Answer: Matan Grinberg argues that the accrual of value in tech is increasingly time‑dependent, the United States lacks cutting‑edge open‑source AI models, and outsourcing AI development can dramatically improve efficiency.
On a Tuesday morning in San Francisco, venture capitalist Matan Grinberg walked into a packed 20VC panel and told the audience that a startup’s valuation now hinges on how quickly it can turn an algorithm into a revenue‑generating product. “If you take a year to ship, you’ve already lost most of the upside,” he said, pointing to a slide that showed the median time‑to‑market for AI‑enabled services dropping from 18 months in 2020 to under eight months in 2024.
Grinberg’s blunt assessment hits a nerve for founders wrestling with the same calendar. The economy and markets sector feels the pressure: Wall Street analysts now cut price‑to‑sales multiples for late‑stage AI firms by 15% if a product launch exceeds 12 months.
Why does this matter?
Speed matters because capital allocation in tech is a race against obsolescence. When a model becomes publicly available—or is replicated by a competitor—the window to capture premium pricing evaporates. Grinberg highlighted that the U.S. “doesn’t have frontier open‑source models” comparable to Europe’s EU‑AI Lab, leaving domestic startups forced to license expensive proprietary stacks.
Can outsourcing AI development close the gap?
Grinberg proposes a pragmatic fix: outsource parts of the AI pipeline to specialist firms in regions with deep talent pools and lower overhead. He cited a recent partnership where a California‑based fintech shaved six weeks off its fraud‑detection rollout by hiring a Ukrainian AI boutique to fine‑tune a transformer model.
The numbers are compelling. The outsourced team delivered 1.2 billion inference calls per month at 30% less cost than the in‑house effort, according to internal metrics shared with 20VC.
Critics, however, warn that off‑shoring could expose sensitive data and dilute IP ownership. Grinberg countered that robust contractual safeguards and encrypted data pipelines can mitigate most risks.
What happens next?
Investors are already recalibrating. Several U.S. venture funds announced a shift toward “speed‑first” capital structures, allocating early‑stage bridge rounds contingent on meeting predefined deployment milestones.
For entrepreneurs, the message is clear: accelerate or risk being left behind. Grinberg’s takeaways suggest a two‑pronged strategy—double‑down on rapid prototyping and consider trusted offshore partners to keep pace with the accelerating AI frontier.
Stay tuned as the tech ecosystem tests whether outsourcing can become a standard shortcut, or if regulatory scrutiny and data‑privacy concerns will curtail its rise.
Meta description: Matan Grinberg warns that tech value is time‑dependent, the US lacks frontier open models, and outsourcing AI development boosts speed and efficiency.