Uber, Lucid & Nuro team up on robotaxis by 2026 | E2173
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this episode of This Week in Startups, Jason Calacanis and Alex dive deep into the future of autonomous vehicles with Dave Ferguson, co-founder and president of Neuro, a self-driving technology company making waves through their groundbreaking partnership with Uber and Lucid Motors. Ferguson, who brings nearly two decades of experience from the DARPA challenges to Google's self-driving program (26:55), shares how Neuro pivoted from building custom delivery vehicles to licensing their technology after capital costs skyrocketed. The conversation explores the critical technical debates shaping the industry—from the LiDAR versus computer vision arguments (60:30) to probabilistic versus deterministic approaches (24:43)—while revealing that most leading companies have actually converged on similar foundational model techniques. Ferguson makes a compelling moral case for rapid deployment, arguing there's an ethical imperative to launch once systems are even marginally safer than human drivers (46:06), given the 40,000 annual traffic deaths in the US alone. The discussion culminates with insights into the global competitive landscape, particularly China's impressive automotive advances, and projections for when autonomous vehicles might achieve meaningful market penetration in the next five to ten years.
Speakers
Dave Ferguson
Cofounder and President of Neuro, Carnegie Mellon PhD who participated in the winning DARPA Urban Challenge team. Former Google self-driving car project veteran with nine years building autonomous vehicle technology and five years of driverless operations.
Jason Calacanis (Host)
Brooklyn-born serial entrepreneur and angel investor, former magazine and blog publisher turned podcaster. Early Tesla investor with IPO shares, owned the first Model S and sixteenth Roadster, and early Uber investor (third or fourth in).
Alex Wilhelm (Co-host)
Technology journalist and podcast co-host covering startup and venture capital markets. Regular contributor analyzing emerging tech trends and startup strategies.
Key Takeaways
Embrace the Data-Driven Partnership Model
Licensing autonomous driving technology to established mobility players like Uber while integrating with premium OEMs creates a capital-efficient path to scale. Rather than burning billions building your own fleet, focus on per-mile revenue sharing and fixed licensing fees that align incentives with operational success. (08:48)
Deploy the "Epsilon Safety" Ethical Framework
As soon as your autonomous system achieves even marginally better safety than human drivers, you have a moral imperative to deploy—despite perfectionist instincts. The U.S. loses 40,000 lives annually to human-driven accidents. Every day you delay deployment for incrementally better performance costs lives that could be saved today. (46:06)
Engineer for Edge Cases, Not Nominal Performance
Modern AI can handle 98-99% of driving scenarios admirably, but commercial deployment demands handling "people lying down in wet suits on skateboards shooting down hills." The massive gap between impressive demos and putting your kids in a driverless car lies in systematically solving thousands of edge cases through diverse real-world data collection. (43:03)
Prioritize Regulator Trust Over Technical Prowess
Cruise's downfall wasn't their autonomous system hitting someone—it was allegedly concealing the full incident from regulators. Your technical achievements mean nothing if regulatory trust evaporates. Invest heavily in transparent relationships with NHTSA and state agencies from day one, before you even tell employees what you're building. (38:42)
Choose Sensors Based on ROI, Not Religious Debates
The camera-versus-LiDAR debate misses the point—successful systems use both, with LiDAR now costing under $1,000 per unit rather than $75,000 in early prototypes. The question isn't philosophical preference but practical value: does the incremental safety benefit justify the incremental cost? For edge cases like black cars at night, LiDAR provides physics-based range data that cameras simply cannot match. (60:49)
Statistics & Facts
- There is a pretty strong ethical argument that as soon as you are Epsilon safer than an average human driver, you are actually costing lives if you delay deploying this technology because we lose forty thousand lives a year on US roads. (04:06)
- Neuro has raised $230 million to build hard tech, highlighting the capital-intensive nature of autonomous vehicle development. (48:38)
- Waymo recently released data suggesting they are something like five times as safe as a human driver, with enough statistical significance behind the claim. (44:22)
