The Twenty Minute VCAre Burn Multiples BS in an AI World? & Sam Altman Needs $1TRN of Energy
At a glance
WHAT IT’S REALLY ABOUT
AI growth warps metrics, funding dynamics, and trillion-dollar energy ambitions
- AI-native companies can look more capital-efficient by burn multiple despite worse cash flow margins because extreme growth swamps burn in the ratio.
- Burn multiples remain useful but are increasingly noisy because ARR quality, churn visibility, gross margin volatility, and CapEx (especially for AI) break prior SaaS assumptions.
- Founders with strong “classic” metrics can still struggle to raise in a winner-take-most, kingmaker-driven market where capital concentrates into a small number of perceived category leaders.
- Public-market comps and recent IPO performance (e.g., Figma, Klarna, StubHub) influence private valuations, with a risk that today’s generous baseline SaaS multiples could revert downward.
- Sam Altman’s stated ~$1T infrastructure/energy ambition may be partly “willed into existence,” but the panel expects economic constraints to force slower adoption or moderated CapEx plans even if AI leadership persists.
IDEAS WORTH REMEMBERING
5 ideasBurn multiple is not obsolete, but it’s no longer plug-and-play.
The metric still compares ARR added per dollar burned, but it assumes “real” ARR, durable retention, stable margins, and minimal CapEx—assumptions increasingly violated by AI pricing, trials, token costs, and infrastructure spend.
Hypergrowth can make heavy burners look “efficient” while still being fragile.
AI startups may post better burn multiples even with very negative cash flow margins because ARR is rising so fast, but absolute burn and cash runway still determine survival if funding windows close.
In 2025, fundraising is a haves-vs-have-nots market driven by breakout status, not just good metrics.
Founders can hit strong growth/efficiency heuristics and still be passed over if the TAM feels capped, the story isn’t perceived as category-leading, or investors believe the company is “years away” from an IPO-scale outcome.
If you can raise a decent round, take it—don’t optimize as if it’s 2021.
The panel argues many boards/VCs still give outdated “wait for better price” advice; for sub-scale companies, getting funded now and extending runway is often more rational than negotiating for marginal valuation gains.
“Non-AI” is becoming a narrative disadvantage because everyone is expected to have agents.
With most public software companies marketing themselves as AI companies, VCs increasingly assume agentic capability is table stakes, making differentiation and defensibility the real questions—not whether AI is present at all.
WORDS WORTH SAVING
5 quotesThere's only two ways of pricing a deal. You price a deal on hope or you price a deal on the multiple.
— Rory O’Driscoll
I hear too many folks that are like, "Oh, you're triple, triple, double, double or better, you're, you're golden. Don't worry, kids." And I think that's terrible advice in 2025. Terrible advice.
— Jason Lemkin
I think he is willing the trillion into existence.
— Jason Lemkin
Whatever the prize is for being the best company in AI, OpenAI's gonna get that prize, right?
— Rory O’Driscoll
A billion here, a billion there, and pretty soon you're talking real money.
— Rory O’Driscoll
High quality AI-generated summary created from speaker-labeled transcript.
