Long Journey's Arielle Zuckerberg on betting early on the weird — from robotic eyelash extensions to deep tech robotics

Summary

Arielle Zuckerberg, general partner at Long Journey Ventures, articulates a pre-seed and seed strategy built around backing ideas that look absurd before they become consensus. The firm's stated mission is to be "the second believer in the magically weird," a posture that has produced early positions in companies spanning cloud seeding, synthetic diamonds, and a robotic eyelash extension startup that cuts application time from roughly 90 minutes to 15 by running both eyes simultaneously.

Robotics and Deep Tech as the AI Second-Order Play

Long Journey's current conviction is that the most defensible opportunity in AI is not at the model layer but in the physical world. Zuckerberg argues that LLMs and physical AI have dramatically reduced the cost and time required to train robots, creating a tailwind for companies in robotics, biotech, and adjacent deep tech sectors. The firm has backed case-picking and distribution-center robotics alongside the eyelash extension play, and continues to evaluate in-home applications, though Zuckerberg is skeptical of general-purpose humanoid robots in the home, preferring ambient, single-purpose devices.

On AI model economics, she draws a distinction between the current enterprise pilot phase, where teams gravitate toward top-benchmark models regardless of cost, and an anticipated deployment phase where token costs will matter significantly. That shift, she argues, is what will unlock wide-scale rollouts of autonomous background agents and AI companions across large employee bases. The public's growing indifference to new model releases mirrors the plateau in consumer excitement around iPhone generations, but Zuckerberg contends that framing misses the compounding effect of continued cost reduction.

Portfolio Construction and Founder Dynamics

Long Journey does not typically take board seats, a deliberate choice that aligns with its tolerance for long, non-linear founder journeys. Zuckerberg acknowledges that early-stage bets on pre-consensus ideas carry a high whiff rate, and that founders usually recognize a dead end before the firm does. Crusoe is cited as the payoff case, where an early seed position preceded the company's significant scale-up.

Sourcing: Edge Institute and Non-Traditional Networks

For deal flow, Long Journey has partnered with Edge Institute, which runs monthly pop-up gatherings under the Edge Esmeralda banner. A month-long event in Patagonia is forthcoming; a prior iteration ran in Healdsburg in June 2025, where Long Journey ran a residency program. The gatherings draw researchers, unconventional thinkers, and people operating outside mainstream entrepreneurial networks. Edge Esmeralda is also in the planning stages of building a physical city, with a site already purchased, led by Devon Zugol.