From screens to streets

Every week, I message 30+ people on LinkedIn who work in fleet. My goal is simple: meet new folks, learn from their perspective, and stay close to what's actually happening in the industry. Some conversations are forgettable. Others stick with me for days.

Two weeks ago I had one of those. We're all drowning in buzzwords right now: AI agents, predictive analytics, autonomous everything. And yes, these matter A LOT as it changes everything. But sometimes a conversation pulls you back to what actually moves the needle today.

I met someone who comes from a multi-generational tire and fleet business dating back to 1938, when his grandfather became the first independent Firestone dealer in the Americas. He's managed 66,000 vehicles across 80 countries. He's seen it all: self-driving trucks, fleet electrification, wireless charging pilots, driver behavior analytics before anyone called it "AI."

A few things from our conversation are still rattling around in my head:

But here's the reminder we all need: When you pitch a fleet, the finance person needs to see "green dollars" real, measurable savings. "Blue dollars" the intangible benefits like driver satisfaction or operational flexibility matter too. But somewhere along the way, we need to translate them into green. Cost per mile. That's usually the KPI. Everything else flows from there.

In 2026, with all the noise about what's next, this felt like a grounding moment. The fundamentals haven't changed what is changed is how we get there.

What is Physical AI and why are people suddenly obsessed with it? For me it’s because most of the innovation we've celebrated for the last 20 years lived inside screens software, feeds, cloud, clicks when we were imagining underground cities, exploring the ocean, robots everywhere. Physical AI is the shift from intelligence that talks to you to intelligence that moves things in the real world: robots, factories, supply chains, autonomous vehicles, long haul, cities. Two stories landed in the same week that made this feel very real. Waabi raised up to $1B not just another round, but capital and distribution betting that autonomy is becoming a repeatable deployment business, not a forever science project. Uber committed $250M of that to exclusively bring Waabi-powered robotaxis to its platform, at least 25,000 autonomous vehicles. One AI brain powering both trucks and robotaxis. If they execute, it means better utilization, lower cost per mile, and a new operating layer fleets will need to master: charging, maintenance, incident workflows, insurance, payments, compliance. Raquel Urtasun is one of those rare leaders proving academia and company building can reinforce each other when the ambition is big enough.

Then, the same week, Tesla announced it's ending production of the Model S and Model X and Tesla converting those Fremont factory lines into a 1-million-unit-per-year Optimus robot production line. No replacements planned. The company that proved EVs could go mainstream is now saying the bigger opportunity isn't vehicles at all, it's humanoid robots. Tesla is betting $20B in capex this year on robots and autonomous vehicles, including the steering-wheel-free Cybercab. For fleet, read both signals together: one company is pouring a billion into making autonomous trucking a deployable business. Another is literally choosing robots over cars. The industry is moving from "how do we manage drivers better" to "what happens when there's no driver to manage." We're not there yet. But for me the signals are getting louder.