ExpertsFM20VC: Why "Pacing the Frontier" is BS | Instinct Raising $1BN at $10BN & Meta Launches Muse | Miro Sells for $1.35BN After a $17.5BN Valuation | Mistral Raises €3BN & Could Sam Bankman-Fried Win His F
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch · September 17, 2026
5 Selected Lessons:
heuristic · 2 reasoning_pattern · 1 priority · 1 case · 1
Scarcity of viable late-stage buyers ('duck duck goose' is ending)
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why "Pacing the Frontier" is BS | Instinct Raising $1BN at $10BN & Meta Launches Muse | Miro Sells for $1.35BN After a $17.5BN Valuation | Mistral Raises €3BN & Could Sam Bankman-Fried Win His F (September 17, 2026) · at 44:23 ▸
When this applies
when a cohort of aging, maturing assets competes for a shrinking pool of willing acquirers
The takeaway
Rank assets within a declining category by relative quality, because once the limited number of buyer 'seats' are filled by the best assets, the remainder simply won't get bought — being 'better than the worst' matters more than being 'good enough' in absolute terms.
Why it works
A small number of buyers (strategics, PE firms like Bending Spoons) each do only a handful of deals a year despite screening thousands of targets, so once the 'best' remaining assets in a declining category are acquired, the rest have no credible buyer and are left to stagnate.
One level up
In a shrinking opportunity set, relative ranking against peers determines survival more than absolute merit.
Where it stops holding
This stops working when new buyer types (e.g., new PE funds, new strategics) enter the market and expand the number of available seats.
Detail
hierarchy:
- Companies better than Miro/Airtable (the current bar) may still find buyers
- Companies at or below that bar will see zero offers and drift into low-growth stagnation
- No buyer is obligated to rescue every asset in a declining category
rationale: Bending Spoons reportedly screens 1,000 targets and does only 5-10 deals a year; large PE (e.g., Thoma Bravo) is mostly sitting out, so very few 'chairs' remain for a whole generation of pre-AI-era SaaS unicorns.
In their words
If there is a seat left in DuckDuck Ooze, grab it, grab it. But otherwise, the game's just ended.
Risk-return profile decays as valuation climbs across funding rounds
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why "Pacing the Frontier" is BS | Instinct Raising $1BN at $10BN & Meta Launches Muse | Miro Sells for $1.35BN After a $17.5BN Valuation | Mistral Raises €3BN & Could Sam Bankman-Fried Win His F (September 17, 2026) · at 35:18 ▸
When this applies
when deciding whether to participate in a later round of a company whose valuation has compounded quickly
The takeaway
Re-evaluate the attractiveness of a bet at every new valuation checkpoint by asking what exit size is now required for an acceptable return, and compare that against the realistic population of exits at that size.
Why it works
As the entry valuation rises, the multiple needed on exit to achieve the same return shrinks the set of plausible outcomes that satisfy it, since very few exits exceed any given high threshold (e.g., tens of billions), progressively worsening the risk-adjusted attractiveness of each successive round.
One level up
A bet's merit is not fixed — it must be recalculated continuously against current price, not judged by its original thesis.
Where it stops holding
This stops working when the asset's fundamentals (revenue, differentiation) are also compounding fast enough to justify the escalating valuation, not just hype.
Detail
approach: At each funding round, recompute the exit multiple required for an attractive return and compare it against the realistic distribution of exit outcomes at that scale.
signals:
- Valuation 20x'ing in under five months
- Comparison to rare precedent (largest-ever M&A outcome, Cursor at $60B)
rationale: The risk-return profile was wildly attractive at $50M pre, good at $500M pre, but by $10B you need an outcome in the rarest tier of tech exits ever to see positive returns.
In their words
The risk return profile was wildly attractive in April and decreasingly unattractive fairly quickly by the time you come to September. So that is the problem with those kind of bets.
Dual-use technology guardrail problem
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why "Pacing the Frontier" is BS | Instinct Raising $1BN at $10BN & Meta Launches Muse | Miro Sells for $1.35BN After a $17.5BN Valuation | Mistral Raises €3BN & Could Sam Bankman-Fried Win His F (September 17, 2026) · at 16:38 ▸
When this applies
when assessing risk of any technology that grants broad capability with no built-in restriction on its application
The takeaway
Assume any sufficiently capable general-purpose tool has a corresponding harmful application of equal strength, and design guardrails accordingly rather than assuming good intent will dominate.
Why it works
Capability is typically symmetric: the same skill that enables a positive application (coding, drug design, structural engineering) equally enables its harmful mirror image, so without active guardrails the negative use is just as achievable as the positive one.
One level up
Power is directionless; restraint must be engineered in deliberately, not assumed.
Where it stops holding
This stops working when a capability is narrow enough or so costly to misuse that the harmful mirror application isn't practically accessible.
Detail
rule: Most positive AI use cases have a corresponding dark version of equal capability.
threshold: Applies wherever open-weight or unrestricted models exist with no effective guardrails.
rationale: Superhuman coding implies superhuman hacking; superhuman drug design implies superhuman poison design; the skill is value-neutral.
In their words
Most positive use cases for AI have a corresponding dark version. If you're super human at coding, you're super human at hacking. If you're super human at structural engineering, you're likely super human at finding structural flaws to knock buildings down.
Avoid bets that require an M&A outcome to succeed
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why "Pacing the Frontier" is BS | Instinct Raising $1BN at $10BN & Meta Launches Muse | Miro Sells for $1.35BN After a $17.5BN Valuation | Mistral Raises €3BN & Could Sam Bankman-Fried Win His F (September 17, 2026) · at 32:53 ▸ · at 35:18 ▸
When this applies
when an investment's only plausible path to a positive return is a single, unpredictable high-value exit event rather than organic cash flow or IPO
The takeaway
Treat any investment whose return depends entirely on an unpredictable M&A outcome as a high-variance speculative bet, size it accordingly within a portfolio, and don't treat it as a base-rate, fundamentals-driven investment.
Why it works
Acquisition decisions are driven by idiosyncratic, capricious factors (a single executive's whim, timing, relationships) rather than predictable fundamentals, so betting on one happening is a much higher-variance wager than betting on organic growth.
One level up
Separate bets with a controllable, analyzable path to return from bets dependent on externally capricious events.
Where it stops holding
This stops working when the asset is so strategically scarce or the team so exceptional that even a small number of 'M&A-outcome-only' bets can carry positive expected value for a portfolio that can tolerate variance.
Detail
rule: Don't take bets that 100% require an M&A outcome to be successful.
threshold: At most a small handful (e.g., three) of such high-variance, M&A-dependent bets in a ~30-company portfolio.
rationale: M&A outcomes are capricious and unpredictable (e.g., a key champion can quit the next day), unlike fundamentals-driven growth.
In their words
The way I was raised to invest in venture was don't take bets that 100% require an MA outcome to be successful, right? They're just too unpredictable.
Matt Mullenweg's bylaw-based board countercoup at Automattic
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why "Pacing the Frontier" is BS | Instinct Raising $1BN at $10BN & Meta Launches Muse | Miro Sells for $1.35BN After a $17.5BN Valuation | Mistral Raises €3BN & Could Sam Bankman-Fried Win His F (September 17, 2026) · at 51:20 ▸ · at 54:35 ▸
When this applies
when a founder with retained structural control (voting power, bylaw leverage) is confronted by a board trying to oust them
The takeaway
Formal governance rights (who can vote whom out) ultimately determine power struggles regardless of the merits of the underlying business dispute — so know exactly who holds structural control before a boardroom conflict escalates.
Why it works
If a founder retains sufficient voting control (via share class, bylaws, or founder agreements) they can use that same governance mechanism the board used against them to instead replace the board itself, nullifying an ouster attempt regardless of the board's substantive concerns.
One level up
In any power struggle, whoever controls the formal mechanism of removal wins, independent of who is 'right.'
Where it stops holding
This stops working when the founder lacks sufficient voting control or when outside shareholders/investors have contractual protections that override founder voting power.
Detail
situation: An independent board majority at Automattic sought to remove founder-CEO Matt Mullenweg over disputes including the WP Engine conflict and strategic drift.
decision: Mullenweg used his retained voting/bylaw power to replace the board that had voted to replace him, rather than accept the ouster.
outcome: The independent board members resigned rather than engage in prolonged litigation over a shrinking, low-stakes prize; Mullenweg retained CEO control of a company both hosts view as strategically declining.
lesson: Structural voting control can override a board's substantive judgment entirely, turning a governance conflict into a formal power contest rather than a merits-based resolution.
In their words
My guess is the founder, CEO went deep into the bylaws and says, you are the board and you can vote to replace me, but I can actually also vote to replace the board. I'm hereby voting. I've seen this once. I'm hereby voting to replace the board.