"Gradually, then suddenly"
On the current moment for AVs coming out of Ride AI 2026
This is the Ride AI newsletter: The most comprehensive weekly digest of news and intelligence at the intersection of technology and transportation.
All of the sessions from the event are now live on our Youtube channel! Huge thanks to our trusty videographer Vin for all the hustle.
The magic, the mundane and the convergence of “real-world” problems in scaling AVs
Mike here, newest member of the Ride AI team, coming to you with additional reporting coming out of the event. As a complement to Sophia’s official recap, below is my take on trying to capture the vibes of the day, based on all the fun B-side cafe, bar room and demo side chatter.
Like many of the attendees, I arrived in style thanks to our friends at Waymo. I got a block away from SF Jazz at about 7AM and there was still construction shutting down one side of the street. The worker manning the STOP - SLOW sign briefly turned it around but then started to chat with a buddy on his left. Hands followed eyes and he was blissfully unaware that he had titled the sign back to essentially showing us neither STOP nor SLOW.
So my Waymo stopped and blocked traffic as I frantically made every hand gesture in his direction mia nonna taught me. For some reason, I assumed at first I was not allowed to open the window. PSA– you can! Luckily, California drivers are much friendlier than those on the mean streets of Queens I was raised on.
The car did its job to keep me safe, the human it was relying on out in the world did not. It was a poetic arrival for me to an event that at its core, connected an industry around the question of “what do we do now that the technology has come this far?”
If my reflections on the day sound a little bit like the perspective of someone new to AVs who still possesses childlike wonder at what you’ve all built, yes.
***********
“Gradually, then suddenly”: The old Hemingway trope about how he went broke is perhaps the most abused cliche in talking about technological progress but in the case of AVs, it just hits.
I heard several recollections of folks’ earliest experiences with autonomy, often in the early-mid 2010s, ranging from a glorified golf cart AV demo that stopped short so violently it hurled a laptop in the air to wild swerves at the sound of sirens. And then one day, it all worked.
The first ride is magic, the fourth is mundane: One story will stick with me the most. It was about how everyday people experience self driving for the first time when they visit SF. The first time East Coast folks come into town they insist on taking a Waymo and are awe-struck. Next ride, there’s a rally cry of “LET’S DO THAT AGAIN.”
But by the fourth ride or so, the childlike wonder has become commonplace. They are opening up Lyft, Uber and Waymo and making a decision strictly based on price. Even magic rapidly becomes a commodity.
No matter how amazing a technology is, the old marketing adage still applies. People don’t want a quarter-inch drill, what they want is a quarter inch hole. The good news here of course is that the quarter inch hole here will save hundreds thousands of lives in the years ahead.
Convergent real-world problems: Erik at Zenesact had a great counterpoint quip on his panel on how advances in AI are upgrading AV development, alluding to the fact that all the compute in the world won’t clean the debris of a sensor. It was one of four mentions of cleaning sensors I heard on the day, alongside in-depth chats on data labeling, tire nuances, EV trucking charging specs, faulty traffic lights and a plethora of other infrastructural challenges of scaling autonomy.
Bot Auto CEO Xiaodi Hu captures an important dynamic related to this well in this post from after the event.
What struck me most is that the people truly doing the work are starting to converge. A few years ago, autonomy still felt like many separate experiments. Today, the serious builders are developing a clearer, healthier consensus of what this industry must become.
Look for us to profile more early stage builders working on picks and shovels for autonomy in future editions of this newsletter.
Automobile America: Robotaxis and to a slightly lesser extent trucking dominate much of the AV narrative, and rightfully so given the real-world applications of both technologies already on the roads at scale. But America is a nation quite literally built on and for the passenger automobile.
On stage, Alex Kendall noted that the addressable market for autonomy in passenger vehicles could well be 100X of taxis, a conversation that continued into the afterparty palomas. Right before I left, I overheard impassioned stories of folks recounting their first adventure after getting their license– my favorite was an old F-150 taken straight to a McDonalds.
Driving does absolutely nothing for me— I grew up in New York and didn’t even get a driver’s license until I was 23. But for many people, when does FSD threaten a small sense of identity? Tensor’s positioning and demo which emphasizes the ability to flex between L0 - L4 positioning captured this well and I expect others to follow suit as we get closer to mass scale L3/4 passenger automobiles.
The regulators at the gate: Autonomy is a uniquely enigmatic policy space that often does not break neatly along expected partisan lines. AV’s are under regulatory scrutiny and in some cases outright attack in red and blue states alike…but also have unlikely supporters on both sides of the aisle.
This obviously creates a massive “what the hell is gonna happen” uneasiness cited by McKinsey on stage and many in casual chats. If there’s a silver lining, it’s that the enigmatic nature of AV policy makes it a uniquely interesting space for the most talented power brokers in Washington to do their thing.
Determinism > Cynicism: The night before the event, I briefly met up for drinks with a friend who has a very good job at one of the buzziest AI application companies in SF. When he rolled in and I asked how his day was, he promptly replied “just another long day in paradise at the slop factory.”
It was a half-hearted comment sure, but it’s a vibe that’s starting to permeate more corners of AI, especially those that don’t touch the physical world. Luckily, any semblance of cynicism in their work was not a sentiment I picked up on from a single person at Ride AI.
But this broader vibe matters for autonomous vehicles because as James shared in “It’s Time to Market”, it is impossible to fully decouple AI from mobility in the zeitgeist, especially as the ever-increasing raw power of the models unlocks more possibilities in autonomy. The unbridled optimism, determinism, and consistent messaging about how AV’s improve the quotidian life of everyday people matters.
Thank you all so much for welcoming me to this community– we’ll see you at the next Ride AI event! In the meantime, please feel free to share any feedback on the event or our broader industry coverage to mike@rideai.org. I’m also naturally on LinkedIn and X.
Takes from attendees on the day
Shobana Sankaran left "buzzing" from conversations that felt like a step-change over year one; her LinkedIn recap flagged that safety parity is being redefined as a statistical bar, TCO in trucking is being rewritten around asset utilization, and Waabi's "shared AV brain" across trucking and robotaxis genuinely blew her mind.
Carlos Oliveros Forero of Robot.com came away convinced AVs have hit a "real inflection point," and his LinkedIn post captured the fun of pairing on-stage conversations with live robot demos on the curb.
Candelaria Cazes of Uber cited James’s opening keynote and reflected further on the fact that the AV hurdle is no longer engineering but positioning.
Ride AI co-host Ed Niedermeyer waxed a bit nostalgic on riding in a Waymo at 65 MPH on the way home, which gave him an "old familiar thrill" and a deep appreciation for the industry's progress; his gratitude goes to the speakers, sponsors, attendees, volunteers, and friends who made the event "enjoyable and effective."
Our volunteer extraordinaire Kofi Agarye-Kwabi (former Uber Ghana country manager, Wharton MBA— hire this man) said the networking alone was worth the cross-country trip; he loved getting up close to Tensor's Robocar and indiGOtech's SmartWheels EV, meeting RoboDock's team on robotaxi maintenance, sitting in TRI's next-gen ADAS, and swapping notes with founders building in everything from AV accessibility to ridesharing for healthcare
Michael Wiesinger of Kodiak noted that his big realization at Ride AI was how wide the perception gap is — most people still ask when AV trucking will be real while their clients operate fully driverless Class 8 trucks
Karen Peck of Voltera cited her excitement around how much both the event and the industry have grown. Her key takeaway from CEO Brett Hauser's talk: AVs scale on infrastructure, and fleet hub networks are the unlock for city-wide commercial deployment.
Chris Lichtmannecker of Mobileye shared a heartfelt nod to San Francisco and his fellow panelists and shared his central takeaway was that scaling AVs is a tightly coupled systems problem where progress is constrained by the weakest link.
Fresh off his Calc II final, Robotaxi Tracker creator and incoming Telsa intern Ethan McKanna posted on X that he “left feeling even more excited about the future of AVs and real world AI.”
Lloyd Lee of Business Insider, recapped the day by citing the shift away from AV-hype coverage toward operations, maintenance, fleet infrastructure, and commercialization.
David Moss dropped a early contender for all-time tweet reply. Hit me up next time you do a cross-country FSD haul to Myrtle and I’ll buy five rounds at McAdoo’s.
Thanks for reading! We’ll be back tomorrow with an essay version of James’s “It’s Time to Market” keynote address.