The Twenty Minute VCInstinct Raising $1B at $10B & Meta Launches Muse | Miro Sells for $1.36B After a $17.5B Valuation
CHAPTERS
- 0:00 – 6:03
“Pacing the frontier”: extinction fears vs practical AI regulation
The group reacts to Dario Amodei’s call to “pace the frontier,” pushing back on doomsday framing while separating real, actionable risks from hype. They debate what governments would do if the risk were truly existential and why proposed monitoring/regulatory solutions are hard to implement globally.
- •Distinguishing Dario’s stated risks (cyber/economic/control) from “10% p(doom)” rhetoric
- •Argument that truly existential threats would trigger immediate state intervention, not voluntary guidelines
- •Skepticism about feasibility of multinational AI regulation (democracies + China/Russia alignment)
- •Concern that third‑party monitoring could become heavy-handed, innovation-suppressing regulation
- 6:03 – 10:24
Anthropic’s IPO risk factor theory & the coming political/media backlash
Jason frames the controversy as essentially an S‑1 risk factor being “written live” ahead of an eventual Anthropic IPO. The panel predicts intense public and congressional scrutiny now that AI safety talk has gone mainstream.
- •“Risk factor in an S‑1 done live”: pre-empting investor concerns ahead of an IPO
- •Why safety statements can trigger investigations, hearings, and potential shutdown risk
- •Public perception shift: friends/family asking if AI will kill humanity
- •Political reaction dynamics: two years of hearings and amplified debate likely
- 10:24 – 15:53
Loss-of-control vs cyber and jobs: which AI risks actually matter?
They rank Dario’s cited risks, dismissing some and emphasizing the hardest problem: losing control via recursive self-improvement/agentic behavior. The discussion lands on CEO responsibility—if it can’t be controlled, the builder should stop.
- •Cyber risk viewed as real and immediate; ‘jobs apocalypse’ framed as historically repetitive tech fear
- •Loss-of-control/recursive self-improvement treated as the genuinely hard-to-assess risk
- •CEO accountability: product liability framing—don’t build what you can’t control
- •Technology diffusion: once capabilities exist, the “cat’s out of the bag” globally
- 15:53 – 20:56
The “dark version” of every AI capability & why markets won’t self-limit
Jason and Rory argue that every beneficial AI capability has an equally potent malicious counterpart, and open-weight models make weaponization easier. They note that despite moral hand-wringing, incentives and capital flows keep acceleration high.
- •Dual-use reality: coding → hacking, drug design → poisons, engineering → sabotage
- •Guardrails won’t hold once dark variants propagate via open distribution
- •Markets barely react to existential-risk talk; capital keeps funding acceleration
- •Biggest danger may come from bad actors/sociopathic founders, not top labs
- 20:56 – 29:10
Meta launches Muse: a serious entrant in the AI assistant race
They dissect Meta’s Muse as a product: fast, polished, and operationally impressive, backed by Meta’s infrastructure and model advantages. The key question becomes whether a standalone assistant has a “killer app” or wins via many small wins.
- •Muse reviewed as ‘great software’ with strong end-to-end task execution
- •Meta’s infra advantage: subsidized VM/compute/storage + in-house model lowers costs
- •Debate: need for a Visicalc-like killer app vs “incrementally better at everything”
- •Standalone assistants vs embedding assistants into existing social/commerce ecosystems
- 29:10 – 35:05
Would you invest in Instinct at $10B? The IC debate
Harry tees up Instinct raising $1B at a $10B valuation, prompting a mock partnership discussion. Jason argues both sides—praising team quality but ultimately rejecting the round due to cost structure and incumbent pressure.
- •Instinct’s rapid valuation step-ups (50 → 400 → 2.5B → ~10B) and what that implies
- •Compute/latency as a signal of high unit costs; fundraising to subsidize usage
- •Meta as the archetypal feared incumbent finally building a real consumer agent app
- •Rule of thumb: avoid deals that require M&A as the only viable exit path
- 35:05 – 41:43
Meta as the biggest threat & the ‘$50B acquisition’ upside scenario
They explore how Instinct could still win via differentiated traction and strategic acquisition, even if IPO economics are tough. The panel debates whether modern venture math is shifting toward expecting $25B+ outcomes more routinely.
- •Acquisition path: Poolside/Nvidia analogy—big players buy talent/models/traction
- •Strategic ties and talent narrative (founder reputation, OpenAI board relationships)
- •Menlo’s framework: aiming for many $25B+ exits; how that changes acceptable entry prices
- •Risk-return compression: early rounds were attractive; later rounds demand extraordinary outcomes
- 41:43 – 45:31
Miro sells for ~$1.35B after $17.5B peak: inevitability of repricing
The conversation shifts to Bending Spoons buying Miro at a fraction of its 2021 valuation. They frame it as an inevitable clearing event for a stale, overvalued asset—and note that late-stage structures can still produce 1x outcomes.
- •Why Miro’s $17B 2021 valuation became the ‘stale sore thumb’ among unicorns
- •Exit seen as ‘not bad’ operationally: founders/employees okay; late-stage investors ~1x
- •Venture vs single-owner operators: syndicates often can’t execute ruthless efficiency transitions
- •Some consideration rolled into Bending Spoons equity—signal of belief in their operating model
- 45:31 – 48:13
The Bending Spoons playbook: raise prices, cut costs, accept churn
They unpack how Bending Spoons creates value post-acquisition through aggressive pricing and cost discipline, not revitalization. The tradeoff is steep churn, but higher revenue per retained customer and strong cash flow.
- •Observed pattern: brutal churn following price hikes; focus on cash-flow extraction
- •Finding the ‘marginal propensity to pay’ by shedding price-sensitive customers
- •Operational philosophy: less reverence for ‘founder’ mythology; performance-now mindset
- •Customer implication: expect large price increases after acquisition
- 48:13 – 51:11
The era of SaaS capitulation: ‘duck duck goose’ and the last chairs
Jason argues many pre‑AI SaaS unicorns are running out of buyers, with PE and strategics less active. The result is a capitulation era where many companies settle into low-growth ‘value mode’ with limited exit optionality.
- •“Only one or two chairs left”: few acquirers with appetite/balance sheet for big SaaS deals
- •Signals: single-offer processes at ~2–3x revenue imply weak competitive tension
- •Forecast: many assets shift to 0–5% growth, potentially never sold
- •Portfolio implication: if a seat exists, grab it—otherwise prepare for independence
- 51:11 – 1:06:23
SBF appeal aside & Automattic/WordPress board drama: founder power vs relevance
A brief detour covers SBF’s Supreme Court angle, then moves into Automattic’s governance conflict and WordPress’s strategic drift. The panel suggests the real issue is product relevance in an AI-native world, not just bylaws and board fights.
- •SBF: debate on proportionality of punishment vs legal/constitutional issues around fines
- •Automattic: board vs founder dynamic; bylaws can allow founder to reset the board
- •WordPress/Automattic viewed as trailing the tech frontier; urgency to pivot toward AI tools
- •Open-source tension: ecosystem monetization conflicts (WP Engine hosting vs platform stewardship)
- 1:06:23 – 1:10:32
Mistral raises €3B: Europe’s AI sovereignty bet (not a frontier race)
They argue Mistral’s mega-round is primarily geopolitical—ensuring Europe has a sovereign AI stack—rather than a sign it can outcompete OpenAI/Anthropic on frontier models. The funders and valuation are seen as strategic, with an asterisk.
- •Core thesis: sovereignty and dependency risk after US access restrictions to advanced models
- •Airbus analogy: Europe funds an inefficient but strategic counterweight until viable
- •Round led by strategic capital (e.g., Samsung; prior ASML), complicating valuation signal
- •Possible outcome: not #1 globally, but durable European ‘champion’ with political tailwinds
- 1:10:32 – 1:19:28
Adobe’s new CEO & when growth companies become value stocks
They treat Adobe’s CEO change as a signal of continuity rather than transformation, noting strong cash flow but limited growth narrative. The discussion broadens into valuation regime shifts: once growth slows, markets price on EBITDA/cash flow.
- •CEO succession framed as ‘nothing changes’; Adobe survives on scale and margins
- •“If it ain’t growing, you ain’t winning”: AI metrics vs true net-new ARR growth
- •Valuation transition: >30% growth supports revenue multiples; slower growth forces EBITDA multiples
- •Canva vs Adobe comparison: deceleration risk and the difficulty of crossing the growth-to-value chasm
