’s quarter confirmed that DRAM—not exotic components—is the binding AI bottleneck and a source of consumer inflation. Revenue rose from roughly $9 billion to $42 billion, Q4 guidance reached $50 billion versus $43 billion expected, and 2026 HBM supply was already sold out; Baker expects DRAM to absorb 30%-40% of hyperscaler capex next year. Apple’s $699 MacBook Neo moving to $799 and Mac Studio pricing rising 25% illustrated the spillover: “Inflation has come to the desktop.”

Scarce terrestrial power strengthens both orbital-compute economics and the value of large energized sites, while public markets remain able to fund the buildout. Baker put a one-gigawatt terrestrial data center at $35 billion of semiconductors plus $25 billion of power and cooling, versus approximately $40 billion in orbit if reusable Starship lowers launch to $5 billion. He separately valued Anthropic at roughly $3 trillion and argued that public markets need absorb only the offered slice, but Cerebras showed why execution matters: once an IPO breaks deal price, mechanical selling and opportunistic shorts can turn disappointment into “a pile-on.”

Private-company secondaries have become a third exit market: 2025 volume is roughly double the 2021 peak, activity equals 31% of primary venture activity, and pricing swung from $0.80 on the dollar to $1.06. Brad Gerstner calls later-stage names “quasi-public companies,” with secondaries now competing against IPOs and acquisitions for distributions. The organized market nevertheless sits beside “wild west” SPVs charging 10% loads and double carry.

Employee liquidity is economically necessary when companies remain private for seven, 15 or even 24 years, but the panel sees that longer private life as a governance trade-off. Paper wealth of $10m or $30m does not buy a house, while founders may prefer avoiding the public-market microscope. Chamath Palihapitiya described private investors as often telling management what it wants to hear to preserve access, making a successful CEO “the most special flower.” His Facebook example supplied the cost: in 2010 he sought $1bn to build a phone, Zuckerberg backed Bret Taylor’s HTML5 strategy, Facebook went public a year later, and Chamath said “that year made all the difference.” Zuckerberg later said public-market pressure might have changed the decision.

Forge’s deal with Schwab is a wager that regulated infrastructure can turn private-company equity into a genuine retail asset class, combining Forge’s roughly 3m investors with Schwab’s 46m investors and $12tn. Kelly Rodriques said Forge’s 2018–19 SpaceX SPVs were permissioned and that Elon publicly discussed broad IPO distribution at the IPO price. Direct shares and SPVs still require accreditation, but products coming to market as interval funds could hold 60 companies including SpaceX and accept unaccredited investors at $500 minimums: “This is a real asset class.”

Democratization does not make entry price irrelevant: after two of the public market’s biggest months in a decade, Brad said he might put $30k of a fresh $100k to work today. He warned retail against double-fee SPVs, leverage and expecting a late SpaceX purchase to triple immediately; he cited 14 levered ETFs launching for its IPO day around a $1.75tn valuation as another sentiment signal. The goal is “durable democratization,” and “we may not be at the top, but we ain’t at the bottom.”

The new sell button is changing venture behavior as much as access: Brad is selling portions at four or five times cost for DPI, while Jason Calacanis now sells alongside founders once companies reach about $500m. Founders rarely welcome it, but Brad called partial sales a fiduciary duty. Gavin Baker sees the other side too: firms lacking exposure to trillion-dollar winners are protecting their franchises by writing speculative “call options” and “chasing” stories.

The panel rejected a 1999-style collapse analogy because Anthropic, OpenAI and SpaceX are real businesses, but it considered 2021-like valuation compression entirely plausible. Unlike CMGI—no revenue, a move from $2 to $2,000, a purchase of Foxboro Stadium, then failure—today’s leaders have substantial operations; that does not preclude a 10–20% semiconductor-index correction and 30–40% drawdowns in high-beta names. Technology remains a “jagged line up and to the right,” provided investors have staying power.

Below the marquee names, the panel’s ideas concentrated on agent software, modern financial infrastructure, AI networking, robotics and autonomous delivery. Picks included Sierra and Parloa, Revolut, Arrcus and DriveNets, Neura Robotics, Vast and Zipline; the recurring thesis was rebuilding legacy stacks or supplying the specialized infrastructure around AI. Gavin’s networking formulation was “the right chip for the right job at the right time,” while Jason called drone delivery “Uber 2.0.”

Gavin Baker’s answer is no: today’s AI buildout does not resemble the 2000 bubble by either valuation or utilization. Cisco peaked around 150–180 times trailing earnings versus roughly 40 times for NVIDIA; 97% of peak-era fiber was dark, while “there are no dark GPUs” and training clusters are pushing chips until they melt. The largest public GPU buyers have gained roughly 10 points of ROIC since ramping capex—though whether that persists through Blackwell spending remains “an interesting and open debate,” and Baker personally thinks it will.

The infrastructure bill is enormous, but its buyers have an unusually deep balance-sheet buffer. David George framed roughly $1 trillion of existing US data centers, another $3–4 trillion planned over five years, and estimated more than $1 trillion of OpenAI commitments; against that, the major spenders generate about $300 billion of annual free cash flow and hold $500 billion of cash. At $40–50 billion per NVIDIA-powered gigawatt, George sees “an $800 billion buffer growing $300 billion every year,” even if near-term buildout creates some mismatch.

The feared round-tripping is real but, in Baker’s view, small and strategically rational. NVIDIA funding OpenAI while OpenAI buys NVIDIA chips looks circular because “money is fungible,” but Baker says the real driver is competition with Google’s TPU, DeepMind and Gemini—not weak underlying demand. Baker estimates Gemini had taken roughly 15–20 points of traffic share in two or three months and suspects Google may already have more AI traffic than OpenAI or Anthropic on an actual-traffic basis; the cited share gain did not include AI Overviews.

AI could reinforce much of the Mag 7, but execution failure remains existential. Incumbents possess the essential inputs—data, distribution, compute, capital and talent—so AI might be a sustaining innovation if they execute; otherwise, “IBM might be a good fate.” David George called ChatGPT “Pearl Harbor for Google,” while Baker cautions that frontier labs will structurally carry lower gross margins than SaaS because scaling laws and test-time compute keep the products compute-intensive.

Application SaaS is not necessarily dead, but winning requires embracing margin compression. Baker has softened his early-2024 view that all application SaaS “might be a zero,” especially for vendors serving fragmented SMB customers; his warning is that protecting 80–90% gross margins can sacrifice the AI opportunity. The operative choice is “10 bucks of revenue with 90% gross margins or 50 bucks of revenue with 60%,” while incumbents can subsidize break-even AI products before leaders such as Cursor accumulate enough tokens to make catching up difficult.

Distribution and reasoning have made consumer AI less hostile to durable platforms. AI-browser launches could let Google watch the pioneers for three to six months before responding through Chrome’s roughly 5 billion users. Reasoning and RL can turn a large user base into the classic product-data flywheel: users improve the algorithm, which improves the product. David George says GPT-5 is not evidence that scaling laws ended because it was “a smaller model” designed to run more economically, not to maximize capability.

Outcome pricing is the likely business-model shift, with robotics as the physical extension of the same logic. Customer support can charge per resolved task because success supplies a verified reward; personal agents may collect affiliate fees for completed purchases, squeezing the advertiser overpayment that made Google search so lucrative. Baker calls robotics “very real,” expects Tesla versus China, and thinks the humanoid debate is effectively over because robots can learn from video or human demonstrations and receive clean task-level feedback.

Tariff risk has narrowed into a mostly reciprocal regime, but China and AI hardware remain the live exceptions. Gavin Baker said semiconductor tariffs would make it “hard to win in AI,” while the apparent China deal—rare earths for NVIDIA H20s and AMD MI308Xs—and Vietnam’s special carveout for goods transshipped from China show where enforcement remains concentrated. Trump’s renewed celebration of market highs should provide a “dampening mechanism on tariff volatility.”

Replacing Jerome Powell would not solve the fiscal problem and could damage confidence in the Fed’s independence. The initial reports drove stocks down 1% and long yields up roughly 10 basis points; Baker warned that an actual firing would cause a much larger reaction. Meanwhile, the 30-year Treasury reached 5%, versus a 3.3% average rate on $36 trillion of federal debt—math that could lift annual interest expense from $1.2 trillion toward $2 trillion if debt were refinanced at those rates.

The deficit matters now because falling interest rates no longer conceal it. Baker noted that the 30-year yield has risen since Powell began cutting, illustrating why easier overnight money cannot dictate long-term funding costs. Without slower spending, additional revenue and faster growth through deregulation, interest could soon exceed Medicare and Medicaid, Social Security or defense: “The deficit finally does matter.”

Grok 4 delivered what Baker called the largest model leapfrog in some time, despite being trained on Hopper rather than Blackwell. He cited roughly twice the performance of state-of-the-art Google, OpenAI and Anthropic models on the semi-private ARC-AGI-2 benchmark, plus scores in the 40s on Humanity’s Last Exam versus roughly 5% for exceptional humans. Grok 5, “‘05” from OpenAI and the next Gemini, trained on Blackwell, could bring another step-function gain.

The real AI prize is leverage over problems that currently consume decades or millennia, not merely automating office work. Baker separated economically useful AGI from superintelligence whose returns are “definitionally unknowable”; Friedberg reframed the latter as “leverage on time,” using fusion’s 40- or 50-year research cycle as the example. The upside ranges from routine productivity to longevity, abundance and scientific breakthroughs, though Baker preserved the hedge that AI could “go sideways.”

A plausible consumer-productivity market alone underlies Jason Calacanis’s $1 trillion revenue and $10 trillion value thought experiment. At roughly $75-$100 monthly across one billion developed-world users, annual revenue approaches $1.2 trillion; Baker called the arithmetic reasonable but stressed that Google alone was spending about $70 billion that year. Because producing more tokens can create more intelligence through test-time compute and reinforcement learning, “being the low-cost producer is really going to matter in AI.”

Model quality will not decide the AI platform war without distribution. Baker’s maxim was that “on the internet distribution wins championships,” making Apple and xAI natural potential partners as OpenAI moves into hardware and Google, Meta and Microsoft bring entrenched channels. Safari, Chrome, Perplexity’s Comet browser, ChatGPT’s virtual desktop and Grok companions all point toward the browser or agent as the controlling interface.

The GENIUS Act turns regulated dollar stablecoins into payment infrastructure and a prospective source of Treasury demand, while CLARITY remains the next legislative catalyst. GENIUS became law after passing both chambers; Sacks said the leading offshore stablecoin provider would have to come onshore within three years, while Baker said Tether would need to be 100% in Treasuries to enter the U.S. market. Sacks argued that global adoption could produce “dollarizing from the bottom up”; Calacanis kept pressing on Tether, and Bo Hines added that it would be the fourth-largest Treasury buyer that year.