# Enigma raises $70M seed to study human-robot interaction through large-scale online experiment
The AMW Read
A novel approach to robotics AI that inverts the capability-first paradigm, introducing a player outside the established roster; significance is sub-segment because the company has no product or revenue.
# Enigma raises $70M seed to study human-robot interaction through large-scale online experiment
Enigma, a research lab emerging from stealth less than a year after founding, has raised a $70 million seed round led by Index Ventures and Ribbit Capital, with participation from Sarah Guo of Conviction Partners. The startup is launching a public online experiment allowing anyone to interact with over 100 proprietary robots housed in hangars in Israel and California, tasking them with painting, sword fighting, or performing chemistry experiments. Co-founders Jonathan Jacobi, formerly Microsoft's youngest employee, and Gal Niv met in Israel's Unit 8200 and assembled a team of math Olympiad winners and PhD dropouts from top AI labs — deliberately avoiding robotics insiders.
Why it matters: Enigma is pursuing an outside-in strategy that bypasses the dominant paradigm in robotics AI, which focuses on building ever-more-capable foundation models through simulation or video training. Instead, the startup treats human-robot interaction design as the primary vehicle for discovering both interface conventions and model training signals — an inversion of the usual capability-first approach. If successful, this could resolve an open debate about whether the bottleneck in physical AI is model intelligence or the quality of human-machine communication, and potentially unlock the kind of intuitive control that has eluded even the most advanced teleoperation systems.
The $70 million seed round for a pre-product startup with no disclosed revenue or enterprise use cases signals that top-tier venture capital is willing to back radical methodological bets in robotics. The bet rests on the premise that collecting real-world interaction data at scale, from untrained users, will reveal interface patterns that no amount of lab-based simulation can produce. Whether Enigma's open-ended experiment yields a commercially viable path — or produces interesting academic findings without a clear go-to-market — is the central uncertainty.
#Robotics #HumanRobotInteraction #SeedFunding #EmbodiedAI #IndexVentures #Startups
