Moonshot capitalism: AI rewrites the venture capital playbook
The AI boom is fuelling a resurgence in ambitious “moonshot” bets, as SpaceX-style returns and falling valuations for traditional software push VCs toward riskier, more capital-intensive deep tech. Excluding giant AI-lab cheques, global deep-tech investment has exceeded $150bn since the start of 2024 — more than the $133bn in the entire decade to end-2019, according to Dealroom. Accel’s Matt Robinson says the companies VCs discuss have “transformed”; Plural’s Carina Namih argues AI simulation is cutting capital intensity for fusion and space. Caveats are explicit: this year’s deep-tech totals have not yet surpassed 2021’s battery/EV peak (Rivian, Northvolt and peers that later soured); commercialisation remains hard even when fundraising gets easier; Alphabet’s X is more grounded on business case after hits (Waymo ~$126bn) and misses (Loon). SpaceX IPO wealth is cited as FOMO fuel (Founders Fund’s ~$600mn stake marked >$50bn at IPO per PitchBook). SaaSpocalypse and Anthropic-era model competition are framed as destroying software “moats,” with General Catalyst’s Hemant Taneja arguing founders must build far bigger companies for outcomes to matter.
Why it matters
Canonical Dealroom-cited datapoint on the post-SaaS VC rotation into deep tech, with named funds (Accel, Plural, Lightspeed, Founders Fund, General Catalyst), SpaceX-exit FOMO, and the tension between easier moonshot fundraising and still-hard commercialisation — directly maps European and global deep-tech opportunity sets Dealroom tracks.
Executive takeaways
- Dealroom: deep tech (ex mega AI labs) >$150bn since start-2024 vs $133bn in the decade to end-2019.
- Driver mix: AI infra creating customers for wild hardware; AI ‘in silico’ cutting experiment cost; SaaS moat destruction / SaaSpocalypse.
- Named space/fusion deal flow: Sierra Space, Axiom, Iceye; Helion, Proxima, Inertia; Star Catcher $65mn (May) for orbital power beaming.
- Counterweights: 2021 peak not yet beaten; Northvolt/Rivian-style bust risk; X more business-case disciplined; no traditional valuation metric for space/biology labs.
What Financial Times may be missing
Dealroom methodology for the $150bn deep-tech total (sector definition, geography, double-counting with AI-adjacent infra) is not unpacked. Limited European founder/LP primary voices beyond Plural; little on defence-tech or EU industrial-policy capital as parallel moonshot channels. Exit paths beyond SpaceX IPO anecdote remain thin.
