Carbon Robotics CEO on laser weeders, Starlink-enabled autonomous tractors, and feeding America without chemicals

Summary

Carbon Robotics CEO Paul Mikesell argues that tech investors in agriculture have made their biggest mistakes by trying to reinvent the sector rather than improve it. Vertical farming, which he estimates destroyed tens of billions of dollars across multiple companies, exemplifies the problem. The mistake was trying to closely manage every input in an artificial environment when natural conditions already work more efficiently. Farmers run hard, competitive businesses with deep expertise in soil health and growing conditions that doesn't transfer easily to a Stanford CS graduate with a data platform.

Starlink as infrastructure

Satellite-based internet has become one of the most underappreciated changes in agriculture. Carbon Robotics built remote observation and command-and-control capability into its LaserWeeder machines to handle exception cases, but connectivity was the bottleneck. That problem is now largely solved, and Mikesell sees the broader agricultural automation wave following from it.

Fundraising sources

Carbon Robotics raised from Bond Capital (Mary Meeker), Sozo, Ignition, Fuze, Voyager, and Anthos, with notably no ag-tech VCs in the group. Mikesell says traditional ag investors kept telling the team that lasers were a stupid idea and chemicals were the obvious solution. It took investors accustomed to flipping industries upside down to see what the company was actually building. Hardware fundraising itself proved harder regardless, given the capital intensity, inventory burden, and R&D spend that doesn't always survive prototyping.

Farm management software

Carbon Robotics runs its own data layer, the Carbon Operations Center, which tracks weed counts, weed types, crop spacing, and plant size across every LaserWeeder pass. John Deere operates a competing command center. When Mikesell asks farmers which platform they want him to integrate with, he gets no clear answer. No dominant farm management platform exists, and he is skeptical one ever will if it is built by outsiders arriving with data but no actionable output. The platform that wins will be mobile-first, because farmers are in their fields, not at desks.

Food security

Mikesell frames a larger concern outside the product roadmap. The U.S. and much of Western Europe cannot compete globally on labor cost. If manufacturing and agriculture continue migrating offshore, vulnerability compounds over time. He describes this as a national emergency in slow motion. The worst effects have not arrived yet, but the trajectory is clear enough that reversing course will be harder once they do. Domestic automation in agriculture and manufacturing is the only viable path to self-sufficiency that does not require a collapse in living standards.

Mikesell says Carbon Robotics is on a path toward an IPO but does not provide a timeline.