E156: Ivy League antisemitism, macro, SaaS recovery, Gemini, Figma deal delay + big Friedberg update
2023-12-08 spoken.md · speaker-labeled ▶ watch ← E152 all episodes E157 →
Every number here comes from the conviction ledger — the same one the pool ranks on. Decay is what the 99 ideas nobody mentioned gave up this week; it applies only when an episode is processed.
Tier crossings
conviction thresholds crossed by this episode — 65 / 45 / 15 · ideas born here show where they landed. why 65 isn't always green
Kill dates that landed since E152
3 hit · 0 partial · 6 miss — windows that closed after 2023-11-03 and up to 2023-12-08, auto-scored against price data and never hand-set. result · return · vs SPY
| idea | result | return | vs SPY | scored |
|---|---|---|---|---|
| 🤖 Generative AI is the next Silicon Valley bubble cycle | RIGHT | +181.7% | +164.9 | 2023-12-03 |
| 🤖 Natural-language chat disrupts Google's search box | WRONG | -32.5% | -49.2 | 2023-12-03 |
| 🤖 SaaS gets replaced by MaaS — Models as a Service | WRONG | -51.9% | -68.6 | 2023-12-03 |
| 🤖 SaaS seat contraction — net revenue retention goes below 100% | WRONG | -94.1% | -110.8 | 2023-12-03 |
| 🪙 DAOs become a real capital-formation primitive | WRONG | -54.1% | -54.3 | 2023-11-20 |
| 📈 FTX's cap table is a who's-who of no diligence | WRONG | -27.1% | -43.1 | 2023-11-19 |
| 🤖 Only foundation models can disrupt search - and Google is too far ahead | RIGHT | +41.5% | +25.5 | 2023-11-19 |
| 🤖 Creator and gig income absorbs the labor force - participation stats understate earning power | RIGHT | +90.7% | +74.5 | 2023-11-05 |
| 🏛️ Split government ends the spending orgy and blocks the interventionist response | WRONG | -3.0% | -19.1 | 2023-11-05 |
Who moved the board
each voice's push on conviction this episode, in conviction points — supports and opposes · share = % of this episode's movement
What got argued (15 ideas)
ordered by how hard each idea moved · quotes are verbatim from the transcript, timestamps deep-link into the episode
The UK CMA is killing Adobe/Figma on a novel theory — not that the two compete today, but that Figma might compete with Photoshop some day — after 15 months of process. Sacks and Chamath argue that once a regulator can veto any deal on an unquantifiable future-competition test, and take a year and a half to say so, M&A activity gets a real chilling effect, the only reliable exit for startups narrows, and companies will deliberately avoid creating a nexus in the offending jurisdiction.
Sacks calls the end of the 18-month software recession off Jamin Ball's data: after four quarters of negative net new ARR, public SaaS printed positive net new ARR growth in Q3 2023, and his own board meetings are showing green shoots. Enterprise software revenue growth re-accelerates from here even though the valuation reset stands.
Chamath's dated rate-path call: rather than argue about whether the first cut lands in Q1 or Q2 2024, look two years out and you get to roughly 2.5% — about 160bp below the December-2023 level. The whole curve reprices lower, which he frames as the single biggest change coming for capital allocators.
If you believe that the 10-year is gonna go back down to, I don't know, this 2.5, 3% range, and if you believe that growth is re-accelerating, then I think there is room for this number, the 5.8 number, to at least grow into the long-term average, which is 7.8.
Episode digest
written during extraction and stored in data/extractions/ep156.json — the auditable source of truth, including everything market-adjacent that did not earn a capture
Rate-cut euphoria ran the macro block: Sacks called a Q1 2024 cut outright as a "Biden bailout" (attaching to Chamath's E120 forced-cuts thesis), Chamath went longer and dated at ~2.5% rates two years out (new idea — a 24-month claim can't be judged inside the E120 window), and both expect a cut to unlock the $5.7T money-market pile into equities. The year-long consumer-recession call died on air: Jason, an E110 supporter, flipped to "the economy is actually doing extraordinary" (logged as a reversal), Chamath waved Fed wealth data at it, and Sacks conceded the data is mixed-to-positive while insisting the electorate doesn't feel it. Sacks formally called the end of the 18-month software recession off the first positive net-new-ARR quarter (new idea); Chamath split it — revenues rebound, multiples don't, with PE clearing software at 1-3x ARR — which feeds the stale-marks and beta-trade-done theses, and Sacks opposed the latter by arguing 5.8x NTM has room back to the 7.8x long-run average. Gemini produced a clean debate: Sacks says AI-for-search caps Google's dominance forever (supporting the E110 erosion call), Jason took the other side on ad units inside AI answers, Friedberg made the YouTube-corpus data-moat case, and Chamath re-asserted commoditization — model costs to zero, Nvidia's CUDA garden broken, value in a cloud-infra-plus-apps barbell (he opposed this same open-model thesis at E129 before flipping at E131). The Adobe/Figma block yielded a new idea on speculative-competition antitrust chilling tech M&A, with Sacks, Chamath and Jason all on the same side. Friedberg's big update — full-time CEO of the stealth multiplex-gene-editing ag company inside Production Board — has no listed instrument, so nothing was coined for it.