Many AI founders now find it necessary to raise at valuations requiring absolute domination of the field; Brex, once valued at $12B, shows the downsides of this
Capital One just announced it's acquiring Brex for $5.15 billion. An incredible, top .1% “exit” in less than 10 years …
The transaction matters because a company can achieve a substantial exit while still landing well below its last private valuation, making the gap between fundraising narratives and realizable outcomes harder to ignore.
First-order effects
Capital One is acquiring Brex for $5.15B, setting a realized outcome below Brex’s previously reported $12B valuation and resetting the reference point for its shareholders.
For AI founders and investors, the deal supplies a visible warning that a high private valuation can narrow the range of later exits that count as successes relative to the last round.
Second-order effects
Investors may apply more scrutiny to whether AI companies’ revenue, margins, and market position can support the valuation implied by new rounds, rather than treating follow-on financing as validation.
Founders raising at aggressive prices face greater pressure to preserve optionality—through pricing, deal terms, and capital needs—because a merely strong acquisition may not clear the prior valuation.
Third-order effects
If similar outcomes recur, private-market financing may shift from headline valuation growth toward structures and prices that leave room for strategic acquisitions and public-market outcomes.
The broader divide may widen between a small set of companies able to substantiate dominance and a larger group whose financing expectations outrun the exits available to them.
The trend: AI’s capital boom is increasing the importance of valuation discipline, as private prices increasingly require exceptional scale or strategic value to be realized at exit.
HUBRISTIC FUNDRAISING “Brex raised $300 million at a $12.3 billion valuation in October 2022—technically 2022, but really the tail end of the 2021 madness before rates spiked and everything crashed. At that valuation, with maybe $200-250M in revenue at the time, Brex was being [i…
Thoughtful piece... but for most VCs the incentives encourage hubristic fundraising. This isn't really a founder problem... it's a VC problem. Also a few ‘modern day AI’ hubristic comps do not deserve to be in the list 👀
@jasonlk I think this is very smart / very much on-point. It can be very dangerous not to play the game on the field though - if your competitors are doing this and you are not, it can be very hard to survive.
Fantastic and balanced piece. It's less about right vs wrong and more about understanding tradeoffs and the rules of the game. It's always a lot more nuance than the average, simplistic twitter/hn take. PS: If you ever get the chance, take Jason's check.
Epic post by @jasonlk Know the trade off If you roll the dice and stumble you won't capture all the upside But obviously still an epic outcome for founders, early employees and early investors With AI companies unlike with Brex unit economics will be another big variable.
I've been in the startup game for a minute, there is a very wide delta between private market valuation & public valuations , especially at this moment. Worth a read if you are a founder who is raising.
@jasonlk My observation is that secondaries have really altered the dynamic Founders can understand that fundraises should be organized for monotonic growth with great investors to get to a great final/max destination It's hard (when they're selling along the way) to take a lower…
My only qualm with this article is the claim that, say, Loveable is “worth” more than Brex. Brex sold for $5.15B. Loveable is “worth” $6.6B. If early-stage valuations are hype; late-stage valuations are fake.
This is quite well written; important reflection on the current environment. I would add one more thing. Most of these rounds are preemptive. The investor suggested it - it wasn't the company's initiative. A finance form of the observer effect. https://grokipedia.com/...
I have no idea what bilt modeled, but once they take away the unsustainable giveaway, all those people are going to churn off. They had a deal that was too incredible to be true. That they couldn't buy the back book is even more damning. They have to reacquire those customers.
My approach to real arr projections: if you are selling primarily to startups, you should discount by likelihood they will still be around in out years. Hint: they won't. And the degree to which you are giving away money. If your product sells for below your token costs, of