Beyond Software: Why the Future of AI Is Physical

From Software to the Physical World

“First gradually, then suddenly.”

Every major platform shift feels slow and almost invisible—until the moment it doesn’t. Just as search engines reshaped how we find information, social media redefined how we connect, smartphones put the internet in our pockets, and the sharing economy reimagined how we use assets, we are now standing at the edge of another transformation: the rise of Physical AI.

I spent the summer in San Francisco—the global testing ground for many of these technologies. What struck me was how Physical AI is no longer theoretical—it’s visible on the streets. Waymo vehicles cruising through neighborhoods.

At first glance, they still seemed like novelties—slightly clunky, cautious, even awkward. But as the summer went on, I began noticing something more profound: the shift from novelty to normalcy. Families waiting for a Waymo instead of an Uber. A 10-year-old hopping into a self-driving car without hesitation. Shorter wait times, and cheaper fares than a Lyft or Uber. Signs of trust and adoption building in real time.

What I learned that summer is simple but powerful: the future of AI is physical. It’s not just code, screens, or algorithms hidden in the cloud. It’s cars, roads, buildings, logistics, and infrastructure. It’s the physical economy—and it’s entering its own “internet moment,” following the same trajectory: first gradually, then suddenly.

The Gradual Build-Up — Early Days in San Francisco

In the summer of 2023, San Francisco felt like a living laboratory for the future. For the first time, I saw autonomous vehicles not as demos or headlines, but as part of the city’s daily rhythm.

At first, they stood out—Waymo cars crawling cautiously through intersections, sometimes hesitating when a dog darted across the street or when a crowd of pedestrians spilled into the crosswalk. They were impressive but also clunky—like watching a gifted student overthink every move on their first day in class.

By early 2024, the technology had matured. The cars were smoother, more decisive. They merged into heavy traffic with greater confidence, handled nuanced situations at complex intersections, and navigated San Francisco’s unpredictable urban choreography with fewer hiccups. Still, they were rare—an occasional sight that made people pause and watch, rather than a trusted, everyday option.

Living through that summer, I was reminded of the early internet era—dial-up modems, buffering videos, clunky websites. You could sense the potential, but the limitations were impossible to ignore. Yet beneath that surface awkwardness, the foundations of a new platform were being laid.

San Francisco is currently showing the world what the early days of Physical AI look like: messy, fascinating, and quietly world-changing.

The Inflection Point — From Novelty to Mainstream

By July 2025, San Francisco felt different. The novelty of spotting a Waymo had worn off—not because the cars disappeared, but because they became too common to notice.

What struck me most was how quickly the value proposition flipped. A ride that cost around $30 on Uber was now $12 on Waymo. Wait times that could stretch 10–15 minutes on Uber or Lyft were now often just 3–5 minutes for a Waymo. Cost, speed, and reliability had shifted in favor of the machine.

Trust was shifting too. Parents were comfortable letting their kids ride in an autonomous car. One afternoon, I saw a 10-year-old hop into a Waymo without hesitation, as if it were no different from the family SUV. That moment stuck with me—because it wasn’t just about technology, it was about culture catching up to capability.

Of course, there are limits. Waymo still doesn’t operate on highways. Runs to the airport are in active testing, with human safety drivers gathering data, but they aren’t yet open to the public. Still, even with those caveats, the arc of progress is unmistakable: what was once awkward and clunky is now cheaper, safer, and faster than the incumbents.

This was the inflection point—the moment where Physical AI shifted from a curiosity to a competitor, from experiment to expectation. And as with every past platform shift—whether the rise of search, social media, or smartphones—once the curve bends upward, adoption accelerates first gradually, then suddenly.

Expansion & Competition — The New AI Platforms

The spread of Physical AI is no longer confined to San Francisco. As of this writing, Waymo has officially deployed in Austin, Texas, with Houston serving as an active testing ground. Each new city is both a proof point and a data engine—training these systems in different traffic patterns, infrastructure quirks, and cultural driving behaviors.

But Waymo is not alone. Uber, the original symbol of the sharing economy, is now racing to reinvent itself in the age of autonomy. In mid-2025, Uber announced a partnership with Nuro and Lucid Motors to launch a fleet of 20,000 autonomous EVs, powered by Nuro’s Level-4 self-driving platform. Rollout is expected to begin in 2026, with plans for rapid scale across U.S. cities.

Tesla, never content to sit on the sidelines, is still pushing toward its long-promised robotaxi network, leveraging its vast real-world driving dataset and vertically integrated fleet.

What’s unfolding feels familiar: a platform war. Just as search engines fought for dominance in the early 2000s, or smartphone operating systems battled for global adoption in the 2010s, Physical AI is entering its competitive phase. The difference now is that the stakes are physical, not digital: cars, roads, logistics networks, and urban infrastructure.

For entrepreneurs, this raises both caution and opportunity. Competing with giants is daunting, but every platform shift opens new niches—the “picks and shovels” businesses, integrations, and vertical applications that emerge when infrastructure changes. This is exactly the kind of dynamic many of my podcast guests have highlighted: where big platforms fight, opportunities for founders multiply.

Beyond Mobility — Physical AI Across Industries

While robotaxis are the most visible expression of Physical AI, they’re only the tip of the iceberg. The same core technologies—computer vision, robotics, reinforcement learning, and real-time decision-making—are beginning to transform industries far beyond transportation.

In real estate, buildings are becoming intelligent assets. AI systems are managing energy usage, adjusting climate and lighting dynamically based on occupancy, and predicting maintenance needs before they escalate into costly repairs. For developers, this isn’t just a sustainability play—it’s a business model shift, where buildings can operate more efficiently and command higher value.

In logistics and supply chain, the potential is even larger. Autonomous delivery fleets are being piloted for groceries and last-mile fulfillment. AI-driven warehouses are orchestrating human and robotic labor with precision that reduces errors and shortens turnaround times. On the horizon, self-driving freight trucks promise to unlock 24/7 long-haul transport, reshaping both costs and capacity in global trade.

Cities themselves are starting to adapt. The smart city model is no longer just about sensors—it’s about AI coordination at scale. Traffic signals dynamically adjust to optimize flow for human and autonomous vehicles. Public transit systems are experimenting with AI-driven routing. Over time, urban planning will evolve not just around people and cars, but around autonomous systems as citizens of the street.

These are the real-world applications of AI that will define the next decade. And just as the rise of smartphones gave entrepreneurs new playgrounds in apps, payments, and media, Physical AI will create entirely new categories of businesses in logistics, urban tech, property management, and beyond.

For entrepreneurs and investors, the message is clear: don’t think of Physical AI as “just cars.” Think of it as the infrastructure layer of the physical economy.

Historical Parallels — Lessons for Entrepreneurs

Every platform shift feels chaotic in the moment, but history leaves a clear pattern: novelty, trust, adoption, ubiquity.

When search engines first emerged, they were clunky directories with inconsistent results. Once Google cracked the model, search became the gateway to the internet—and hundreds of businesses (from SEO to e-commerce) were built on top of it.

Social media began as niche communities for students and hobbyists. Then Facebook opened to the world, Twitter (now X) became a global conversation engine, and Instagram turned into a visual marketplace. Entire industries in marketing, media, and brand-building realigned around these networks.

Smartphones were initially a luxury for tech enthusiasts. Then the iPhone in 2007 shifted the paradigm, unleashing an app economy that created everything from ride-hailing to food delivery to fintech. Founders who recognized that new distribution channel early built billion-dollar companies.

The sharing economy sounded strange at first—renting your home to strangers, getting in someone else’s car. But trust mechanisms (reviews, insurance, payments) created enough confidence to scale, and today Airbnb and Uber are verbs.

Physical AI is now at this same inflection point.

Autonomous vehicles may feel clunky today, like early search engines.
Trust is building, just as it did with social networks.
The infrastructure (smart cities, 5G, AI chips) is emerging, like the smartphone era.
And skepticism is giving way to adoption, just like the sharing economy.

For entrepreneurs, the lesson is simple: watch the patterns. Every one of these shifts created new categories of companies—not just the platforms themselves, but the tools, services, and integrations surrounding them. The winners weren’t always the ones building the infrastructure. Many were the ones who saw what the infrastructure enabled.

Physical AI is the next infrastructure layer. The real question is: what will you build on top of it?

Opportunities & Risks — Building in the Physical AI Era

Every platform shift creates both massive opportunities and new challenges. Physical AI will be no different.

🚀 Opportunities for Builders and Founders

New infrastructure layers: Just as AWS and cloud platforms unlocked SaaS, Physical AI will demand tools for routing, coordination, payments, and compliance in the physical world. Founders who build these enabling layers will capture enormous value.
Vertical applications: From autonomous delivery logistics to AI-managed real estate, Physical AI will create dozens of vertical markets. Early entrants can carve out niches before incumbents adapt.
Efficiency + safety gains: By reducing human error, downtime, and inefficiency, Physical AI unlocks significant cost savings—opportunities entrepreneurs can capture and scale.
The picks-and-shovels strategy: Just as the app stores created room for developers, the Physical AI ecosystem will need startups that handle trust, security, UX, and edge cases.

⚠️ Risks and Challenges to Navigate

Job displacement: Millions of professional drivers—across ride-hailing, trucking, and delivery—face disruption. Entrepreneurs will need to balance innovation with reskilling, workforce transitions, and social impact.
Public trust: A single accident involving an autonomous vehicle can dominate headlines and slow adoption. Building trust mechanisms (transparency, insurance, oversight) will be just as important as the technology itself.
Regulation and policy: Governments move slower than technology. Founders entering Physical AI markets must be prepared for a patchwork of rules, approvals, and delays. Navigating this is not just legal—it’s strategic.
Concentration of power: Just as a few companies dominate search, social, and cloud, Physical AI may consolidate around a handful of players. For startups, survival may depend on finding the right niches or building symbiotically with the giants.

⚖️ The Dual Reality

The stakes are high. On one hand, Physical AI could make cities safer, cleaner, and more efficient, creating massive entrepreneurial wealth along the way. On the other hand, it could deepen inequality and concentrate power if adoption is mismanaged.

For entrepreneurs, this paradox is the signal: where there are risks, there are also opportunities. The winners of the Physical AI era will be those who not only build great products but also navigate trust, regulation, and impact with foresight.

Conclusion & Call to Action

The story of Physical AI is unfolding in real time. What I saw this summer in San Francisco—autonomous vehicles cruising neighborhoods, shorter wait times than Uber/Lyft, and families treating robotaxis as routine—signals a deeper truth: the future of AI isn’t just software. It’s physical.

As with search, social, smartphones, and the sharing economy, adoption will feel gradual—until the moment it isn’t.

And if you’ve read this far—congratulations. It signals that you’re already ahead of the curve and genuinely curious about where AI, technology, and entrepreneurship are headed.

That curiosity is exactly what these upcoming webinars are designed to serve.

I’m extending these conversations beyond the podcast into a new webinar series. In this format, past podcast guests will return with 10–15 minute slide-deck presentations that go deeper and with more detail than a podcast conversation allows. Think of it as an “extended cut”: frameworks, hard-earned lessons, and predictions, delivered directly by the founders and technologists building the future.

For you, it’s a chance to get greater exposure to these leaders and their ideas—an inside look at what they’re learning as they scale companies, test in the market, and navigate the shift to the Physical AI era.

👉 Want an early invite? Just reply to this email with “Webinar” and I’ll add you to the priority list for the first priority recording sessions.

You’ll also keep getting these deep dives here in the newsletter, plus updates as new speakers confirm.

The future of AI is physical—and it’s arriving faster than most people realize.

—
P.S. If there’s a specific founder or topic you’d love to see in this format, just reply and let me know.

Warm regards,

Chris