Featuring
The original quartet tours the Trump fundraiser, the COVID-origin investigation, Bitcoin's halving, Salesforce's face-plant, and one breaking guilty verdict. The courtroom recap supplies the habanero: Jason argues that a political prosecution can still charge a real crime, while Sacks calls the theory a pretzel in a suit. Jason has the best episode overall—his Biden switcheroo forecast lands almost on the calendar—even if his legal fact pattern needs a red pen.
Spice rack
Was Trump's New York conviction a valid prosecution or politically motivated lawfare?
Original point: The guilty verdict rested on real business-record and election-law concepts, even though the prosecution was politically motivated and the conduct was minor.
What everyone argued
Jason Calacanis
Jason argued that Trump did falsify business records to protect his election prospects and that the jury applied recognizable legal concepts. He simultaneously called the conduct 'ticky-tacky,' the prosecution politically motivated, and the likely consequence a speeding ticket rather than imprisonment.
David Sacks
Sacks called the case a sham built from an expired misdemeanor, a novel felony theory, and a politically driven prosecutor. He stressed that jurors could choose among unlawful means without agreeing on one and predicted that the conviction-as-campaign-message strategy would fail.
Winner circle
Jason wins narrowly on the central legal question because he recognized that an aggressive and politically useful prosecution can still rest on valid statutory elements and evidence. Sacks raised the better appellate objection and was plainly right about the conviction's limited political force, but he did not carry the heavier claim that the case was a sham with no real crime underneath it. Jason's factual errors keep the ruling at medium confidence, especially while the direct appeal remains unresolved.
Commentary
Jason Calacanis
Assumptions and fact checks
The prosecution was politically motivated even though its legal theory was valid.
Why it mattersPolitical incentives are plausible, but motive is difficult to prove from timing and rhetoric alone. The trial court had rejected the improper-purpose argument before trial.
The practical consequence would be closer to a speeding ticket than imprisonment.
Why it mattersTrump received an unconditional discharge: the conviction remained, but the sentence imposed no jail, probation, or fine.
Trump was convicted on 34 counts of falsifying business records in the first degree.
CheckThe New York court's case record and post-trial orders confirm the jury returned guilty verdicts on all 34 charged counts.
Allen Weisselberg outlined the payment terms for the jury at this trial.
CheckWeisselberg did not testify in the criminal trial. Other witnesses, including Michael Cohen, Jeffrey McConney, and Deborah Tarasoff, supplied evidence about the payment arrangement and records.
The jury convicted Trump of a separate election-interference crime that served as the second crime.
CheckThe only charged offenses were 34 first-degree falsifying-business-records counts. Intent to commit, aid, or conceal another crime was an element; the instructions identified conspiracy to promote an election by unlawful means and did not require a separate conviction on that object offense.
David Sacks
Sacks was strongest on the exact unanimity issue and on political hindsight. He lost ground by turning a real appellate argument into certainty that no valid crime existed, then seasoning the legal analysis with motive claims he could not prove.
Assumptions and fact checks
An aggressive or novel prosecution of a former president is necessarily a sham.
Why it mattersNovelty raises the burden for clarity and restraint, but it does not erase documentary evidence or statutory elements. The better argument attacks specific instructions and proof rather than treating novelty as dispositive.
Bragg brought the case chiefly to create the phrase 'convicted felon' for the campaign.
Why it mattersThe political benefit was obvious, but the episode offered inference rather than evidence of the prosecutor's controlling motive.
The jurors did not have to agree unanimously on which unlawful means supported the election conspiracy.
CheckThe judge instructed jurors that they had to conclude unanimously that unlawful means were used, but did not have to agree on which of the three described unlawful means was used.
The case was legally barred because the misdemeanor statute of limitations had expired.
CheckTrump was charged with first-degree felonies, not time-barred misdemeanors, and the trial court rejected the limitations and object-offense challenges. Whether the felony theory survives direct appeal was still a live question, but the categorical claim that the filed case was already barred was wrong.
Using the conviction as the Democrats' central campaign answer would not work.
CheckTrump won the 2024 election with 312 electoral votes to Kamala Harris's 226. That result does not isolate the verdict's causal effect, but it decisively confirms that the conviction did not block his return to office.
Would Democrats replace Joe Biden after the first 2024 debate?
Original point: Democrats would replace Biden within 30 to 60 days after an early debate exposed his weakness.
What everyone argued
Jason Calacanis
Jason predicted with '100%' confidence that Democrats would perform a switcheroo after Biden struggled in an early debate. He expected party leaders to give Biden a graceful exit and nominate someone more capable of campaigning against Trump.
David Sacks
Sacks argued Democrats had made their bed through the primaries and policy record, so it was too late to substitute a new candidate. Even if they tried, he said voters would not forget the Biden administration's record five months before Election Day.
Winner circle
Jason wins decisively on the exact question. He predicted the trigger, the mechanism, and a 30-to-60-day window, and Biden withdrew 51 days after the episode. Sacks gets the important footnote that replacement would not erase the record or guarantee victory, but his claim that it was too late did not survive the summer.
Commentary
Jason Calacanis
This was Jason's cleanest call of the episode: mechanism, timing, and outcome all landed. He should have stopped at the switcheroo; the landslide add-on turned a sharp forecast into a bundled overclaim.
Assumptions and fact checks
Party leaders could still persuade Biden to release his delegates before the convention.
Why it mattersThat was procedurally and politically possible, and the debate created the coordinated pressure needed to make it happen.
A more agile nominee would erase the administration's liabilities and beat Trump easily.
Why it mattersChanging the candidate improved presentation but did not remove inflation, immigration, foreign-policy, or incumbency burdens. The final result disproved the landslide claim.
Democrats would replace Biden within 30 to 60 days of the episode.
CheckBiden announced his withdrawal on July 21, 2024, 51 days after the May 31 episode, and endorsed Kamala Harris.
The early debate would be the trigger for reassessing Biden.
CheckThe first debate occurred June 27, and Biden's performance immediately triggered the pressure campaign that ended with his withdrawal.
A replacement Democratic candidate would win in a landslide.
CheckHarris became the nominee and lost the Electoral College 312-226 to Trump.
David Sacks
Sacks correctly saw that candidate substitution was not policy amnesia. But that was adjacent to Jason's actual prediction. On whether the party could and would replace Biden, 'way too late' was simply wrong.
Assumptions and fact checks
Primary victories made Biden effectively irreplaceable before the convention.
Why it mattersThe delegates were pledged, but Biden could withdraw and party actors could coordinate around a successor—as they did.
A replacement nominee would inherit the administration's policy record.
Why it mattersHarris could change tone and emphasis, but as vice president she could not plausibly campaign as a clean break from the incumbent administration.
It was too late for Democrats to replace Biden with another nominee.
CheckBiden withdrew before the convention, released the path to his delegates, and Harris became the Democratic nominee.
A new nominee would not necessarily prevent a Trump victory.
CheckHarris replaced Biden and Trump nevertheless won 312 electoral votes.
Was Salesforce facing structural AI disruption or a temporary valuation and macro shock?
Original point: Salesforce's slowdown could reflect macro weakness, cheaper AI-built alternatives, or customers' unwillingness to pay separately for generative AI.
What everyone argued
Chamath Palihapitiya
Chamath called Salesforce a legacy product on the wrong side of a replacement cycle. He argued new companies could deliver 80% of its features at a 90% discount, forcing business-model recycling and layoffs over five to ten years.
David Sacks
Sacks viewed the selloff as a probable buying opportunity. He argued Marc Benioff had repeatedly adapted Salesforce to major platform shifts, was unconcerned about the per-seat model, and thought the soft forecast and broader economic slowdown mattered more than imminent product extinction.
David Friedberg
Friedberg framed three competing mechanisms: macroeconomic slowing, AI-enabled software commoditization, and weak willingness to pay for generative-AI add-ons. He later joined Sacks in refusing to count out Benioff and stressed founder-led companies' ability to maneuver.
Winner circle
Sacks and Friedberg win. Sacks correctly bet on Salesforce's capacity to adapt, while Friedberg supplied the better causal model: macro pressure, product commoditization, and willingness to pay had to be separated. Chamath earns partial credit because Salesforce did broaden beyond seat pricing, but his 80/90 replacement claim ignored the integrations, data, governance, and switching costs that make enterprise software sticky.
Commentary
Chamath Palihapitiya
Chamath spotted the pressure but used startup build cost as a stand-in for enterprise replacement cost. The strongest version of his case was pricing-model change, and Salesforce later validated that narrower claim; the sweeping rip-and-replace thesis did not land by 2026.
Assumptions and fact checks
Matching 80% of visible features at 10% of the price is enough to replace Salesforce for large enterprises.
Why it mattersEnterprise switching decisions also price data migration, integrations, compliance, reliability, retraining, procurement, and vendor risk. Feature parity is not system replacement.
AI would force Salesforce to change how it prices and packages value.
Why it mattersSalesforce later added action-, conversation-, and credit-based Agentforce pricing alongside user licenses.
The May 2024 drop revealed a settled structural verdict against legacy SaaS.
Why it mattersLater growth and AI monetization show the selloff was not a clean referendum on inevitable displacement.
David Sacks
Sacks got the company-level call right: distribution, data, and a founder-led product machine mattered more than a demo-level feature comparison. His blind spot was pricing architecture, where Salesforce itself later adopted the consumption logic he waved away.
Assumptions and fact checks
Founder-led adaptation would matter more than the installed base's vulnerability to cheaper AI tools.
Why it mattersThe installed base, data layer, distribution, and ability to bundle Agentforce gave Salesforce defenses that a feature-cost comparison missed.
The per-seat model would not face meaningful disruption.
Why it mattersSalesforce retained user licenses but also introduced Flex Credits, per-action, and per-conversation pricing. AI changed the meter even without killing seats.
Salesforce's reported Q1 FY2025 revenue was about $9.13 billion and grew 11% year over year.
CheckSalesforce's official release reported $9.13 billion of quarterly revenue, up 11% year over year.
Salesforce could adapt to AI rather than miss the platform shift entirely.
CheckBy FY2026, Salesforce reported more than $2.9 billion of Agentforce and Data 360 ARR, and Q1 FY2027 total revenue rose to $11.13 billion. Company-reported metrics do not prove every product claim, but they disprove near-term failure to participate.
David Friedberg
Friedberg argued best by refusing a false choice. His three-part frame captured the different risks, then his Benioff point supplied a mechanism for adaptation rather than mere founder worship.
Assumptions and fact checks
The earnings shock could not identify whether macro demand or AI commoditization was the dominant cause.
Why it mattersA single quarter's guidance embeds customer budgets, valuation, execution, and product transitions; the causal ambiguity was real.
Founder leadership materially improved Salesforce's odds of adapting.
Why it mattersBenioff could redirect product, packaging, acquisitions, and sales distribution, which is exactly the response a structural transition required.
First-quarter 2024 real U.S. GDP growth had been revised to about 1.3% at an annual rate.
CheckThe BEA's May 30 second estimate put annualized real GDP growth at 1.3% for Q1 2024.
Customers might resist paying for generative AI as a conventional add-on.
CheckSalesforce ultimately offered several value meters—per action, per conversation, and per user—rather than relying only on a traditional seat add-on. That supports the pricing concern, though not a claim that customers rejected AI outright.
Jason had the right framework—separate guilt, motive, and proportionality—but weakened it by claiming testimony that never happened and flattening the object-offense instructions into a separate conviction. Precision mattered most in exactly this debate.