BREAKING: Tesla Launches UNSUPERVISED Robotaxi Dallas and Houston

Added: 2026-05-05

[00:00:00]Just now Tesla announced they've launched their robo taxi service in both Dallas and Houston. Both Elon Musk and Ashra Koswami joining the announcement with Elon saying to try the service.

[00:00:09]This seems to mean it will be open wide to the public. And in the video they shared it appears these are unsupervised robo taxis. If this is the case, this is a very big deal. It might mean that from now on or at least for certain states, Tesla will no longer need to start with safety riders. Both Dallas and Houston are geofenced right now, but as expected with Houston's area estimated at 25 square miles. The rest of this video is a recording I did yesterday before this news. It's even more relevant now that we've got more evidence Tesla truly intends to grow robo taxi at scale. Hope you enjoy it.

[00:00:40]>> Try something today. This is the most comprehensive list I've ever put together of everything Tesla needs to scale robo taxi. So, not the headlines or the hype, the full list. every single thing that has to be true for Tesla to go from 13 cars in Austin to let's say a million cars across the country. And I'll just tell you up front, Tesla's actually hiring in 34 cities right now, which means they already know what they need and they're paying for it. Here's the deal. Tesla investors, they love going to Robo Taxi Tracker. They keep pointing out the same number. There's 582 vehicles in a total fleet. 13 of them driving around Austin with nobody behind the wheel. And then people are looking at Whimo with thousands of cars and they'll say, "Yeah, Tesla's really way behind." But I think that's actually the wrong scoreboard. The right scoreboard has four boxes on it. Okay, four pillars. Every single one of them has to be ready. So first is software, two is operations, three is hardware, and of course four is regulation. If any of these is missing, the whole thing collapses. All four have to be in place at the same time or it doesn't scale.

[00:01:43]And here's the thing about hiring. You cannot fake hiring. You can fake a press release. You can fake a demo. You cannot fake payroll. When Tesla posts jobs in Phoenix, Miami, Jacksonville, Belleview, Henderson, that's where real money is going. That's where the actual strategy is pointing. So, right now, there's 34 cities active postings right now. So, what I want to do today is walk you through the full checklist, every piece of the stack, every role Tesla's hiring for to build it, and every number that tells you how close we really are. No hype, just the complete list. By the end, you'll know exactly what Tesla has and what it still doesn't have. So, pillar one, of course, is software.

[00:02:22]Software is number one. Buy a mile.

[00:02:24]Nothing else matters if the car can't drive itself safely. You can have a million cars charging on every corner.

[00:02:29]Hiring in every city, doesn't matter. If the software isn't dramatically safer than a human driver, the whole business is dead on arrival. So, where are we actually? Tesla rolled out FSD version 14.3 in early April. firmware version 2026 296 and the word inside Tesla community is that version 14.3.1 is the version clean enough to go unsupervised at scale. So here's the piece that stopped me dead in my tracks.

[00:02:57]The most recent safety data on the unsupervised fleet in Austin. Zero incidents last month. Zero. 13 cars driving around a real American city with nobody behind the wheel for a full month. No incidents.

[00:03:11]Think about what that actually means. So you got human drivers in Austin have crashes every single day. Tesla's unsupervised fleet had zero for the month. Now of course, yes, the fleet small. I'll be honest, you need more miles to say anything definitive. But the direction's right. The curve is bending towards safer, not less safe, and version 14.3 just cut the reaction time by another 20%. Plus a rewritten AI compiler under the hood. Elon called version 14.3 the last piece of the puzzle. Is it not really sure yet? I don't think so. Elon does say a lot of things. But when you combine his claim with the real safety data showing zero incidents, that's the first time software and reality have actually been lining up like this. On the hiring side, Tesla's still aggressively hiring AI engineers, simulation engineers, and autonomy researchers. If software were done, those job postings would be shrinking. They're not. They're growing, which tells you Tesla's still investing hard to keep improving the stack. But the end of the day, I think most people will say that 14.3.1 box one is checked.

[00:04:13]Pillar two is operations. Okay, so this is the one investors almost never talk about. And honestly, this is where the biggest signal is hiding. So running a robo taxi service isn't just software and cars. You need to operate a fleet at scale in cities 24 hours a day, every day. Here's where Tesla Yoda's post really opened my eyes. Tesla's currently hiring for robo taxi roles in 34 cities.

[00:04:37]Yeah, 34 cities. Not promised or planned, active job postings right now.

[00:04:43]If you went to Tesla's careers page this afternoon, you could apply. But it's not just the number of cities. It's the variety of roles. Each role tells you something different about what Tesla's actually building. So, let's walk through them. Role one is the AI safety operator. This is the person who sits in the car during the supervised testing, not driving, monitoring, watching what FSE does, taking over if anything looks off. Tesla's hiring AI safety operators in Tempe, Belleview, Austin, Dallas, Houston, Tampa, Orlando, Doral, Jacksonville, Peabody, Marina, Del Rey, Henderson, Flushing, New York, and Bridgeville, Pennsylvania. So think about what that list tells you.

[00:05:25]Tesla's not just building in the four states with regulatory approval today.

[00:05:29]Tesla's prepositioning people in cities where approval hasn't even happened yet because they know it's coming. You don't hire safety operators in Flushing, New York if you're not planning to run rides through Flushing, New York. Role two, the vehicle operator for Robo Taxi.

[00:05:44]That's a different role. This person drives an engineering vehicle to collect data, camera data, audio data, real world driving scenarios. All of it feeds back into the training system. These postings are heavy in Austin, California, and Texas, but also spreading. So, when a new vehicle operator posting shows up in a new city, that's Tesla starting local data collection 3 to 6 months before the rides actually launch. Ro is the remote operator and tea operator. This one makes headlines. Tesla's hiring SPAS software engineers for the tea operation wing of the Tesla and robo taxi division. plus actual remote operators, humans who sit at a station. The job isn't to drive the car, it's to help when the car gets stuck. Construction zone, a weird situation, passenger issues. The remote operator clears the problem, the car resumes. Every real robo taxi service on the planet does this. Whimo does it. Cruise did it. It's how you bridge the last 1% of edge cases. Tesla building this team means they're serious about real deployment, not a real demo. Role four, robo taxi fleet support specialist. This is the night shift crew. Tesla's hiring fleet support specialists for overnight operations in Austin, Dallas, Houston, Las Vegas, Tempe, Palo Alto, Tampa, Dural, and Orlando. Nice shift tells you something huge. If you're only doing a daytime demo, you don't need overnight staff. If you're running a real 24/7 commercial service, you do. Tesla staffing like the service never sleeps because it won't. Role five, service technicians. Robo taxis will break down.

[00:07:19]They scratch. They need tires and they need fixing fast because every hour a robo taxi is off the road, it's lost revenue. Tesla already runs 1,300 plus service locations globally. Now they're adding robo taxi specific technician roles. Fleet maintenance, detailing, fast turnaround service. Ro six, mapping and localization engineers. For every new city you launch in, you need mapping, street geometry, traffic patterns, local rules. Tesla's hiring mapping engineers continuously, one of the least glamorous and most necessary parts of the staff. Role seven is customer support and ride support.

[00:07:56]Somebody has to pick up the phone when a rider has a problem. Tesla has been staffing this up. Riders in Austin already report realtime support inside the app. That's a hired team in a call center answering tickets, fixing rides.

[00:08:09]Role eight is charging operations.

[00:08:11]Here's the one people forget. Tesla's hiring people to manage robo taxi specific charging operations because the fleet can't just share superchargers with random Tesla owners when demand is high. You need dedicated charging slots, dedicated cleaning, dedicated routing.

[00:08:26]Put all this together, you have an AI safety operator, you have vehicle operator, you have remote operators, teleoperation engineers, fleet support specialists, service technicians, mapping engineers, support reps, charging operations. Tesla's hiring every one of these roles in 34 cities right now. This is not a pilot project.

[00:08:46]This is a commercial service roll out in motion. And this is on top of the infrastructure Tesla already owns.

[00:08:52]80,000 Supercharger stalls at 8,500 sites globally. 2,500 new stalls added just in Q1 2026.

[00:09:03]Whimo doesn't have a charging network.

[00:09:05]Cruise doesn't have one. Zuks doesn't have one. Nobody has one. Tesla is the only robo taxi operator in the world that owns its charging infrastructure.

[00:09:15]Add 34 cities of active hiring on top of that and you start to see the real picture. So box two checked. Pillar three is hardware. Okay. So software is ready. Operations is built and hiring.

[00:09:28]Now you need cars. Not just any cars.

[00:09:31]Cars that can run the software. Cars with the right cameras. Cars with the AI4 chip on board. And eventually cars built from the ground up for this job.

[00:09:39]Here's where Tesla's position isn't fair. Every Model Y, Model 3, Model X, Model S built since mid 2023 has AI4.

[00:09:47]Every single one of them. That's millions of cars already on the road, all potential robo taxi candidates.

[00:09:55]Fortunately, version 14.3 runs on AI4.

[00:09:58]That's the key unlock. The chip is powerful enough. The cameras are positioned right. The onboard compute can handle unsupervised driving with no remote brain required. So, when you ask, "Okay, Tesla, show me you can build the cars." They already built millions of them. But here's where it gets really interesting. The Cyber Cab productions just started. at Gigafactory Texas this month, April. Drone footage shows more than 50 cyber cabs on the Giga Texas campus going through crash testing. The target, 2 million units per year, priced under $30,000.

[00:10:32]Right now, the Tesla Robbo taxi fleet is Model Wise. Works great, but Cybercap is purpose purpose-built. No st steering wheel, no pedals, two seats built for this exact job. If you want to scale from 582 vehicles to a million, you can't do it pulling Model Y's off the consumer line. You need a dedicated vehicle at dedicated scale. And that vehicle is rolling off the line right now this month. On the hiring side, Tesla's Giga Texas postings for Subcap production are active and growing.

[00:11:02]Manufacturing engineers, robotics technicians, paint shop staff, assembly operators. You're watching a factory ramp in real time through the careers page. Box three checked. Pillar four is regulation. This is the one that makes investors nervous because regulation is the wild card. You can have perfect software, perfect cars, the best operations teams on earth, but if governments don't let you operate, none of it matters. So, let's walk through it. In the US, Tesla's approval to operate robo taxi service in three states: Texas, Nevada, Arizona, and a ride hole service, California.

[00:11:37]Arizona is interesting because that one is statewide, not city by city, the whole state. Then on the Q4 earnings call back in January, Tesla announced seven new cities for the first half of this year. So Dallas, Houston, Phoenix, Miami, Orlando, Tampa, and Las Vegas, going from one city to eight cities in about 6 months. But this isn't really about individual cities. The real prize is a clear federal framework for autonomy. And that's moving. The Department of Transportation has been pushing a unified framework. No final rule yet, but momentum is in one direction and it's favorable. Then internationally, this is the one almost nobody talks about. The Netherlands just approved Tesla's FSD supervised on April 10th. First country in Europe to do it.

[00:12:18]Now supervised isn't unsupervised. I know it's level two driver stays legally responsible. But the approval took 18 months of testing, 1.6 million km they said of European road data. More than 400 compliance requirements. The reason this matters isn't the Netherlands specifically. It's what comes next.

[00:12:38]Germany, France, Italy, Belgium, Spain, all lined up. EUwide approval targeted maybe by the end of summer. So the question is no longer will regulation block this. The question is how fast regulation open up. In the last 6 months, every regulatory move has been forward. So far, zero blockers. And you see it in the hiring too. Tesla's policy and regulatory affair roles open across multiple US states and in Europe. That's the people who actually sit in rooms with regulators and get approvals across the line. They're being hired right now.

[00:13:11]I think box four is checked. So, putting it all together, let's zoom out a bit.

[00:13:16]You have four boxes. Software, operations, hardware, and regulation.

[00:13:20]All four checked. And when I say checked, I don't mean perfect or done forever. Every critical blocker is no longer a blocker. Two years ago, not a single one of these boxes was checked.

[00:13:30]One year ago, maybe one. Now, all four in about 12 months. And the hiring is the thing that convinces me more than anything else. Because you can argue with a press release. You can argue with an Elon tweet. You cannot argue with 34 cities of active job postings. That's money being spent today by a company that has a long history of cutting costs, not expanding them. If Tesla didn't think robot taxi scale was ready, they wouldn't be paying recruiters, background checks, and onboarding teams in 3 to four cities at once. They'd be shrinking, not hiring. Let me play the devil's advocate part of this. You should always hear the bare case. One, of course, it's Elon's timelines. Missed robot taxi in 2020, missed in 2022, missed in 2024.

[00:14:15]So why believe them now? Honest answer, we're no longer debating capability. We have actual cars on actual roads serving real riders right now. The question isn't can this happen anymore. It's how fast does it scale. I think very different questions. Two, the Nitsa investigation. It's real and Nitsa upgraded the probe in March to cover 3.2 million Tesla vehicles focused on FSD in low visibility conditions. If it turns into a mandatory recall, it could slow things down. But the Tesla response has been version 14.3. 20% faster reaction time, rewritten AI compiler specifically addressing the failure modes Nitsa flag and the unsupervised Austin fleet has zero incidents for the month. So data and regulatory concerns are both pointing in the same direction. Three, Whimo. They're not sitting still. Five cities, they have a data advantage, Google's checkbook, but Whimo's cost per vehicle is around $200,000. Tesla's target cost per cyber cab is under 30,000. When you're scaling from thousands to millions, that ma matters enormously.

[00:15:19]Four, consumer demand. Will people actually take a robo taxi instead of a regular Uber? Early Austin data says yes. Ridership growing, word of mouth positive, but a real scale we don't know yet. All legitimate concerns. But notice something. None of them are this technology doesn't work or this company can't operationalize it. Every bare case is about timing or competition, not fundamentals. Different risk profile than it was two years ago.

[00:15:47]Okay, so you're a Tesla investor. What do you do with this? Here's how I think about it. There's three scenarios, right? Scenario one, robo taxi scales slowly. End of 2027, we have a few hundred cars in a few cities. Tesla stays mostly a car company. Stock trades on auto margins plus a little energy.

[00:16:05]Stock drifts. Not bad, not great.

[00:16:07]Scenario two. Robbo taxi actually works in multiple cities by the end of 2027.

[00:16:13]Teslas gets rerated from a car company to a transport platform. Multiples expand. Stock rerates significantly higher. Scenario three. Robo taxi hits escape velocity 2028 2029. Tens of thousands of vehicles running at 75% margins on rides. Tesla becomes a cash printing machine. Multiples expand to tech platform levels. Stock does something most people can't picture up.

[00:16:37]The question isn't which scenario you believe, it's which scenario the market is pricing in today. Based on current multiples, it's scenario one, maybe a sliver of scenario two. Which means if all four pillars really are in place and hiring is actually a forward signal, you want to own it before the rerating, not after. Here's what I would actually track going forward. One, the hiring page. Go to tesla.com/careers.

[00:17:03]Type robo taxi in a search. Watch the city count every week. 34 today. If that goes to 50, then 70, then 100 in the next months and years, scale is accelerating. This is the single most undertracked signal in Tesla today. Two, the Austin unsupervised count 13 today.

[00:17:22]Watch for 25, 50, hopefully 100 leading indicator for everything else. Three, the seven city rollout. Dallas, Houston, Phoenix, Miami, Orlando, Tampa, and Las Vegas. They were all promised by the end of June. Hit five of those. Huge. Only two or three raises flags. Four. The EU approvals cascade. Netherlands is done.

[00:17:43]Watch for Germany, France or Italy.

[00:17:45]Next. Second European country approves.

[00:17:48]The continent unlocks fast. Five.

[00:17:50]Cybercap production numbers. Q2 earnings in late July gives the first real number. Over a,000 units for the quarter is bullish. Over 5,000 is the signal that scale is real. If these five things keep moving forward, you're watching the most important infrastructure buildout in modern business happen in real time.

[00:18:09]Look, here's the thing I keep coming back to. Yes, 13 cars, 582 total fleet.

[00:18:16]Those are the numbers everyone sees. But the numbers that actually matter are different. It's software is ready.

[00:18:22]Operations are built. They're hiring in 34 cities. The hardware is rolling off the line. Regulations moving. That's the real scoreboard. a full list. And right now, Tesla's winning every line on it.

[00:18:34]If you found this useful, please hit subscribe. I'll keep tracking these four pillars and that hiring number every single month. See you in the next video.