Featuring
No guest this week—Tucker Carlson's afterglow is strictly a cold open. The besties move from campus speech to household economics, Gemini, and Adobe–Figma, with the sharpest receipts arriving when hindsight tests their predictions about AI search and design competition. Jason has the best episode: his Google monetization call ages beautifully, and his first instinct on Figma beats the panel's later antitrust pile-on.
Spice rack
Would generative AI erode Google's search economics, or give Google more queries and better places to monetize intent?
Original point: Google would remain a major AI player, but direct answers would cannibalize its uniquely dominant search franchise and make the old ad model harder to reproduce.
What everyone argued
Jason Calacanis
Jason takes the opposite side: conversational AI should multiply queries, expose richer purchase intent, and place useful commercial links directly inside planning flows for travel, shopping, parties, tutoring, and other tasks. He predicts a gold mine rather than an ad desert.
David Sacks
Sacks says direct answers are a better user experience than ten blue links but a worse starting point for Google's ad machine. Even if Google builds excellent models, he argues, AI competition and answer-first interfaces should make its franchise less dominant and harder to monetize.
David Friedberg
Friedberg argues that shipping Gemini shows Google is willing to cannibalize Search in service of a better product and trusts the company to solve monetization later. He points to Google's data, product muscle, and user-first tradition as reasons it can manage the transition.
Winner circle
Jason wins the business-model debate. He predicted that AI would generate more queries and expose richer commercial intent, and Alphabet later described almost exactly that pattern while Search revenue accelerated. Friedberg also deserves credit for seeing that deliberate cannibalization was a strength; Sacks identified a genuine strategic risk, but his near-term monetization pessimism aged poorly.
Commentary
Jason Calacanis
Assumptions and fact checks
AI interfaces would create more total questions and more commercially useful intent for Google to monetize.
Why it mattersAlphabet's later disclosures say AI Overviews and AI Mode drove incremental query growth, including commercial queries, and enabled ads on longer searches that were previously hard to monetize.
Query volume would rise by 10x, 20x, 50x, or 100x.
Why it mattersUsage has grown and AI Mode queries are longer, but public disclosures do not support Jason's giant numeric range for total Search activity. The direction aged better than the magnitude.
David Sacks
Sacks asks the right incumbent's-dilemma question but treats the old ad layout as the only possible monetization design. He is stronger on future market share than on the near-term revenue mechanics actually disputed here.
Assumptions and fact checks
Direct AI answers would materially displace traditional search faster than Google could adapt its advertising model.
Why it mattersThe available hindsight shows adaptation keeping pace: AI features increased queries, ads were inserted into the experience, and Search revenue continued strong double-digit growth.
Google's AI market position could never be as dominant as its traditional search position.
Why it mattersThis may still prove right in a more competitive model market, but it is a different claim from whether AI damages Google's Search economics. The latter has not happened in reported results so far.
David Friedberg
Friedberg reads the organizational response correctly but leaves the hardest step inside the phrase 'figure it out later.' Jason supplies the missing mechanism.
Assumptions and fact checks
Google's willingness to cannibalize parts of Search would let it protect the broader franchise rather than freeze in an incumbent's dilemma.
Why it mattersThat is the pattern visible through 2025: Google deployed AI Overviews and AI Mode broadly while Search usage and advertising revenue continued to grow.
Google's unique data advantage was enough to ensure successful monetization of generative AI.
Why it mattersData and distribution helped, but monetization also required product design, advertiser tools, inference capacity, and execution. The advantage was real without being sufficient by itself.
Was the CMA's Adobe–Figma competition case substantively justified, even if the review process took too long?
Original point: The standard competition concern was real because Adobe XD competed with Figma and designers were already using Figma beyond interface work in areas overlapping Adobe's creative tools.
What everyone argued
Chamath Palihapitiya
Chamath agrees with Sacks that a more-than-50% expectation of diminished competition is too speculative and that a 15-month process freezes investment. He treats internal Adobe planning documents as normal strategy work, not proof that successful execution should become legally suspect.
Jason Calacanis
Jason initially argues that XD and Figma competed directly and that Figma's templates and customer use were expanding into decks and marketing collateral. He later joins the criticism of the future-competition standard and proposes aggressive geographic workarounds, so his position becomes less consistent as the segment continues.
David Sacks
Sacks argues the firms occupied distinct markets, Adobe XD had been shut down, and potential future competition was an arbitrary substitute for an objective market-share test. He says the UK was using a novel theory that could chill M&A and drive startups away.
Winner circle
Jason wins the narrow merits question because he identifies the current XD overlap and Figma's broader design trajectory, both later reflected in the CMA's evidence. Sacks and Chamath win an important side point about process speed, but they turn that into an argument against considering future competition at all. The deal's collapse and Figma's later independent growth do not prove every remedy was right; they do make the competition concern look far less fanciful than the panel claimed.
Commentary
Chamath Palihapitiya
Chamath is persuasive on speed and unpersuasive on evidence. Strategic documents are not incriminating because executives used their brains; they are relevant because they reveal how competition was already shaping product choices.
Assumptions and fact checks
Using internal plans to infer future competition unfairly punishes companies for ordinary strategic thinking.
Why it mattersCounterfactual merger analysis necessarily considers plans, incentives, and likely entry. The important safeguard is evidence quality, not refusing to consider future competition at all.
A much shorter and predictable review deadline would reduce unnecessary damage to investment and deal planning.
Why it mattersLong uncertainty imposes real costs on both firms. That process critique can be valid even when the competition concern is valid too.
The CMA said it had to decide whether there was a greater-than-50% chance that the merger would substantially lessen competition.
CheckThe provisional findings explicitly define the expectation standard as a more-than-50% chance of a substantial lessening of competition.
Jason Calacanis
Jason had the winning point in his first minute and nearly talked himself out of it. His direct product evidence aged better than the panel's later Minority Report jokes and VPN workaround.
Assumptions and fact checks
Figma's expansion into broader design workflows made future competition with Photoshop and Illustrator plausible even without current equal market shares.
Why it mattersThe CMA's evidence-based finding supports this direction, and Figma's later independent platform growth makes the competitive trajectory more credible.
Adobe and Figma could neutralize UK review by simply blocking UK customers and proceeding elsewhere.
Why it mattersThe deal also faced a European Commission objection, and jurisdiction cannot generally be escaped with a theatrical consumer geofence after firms have established market activity.
Adobe XD competed directly with Figma in product design.
CheckThe CMA provisionally found that Adobe XD exerted an existing constraint on Figma in product design and that Adobe abandoned development of a closer competing product after the merger agreement.
David Sacks
Sacks has a good process brief and a poor merits brief. He repeatedly answers 'Was the evidence persuasive?' with 'Future competition should not count,' which avoids the strongest version of the regulator's case.
Assumptions and fact checks
Antitrust review should rely mainly on current quantitative market shares rather than evidence about likely future competition.
Why it mattersCurrent shares are important but can miss nascent threats, product roadmaps, innovation incentives, and an incumbent's decision to abandon a competing product. A disciplined counterfactual is not the same as arbitrary veto power.
A prolonged, uncertain review can chill otherwise useful M&A and startup investment.
Why it mattersThis is a legitimate cost and the panel is right to demand timelier decisions. It does not settle whether this particular merger would lessen competition.
The CMA had effectively conceded that Adobe and Figma operated in distinct markets and relied only on possible future competition.
CheckThe CMA provisionally found existing competition in product design through Adobe XD, as well as future competition in image editing and illustration. Its case was not limited to a hypothetical Photoshop entrant.
The UK regulator was essentially acting alone while the EU and U.S. were not presenting the same obstacle.
CheckWhen Adobe and Figma terminated the deal, they explicitly said there was no clear path to approval from both the European Commission and the CMA. The UK was not the only material regulatory obstacle.
Was the late-2023 economy genuinely strong for ordinary households, or did inflation make the good headline data a poor guide to lived conditions?
Original point: The economy could look healthy in aggregate while many working households still felt poorer because the price level had jumped faster than their purchasing power over the preceding inflation shock.
What everyone argued
Chamath Palihapitiya
Chamath argues that raw wealth data shows an economy doing more for more families than the prevailing emotional narrative admits. He blames media and social reinforcement for widening the gap between measured conditions and public perception.
Jason Calacanis
Jason accepts the sticker shock but says the contemporaneous labor market and GDP numbers were extraordinary: wages had begun beating inflation, unemployment was near generational lows, and third-quarter output growth was close to 5%. He treats those figures as evidence of underlying confidence despite household complaints.
David Sacks
Sacks argues that disinflation was not deflation: prices were still rising from a much higher base, so households could remain worse off even after the inflation rate fell. He says cumulative purchasing power, not a single good quarter, explains the public's sour view.
David Friedberg
Friedberg separates the economist's aggregate scorecard from the household's own test: whether income adjusted for purchasing power is producing visible lifestyle progress. He argues that both accounts can be true without dismissing the public as irrational.
Winner circle
Friedberg has the best answer because he keeps the national scorecard and household scorecard separate. Jason is right about the strong snapshot, and Sacks is right about the accumulated price-level pain, but each overstates one side of the ledger. Hindsight supports a strong macro recovery with frustratingly little median-household progress—not a hidden depression, and not an economic victory lap.
Commentary
Chamath Palihapitiya
Chamath is right to demand data, then asks one wealth chart to carry far more weight than it can. The argument needed a clean bridge from balance-sheet wealth to the median household's disposable purchasing power.
Assumptions and fact checks
Negative economic sentiment was driven more by partisan and media framing than by a broad deterioration in household purchasing power.
Why it mattersPartisan framing plainly affects surveys, but cumulative inflation and essentially flat real median household income versus 2019 gave households a material reason to feel little progress. Both mechanisms mattered.
A rising share of millionaire households is strong evidence that the economy is broadly serving ordinary families.
Why it mattersNominal asset thresholds rise with inflation and asset-price gains can be concentrated. The statistic does not, by itself, answer what happened to median income or day-to-day purchasing power.
Jason Calacanis
Jason has the strongest snapshot and the weakest adjective. The economy was not secretly collapsing, but 'extraordinary' skipped the distinction between fast aggregate output and a median household that had only begun regaining lost purchasing power.
Assumptions and fact checks
High credit-card balances and 401(k) hardship withdrawals mainly showed that consumers felt secure about jobs and future income.
Why it mattersBorrowing can reflect confidence, distress, or both; hardship withdrawals are especially weak evidence of comfort. Jason assigns a favorable motive without separating these channels.
Strong point-in-time GDP and employment data were sufficient to describe the household economy as extraordinary.
Why it mattersThey justified calling the macroeconomy strong, but not erasing distribution, price-level, housing-affordability, and cumulative-income concerns.
Wages were slightly outpacing inflation by late 2023.
CheckBLS reported real average hourly earnings were positive year over year in November 2023, and its full-year review found real hourly earnings rose in 2023 for both all employees and production and nonsupervisory workers.
Unemployment was at a 50-year low when the episode recorded.
CheckThe November 2023 unemployment rate was 3.7%. It was historically low, but above the 3.4% reached earlier in 2023, so the precise present-tense claim was overstated.
GDP growth was about 5%.
CheckReal GDP grew at a 4.9% annualized rate in the third quarter of 2023. That was a quarterly annualized rate, not the full-year growth rate, so the number was accurate but narrower than Jason's broad conclusion implied.
David Sacks
Sacks identifies the stock-versus-flow error that the optimistic case underrates. He would have been stronger with a defined cumulative income series instead of treating 'working class' as a self-proving statistic.
Assumptions and fact checks
Cumulative inflation explained a substantial share of why many households rejected upbeat descriptions of the economy.
Why it mattersThis is strongly plausible and consistent with income data showing the median household had made little statistically meaningful progress over the full 2019-2024 span.
Working-class wages had broadly failed to keep pace with prices over the relevant multi-year period.
Why it mattersThe answer depends on endpoints, worker group, hours, and income measure. Real wages were rising again in 2023, while median household income remained roughly flat versus 2019; Sacks's broad wording hides that split.
A falling inflation rate means prices are rising more slowly, not that the overall price level has fallen.
CheckThat is the correct distinction between disinflation and deflation, and the November 2023 real-earnings release still showed a positive year-over-year CPI increase alongside positive real wage growth.
David Friedberg
Friedberg wins by refusing the false choice between 'the data is fake' and 'the public is irrational.' His framework survives hindsight even though his 10% rule of thumb does not.
Assumptions and fact checks
Consumer sentiment is driven more by personal purchasing-power progress than by the indicators economists use to summarize national output and employment.
Why it mattersHouseholds experience budgets, wages, rents, and prices directly. GDP and unemployment matter, but they do not map one-for-one onto perceived living-standard gains.
People generally need about a 10% annual purchasing-power improvement to feel satisfied.
Why it mattersThe transcript supplies no evidence for that threshold, and a sustained 10% real annual gain would be an unusually high expectation. The broader argument does not need this number.
Jason wins because he describes the commercial interface Google eventually reported building. The flourish is oversized; the business-model intuition is excellent.