Patrick
Welcome to the last episode of the Barclays Brief Podcast for 2025. We're finishing the year on a high as I'm joined today all the way from San Fran by my colleague Ross Sandler, who's our Head of US Internet Research at Barclays. Ross, thanks a lot for joining the podcast today.

Ross
Hey, Patrick, super pumped to be here. And you know what they say, ‘save the best for last’. So happy holidays to everybody listening to the pod. And let's get going.

Patrick
Okay. Cool. So, Ross, you had quite a run recently. You've hosted a massive tech conference. You've joined our autos conference. You've been at our eat, sleep, play and shop conference. You've been doing lots of other things as well. So, what's the mood amongst investors at these gatherings?
You know, what are the hot topics that everyone's talking to you about? Are there any themes that the market is overlooking right now in your sector? You know, what's the buzz?

Ross
It's all over the place. And it changes on a weekly basis. So, we've kind of gone from being very excited about the entire AI basket to bifurcating between, you know, names that are exposed to Google or Gemini winning the AI race. Then we flip to names that are exposed to open AI and ChatGPT winning the race. And then we've had a risk off where investors have kind of moved away from the AI infrastructure related, part of the trade. So, it's all over the place. And usually what happens next is we get to a point where, you know, the valuations just look so compelling, that, that it's hard to pass up on some of these things.

And then we'll get like an AI model or an AI product breakthrough that happens and the whole cycle repeats itself and everybody gets excited again.

Outside of the AI trade, we have a lot of things going on. There's the debate around robotaxi versus human ride, hail. There's, you know, web 1.0 companies that we cover that are getting choked off by Google search referrals.
There's a fresh IPO cycle and there's just a lot going on. So, happy to talk about, you know, whatever you guys would like.

Patrick
Okay, great. Well, as you know, us, because there's been very limited prep here. This is completely unscripted, this podcast. So, you know, we could go anywhere. Robotaxis. You mentioned that, as a theme, it's very interesting.
I've got to admit; I've never ridden in one yet. What's the experience like? The consumers really trust them. And can you just explain to me the different types of robotaxis? There are some with people in the front, some without, you know, steering wheels moving by themselves. It's all a bit sci fi to me.

Ross
Absolutely. Robotaxis are amazing.

You just got to come out to San Francisco, and you'll see what all the buzz is about the first time you take one, you're kind of blown away. I liken it to, you know, when, you kind of transition from having a BlackBerry to having your first iPhone. This happened about 18 months ago for me when I, when I first got into a Waymo without the cockpit driver.

The robotaxis just quietly and smoothly kind of cruise around without any sense of urgency. But the coolest thing is you're in your own little world. You're not talking to any annoying driver. There's no noise. You can either, you know, chill out, listen to music. You can work, you can do conference calls. You can even do a Zoom all with like the total privacy and safety. So, Patrick, you need to come out and see what all the buzz is about.

Patrick
I will, I don't need much excuse, Ross. I'll be there on the on the next flight, if I could. So. Yeah. Talk to me about the regulation and the rollout of these things. So, you know, 2025 feels like it was a test year for the robotaxi market. 2026 what's it shaping up to be, you know, in terms of rollout?
And can you just roll these things out in different states without much friction? Or, you know, what's the regulatory challenges that the companies involved in robotaxis face?

Ross
So, you're right that 2025 was mostly about announcements and about testing of small fleets. The only city, like I said, where you see the glimpse into the future is in San Francisco.

You kind of walk down, you know, a city block. You can't go, you know, ten feet without seeing half a dozen Waymo's or other robotaxis cruising around. Every other city's going to look like this eventually. We have about, 30 cities going from, you know, kind of initial announcement to deployment stage in 2026. And we're going to see a similar trend, outside the US and some of the top international markets.

As far as the regulatory process, it's a little bit of a whack-a-mole. So, there's no, national, autonomous, you know, regulation in the US, you kind of have to go state by state. And what the regulators do, it's generally like, the California Transportation Authority, they will look at the data from a robotaxi fleet. They'll assess; how safe the service is.

These companies have to provide what's called a disengagement report, which is, you know, the number of miles travelled before the cockpit driver has to override the system. And as long as those engagements are not critical and are spread out over, you know, tens of thousands of miles, then the system is deemed to be generally safe, and you kind of get the okay to go ahead and operate without the cockpit driver.

So that's where Waymo is. Everybody else is kind of in step one, which is, you know, driving around logging miles and testing with the cockpit driver. And, you know, it's somewhat agnostic. San Francisco's kind of a crazy city with, you know, a little bit of a grid, a little bit of a windy road. There's, you know, rain, there's fog, there's, you know, bad weather.

Depending on what city you're going to have, you know, different constraints on the system. And, you know, that's what these companies are trying to overcome as we, as we get to the deployment stage in 2026.

Patrick
Okay. Can you just unpack for our listeners a bit more about the different types of robotaxis out there, the different technologies, because, yeah, I've looked at lots of pictures.
I've watched lots of videos about this in the past. But, you know, there's a massive range, isn't there, in terms of different companies doing different things.

Ross
That's right. Right now, the space is significantly supply constrained. The only company that's operating with any scale is Waymo. They've got about, you know, 2000 vehicles deployed right now, about half of which is in state of California. That number is going to go up to, you know, about double that by the end of 2026. And that's not nearly enough to meet the demand that's out there.

Tesla is probably the next company to watch on deployment. Their system is a little bit different than, Waymo. Waymo has got a LIDAR system. Tesla has a, camera-based system. They both use a significant amount of AI in terms of the intelligence on the vehicle. So this is like the first manifestation of AI kind of moving out of the data center into the physical world.

Robotaxis are essentially, driving robots with an intelligence layer added on. And you're going to want to watch Tesla because they have a unique approach with their full self-driving, their software, their intelligence stack.
And they also are the one company that, has in-house production, unlike Waymo, who buys their cars from Jaguar or from, you know, Hyundai or others in the future.

So Dan Levy, who covers Tesla in the auto space for us, will tell you that these guys are in a position to really turn things up. You know, rather quickly so they can produce somewhere in the neighbourhood of 50,000 model Ys in a given week. And that's a massive number relative to the size of the deployments that that we're seeing from Waymo and from, you know, other pure play autonomous providers.

The one gating factor for them is what we call the remote operator ratio. So if you look at the different, robotaxi fleets, they tend to have about one person, in a remote kind of a air traffic controller type role, looking at about five vehicles on the road, we think the, you know, the future ratio between remote operator and vehicles deployed could get up to, you know, as high as 50.

And so Tesla, while they could produce a huge amount of robotaxis and deploy them, are going to be somewhat gated by this, this ratio between, the air traffic control and the cars on the road.

Patrick
Understood. And, so headlines, just a couple of days ago, but that robotaxi begins testing without, in vehicle safety monitors at Tesla. So clearly, they're making strides as well.

Let's zoom out and jump forward sort of five, maybe ten years. Right. So autonomous vehicles are everywhere. Accidents are down. Consumers are on board. No one's nervous about these things. What happens to the traditional ride hailing players in that world? Are they going to have to shift from asset light business models to something a little bit heavier?

Ross
Absolutely. the human ride hail providers are in a bit of a tricky situation. Some of them are partnering with a bunch of, different technology companies but the big problem that they have is, is that around 70% of human ride hail trip volume, lives in just the top 20 cities in the US.

The remaining kind of suburbs and tier two cities only represent about 30% of trips, which makes sense, because if you're in the suburbs, everybody's got two cars, they've got a garage, a driveway they can park, you know, at the grocery store. They don't need to take robotaxis around town the way that you do. If you live in downtown San Francisco, for instance.

And if you look at the concentration of those trips, basically the bulk of the volume is in the downtown metro and on trips that take the passenger to or from the airport. So, to, to disrupt the human ride hail industry, you only need about 1000 vehicles in each of the top 20 cities before you can really put a dent in, in that human ride hail 70%.
If you look at San Francisco data, which we have, you know, pretty good handle on Waymo was able to take about 30% of the downtown metro area in terms of market share, with only a few hundred cars and without the airport. So, you know, stepping back, if if we fast forward five years and there's a handful of different, you know, fleet owners that are operating robotaxis, each of them having, you know, a thousand cars in each of these top 20 cities, it's going to put a huge dent in the human ride hail, industry.

So, it's up the traditional ride hail partners to deploy their own robotaxis. And they're doing some of this already. So that's what they're trying to do to, to kind of bridge to the other side.

Patrick
Okay. So, we're going to have to bridge the end of the podcast, sadly. But Ross, thanks so much for joining.
I mean, I would have happily spoken to you about any of the other topics you brought up at the start of the pod. You know, we could have talked about AI funding, CapEx and the narrative around compute that, for a long time. So, can you just promise the listeners you'll be back on the podcast next year?

Ross
Absolutely. 26 is going to be a massive year for AI. We've got, you know, the Blackwell models coming out, GPT six equivalent. We've got agentic we've got all kinds of new things coming to the forefront. Means a lot more CapEx coming as well. So happy to hop back on and, go through that next year.

Patrick
Look forward to it. Thanks, Ross.

So as Ross heads into San Fran in a autonomous vehicle and I head back down to my desk here in a miserable and cold, dark London, here's what I'm taking away from this conversation. Robotaxis once felt like sci fi to me, but then now, clearly a reality in major US cities. Still, widespread adoption is going to take time, scale and production regulatory challenges, building trust with consumers and integrating with existing models are big hurdles.
And as AI funding and compute costs reshape the tech narrative, 2026 could be a pivotal year for autonomous vehicles and the broader internet ecosystem.

So, I'm looking forward to getting Ross back on the show next year. Thanks a lot for tuning in to the Barclays Brief. We're going to take a short break for the rest of the year and be back to it in early January. We'll have more unique insights to help you navigate a volatile market, and we've already secured a great pipeline of guests. So hit subscribe. And both Ronnie and I wish you all a relaxing holiday season.