Uber’s Robotaxi Playbook, End of Human Driving & $10B Bet on Robots | Dara Khosrowshahi (Uber CEO)
In a Nutshell
Uber CEO Dara Khosrowshahi predicts autonomous vehicles (AVs) will surpass human safety in 25 years, potentially ending routine human driving and driver's licenses while preserving fun activities like racing; Uber's $10B bet on robotaxis leverages partnerships with Pony.ai, WeRide, and BYD for scalable, cheaper mobility. Key opportunities include vertiports for flying taxis, embedded AV insurance, multi-robot coordination via Uber's real-time dispatching expertise, and extending the flexible work platform to 20M people by 2035 amid job shifts. Upholding "do the right thing, period" culture, Uber pursues "rhyming" adjacencies (rides to Eats, Freight, air mobility) with big bets now that it generates $10B cash flow annually.
These notes were generated by AI and may contain inaccuracies.
Teen asks if they need a driver's license in a couple of years given autonomous vehicles (AVs). Dara Khosrowshahi responds that they may not have to, but technology will always be part of the human experience. Humans will be demonstrably less safe than autonomous systems in 25 years, leaving it to regulators to decide human license requirements.
Teen question: In 15-20 years, when most cars are self-driving, will human driving become illegal as a safety hazard? Will next generation learn to drive? What about race car driving?
Dara: Race car driving will always exist because it's cool, possibly augmented by technology. Mass production of AVs will take time as they are currently much more expensive than regular cars, but progress is faster than expected. Society's adjustment speed is uncertain.
Slido question: What's harder—convincing regulators AVs are safe or convincing passengers?
Dara: Passengers are easy; in Atlanta and Austin, 80% accept AV matches and love the experience, 20% decline. Regulators move slower, but more people are excited than not.
Dara: In next 15 years, unlikely; in 25 years, humans less safe than AVs. Regulators decide license standards, possibly more demanding tests. Likes free will—if you want to drive, fine. Foresees fewer drivers like fewer horse riders today.
Comment on Middle East deaths from young teens with fast cars: AI overlay could allow fast driving but prevent swerving off road like bumpers.
Question on flying cars (e.g., Joby): Real estate opportunities—rooftops or heliports?
Dara: Discussion on vertiports for mass market. Designed for multiple vehicles landing/taking off in high-traffic areas like airports, city centers. Uber data identifies optimal locations for business and traffic reduction. Real estate opportunity—contact Uber or Joby. Also opportunities for small islands.
Elchin (solid shape construction materials marketplace, uses Uber for LTL): Starting a marketplace now with AI, what to focus on first?
Dara: Uber's marketplace always supply-led (vs. Expedia's audience focus). Build liquidity (more cars, restaurants, retailers). With product-market fit, demand follows. Build easiest-to-use tools for most liquid supply.
Burgundy (lawyer, pro-automation): Opportunities for insurance via products liability to remove human error (speeding, intoxication) while assuring coverage.
Dara: Uber professional drivers safer than average (data shows). AV era needs layers: AV provider insures car/driver quality; additional layer covers rest. Half of US court cases are car accident-related—huge societal benefit to eliminate.
Shilpi (led robotic systems at Amazon Air Hubs, now Quin Labs continuous bio-sensing): Vision for multi-robot systems (Joby aircraft + ground AVs)? Bottleneck: coordination, monitoring (faced at Amazon).
Dara: Uber manages 9-10 million drivers globally, recalculating city states real-time. Dispatches predict 5-6 seconds ahead for network efficiency (e.g., reserve closer driver for predicted hail). Humans unpredictable; machines predictable with higher acceptance rates. Uber does 40 million trips daily across diverse markets—excited for machine coordination.
Mark: Affordable Living as a Service (housing, food, energy, connectivity, mobility for $250/month)—won Xprize Visioneering, $50M prize. Challenge Uber to fund/compete for stability in transition.
Dara: Uber ensures service in outer areas/transport deserts. AVs promise safer, eventually cheaper transport for all.
Peter Diamandis: Addresses fear/uncertainty; new Xprize for positive future films. Universal basic services enable dreaming. Disruptive childhood in Iran context.
Dara: Decouple housing from land possible.
Charles (Zoom): Uber exited China; China expanding to Middle East/emerging markets, 70% cost savings via AVs (no driver). Lessons and plan?
Dara: Exit pre-his tenure was right—China hyper-competitive (Expedia experience). Big stake in Didi (largest Chinese ride-share). Partners: Chinese OEMs like BYD (affordable EVs globally), AV players like Pony.ai, WeRide, hopefully Baidu outside US. Build bridges—AV industry where China excels.
Andy (Hong Kong, autonomous insurance for EV/AV/robotics): Challenges/desires for underwriters/actuaries? Embed insurance in AVs?
Dara: Yes to embedded. Model TBD (driver, platform, combo). Need volume for actuarial data—AV trips <1% of Uber's growth last year. Law of large numbers needed for pricing.
Ilia (13-year-old teen): Need license at 16 if AVs/drones do everything?
Dara: In 3 years, still get it—AV mass production takes time, expensive now. In 10 years, no need to drive, but fun—encourage it.
Predictions: Sons (now adults) won't need licenses soon.
George Bain (Untapped Ventures, invests in autonomous economy): Uber's role amid billions of robots? Hundreds of millions out of work?
Dara: Press dramatizes machine replacement. Historically, automation augments (e.g., Chinese OEM factories: robots + humans overseeing). AI automates 20-30% tasks, not 100%. Society adjusts with new work, but pace faster than expected—open question if fast enough.
Uber as flexible labor platform: Extend to 20M people by 2035 (vs. 10M now), different tasks. Private/public retraining/new jobs needed. Don't worry—act.
Peter: Uber as societal capacitor for flexible work (job for week vs. life).
Per (Chile, family shipping oil/gasoline/jet fuel): Will EVs fully replace combustion? Timeline?
Dara: Physical tech transitions slower than digital. EVs better product, but US charging infrastructure lacking (vs. Europe traction). AVs are EVs, so network shifts. In 20 years, few combustion engines.
Sagar Chopra (Empower Sleep co-founder, ex-Uber 4.5 years international expansion): Preserve entrepreneurial culture (run to fire, agency) as company scales/doubles?
Dara: Companies face this scaling challenge and we don't get it
One of Uber's core values is do the right thing, period. This value was introduced during a difficult time when the company wasn't always doing the right thing. Unlike most company values with headlines and descriptions, the description here is simply "period." When introduced to employees, the explanation was that every single person is responsible for doing the right thing personally. Sometimes that means playing safe, taking a risk, running fast, or folding your cards—you never know exactly what it is, but individuals three or four levels down can impact the company's direction.
This culture of go-getters came from Travis, which Dara inherited and pushes forward. Do the right thing is a message of responsibility at all levels.
Encouraging Big Bets at Scale
As companies get larger, they tend to get more conservative, but the opposite should be true. When Dara joined Uber, the company was losing $4 billion; losing another billion would have been a disaster. Now, Uber will cash flow $10 billion this year, so it should take big bets. Some will fail, but the company can afford it. Dara fights this conservatism daily—with himself, the board, and the team—emphasizing that the team are the doers.
To decide on pursuing new fields rather than staying conservative and collecting cash flow, Uber continually experiments. New pursuits must rhyme with what they do—a startup is better for something completely new and crazy.
Uber's path:
- Started with rides (moving people).
- Then Eats (moving food, i.e., things).
- Then Uber Freight (moving bulk).
- Elevate (moving people in the air).
The pattern is identifying areas where Uber has a core skill set giving a right to win in an adjacency.
Uber AI provides flexible work opportunities to drivers or new people for designing models, labeling, or testing—rhyming with Uber as a platform for flexible work.
If it rhymes, I'll listen. If it doesn't rhyme, it's better for someone else to do it.
Keep Peter H. Diamandis in your library
Save the videos and channels worth coming back to, and find them again in one place.
