A tea chest from 250 years ago, a Star Trek communicator, and an Apple Newton made the point fast: the future rarely arrives in a clean, polished form. At a Forbes panel on the next 25 years of American innovation, John Werner used those objects to connect rebellion, science fiction, and failed early tech into one idea.
If you’re trying to figure out where the next big markets may come from, that framing matters. The most important shifts often look awkward at first, and the biggest story is rarely one gadget or one app.
Why this conversation felt bigger than another AI panel
Werner’s opening props were more than stage theater. The Boston Tea Party artifact pointed back to an earlier American turning point. The Star Trek communicator pointed ahead to a future that once felt remote and now looks oddly familiar. The Apple Newton sat in the middle as a reminder that first-generation products often miss badly before later versions change daily life.
That Newton example landed because it wasn’t only about product design. Werner called it one of the first consumer devices to use AI, and the AI failed in public. It couldn’t read handwriting well enough. At the time, that weakness made the product feel clumsy. Years later, the smartphone absorbed many of the same ambitions and turned them into ordinary behavior.
The panel’s three guests each worked on a different piece of that same pattern.
Here’s the simplest way to read their views:
| Speaker | Focus | Core argument |
|---|---|---|
| Alvin Wang Graylin | AI, policy, US-China tech | America still leads in invention, but it risks weakening the systems that made that possible |
| Daniel McGill | Joby Aviation, eVTOL | Air mobility could change commuting, housing, and the shape of cities |
| Cinnamon Sipper | Godella, physical AI | AI trained on text isn’t enough if you want machines to understand the physical world |
Put together, their arguments painted a more grounded picture of the next tech cycle. This wasn’t a parade of slogans about AI magic. It was a conversation about education, immigration, energy, transportation, engineering tools, and the human habit of outsourcing too much thinking to machines.
By the end, Werner’s larger point was hard to miss. Some things that looked like science fiction not long ago are already stepping into the real economy, and they may arrive faster than most people expect.
Alvin Wang Graylin on America’s edge, China’s strength, and AI discipline
Graylin’s view of innovation started with a blunt contrast. He said the United States still gets one big thing right: it remains the source of major technology innovation across several decades. Yet he also argued that America is now putting that advantage at risk by weakening education, reducing research support, and damaging the immigration channels that bring top talent into the country.
His comparison with China was just as clear. In Graylin’s view, China is strong at diffusion. It takes technology invented anywhere and makes it accessible, affordable, and widely deployed, both at home and abroad. What China doesn’t do as well, he said, is build a positive enough relationship with the United States to support more shared development and trust.
That argument fits a broader pattern in McKinsey’s technology trends outlook 2025, which treats AI as part of a wider shift that includes robotics, autonomous systems, and responsible deployment. In other words, the race isn’t only about who invents first. It’s also about who educates well, builds well, and spreads tools across society.
Graylin also pushed back on the tendency to obsess over whatever trend is loudest. He said most things are overhyped, but one area still feels underplayed: bio and personal medicine. His logic was simple. If health fails, every other advance matters less.
His strongest point, though, was about human agency. Graylin warned that people are starting to hand over too much decision-making to AI, especially students who go straight to a chatbot for homework answers instead of working through the problem first.
He used the story of Odysseus and the sirens to explain what discipline should look like. Odysseus didn’t avoid the song. He listened while tied to the mast, so he could hear without surrendering judgment.
“First, come up with your own answer, then ask AI what you’re missing.”
That habit may sound small, yet it gets at a bigger fear under the current boom. A tool that sharpens thinking is helpful. A tool that replaces it too early can leave people less capable than before.
Daniel McGill on why VTOL could redraw the map of a city
McGill’s part of the conversation brought the future down to street level, even though his aircraft fly above it. He explained VTOL in plain terms: vertical takeoff and landing. In Joby’s case, the more exact label is electric VTOL, or eVTOL. These aircraft take off like helicopters, then shift into wing-borne flight, which makes them more efficient in the air.
He argued that this new class of aircraft could do more than shorten a few premium commutes. It could widen the practical footprint of a city. If people can travel farther in less time, then expensive urban cores no longer hold the same grip over where people live and work.
McGill used the Bay Area as a concrete example. A trip from the East Bay to Santa Cruz can take 90 minutes to two hours by car. By air, he said, that same route could take about 15 minutes. The value there isn’t only speed. It’s time returned to people who now lose large chunks of the day to road traffic.
He tied that promise to a historical parallel. After World War II, the interstate highway system opened huge amounts of land to new communities and businesses. McGill sees eVTOL as a smaller but similar shift, one that may help relieve housing pressure by changing what counts as a manageable commute.
Still, aircraft alone won’t do it. McGill kept coming back to infrastructure. He said the biggest smart bet right now is investment in vertiports, power access, and charging systems. His shorthand was memorable: large Teslas in the sky. If those systems don’t exist, the aircraft stay a demo instead of becoming a transport network.
He also stressed that the first wave will be piloted, not autonomous. That matters because safety, certification, and public trust come before scale. McGill said Joby has seen support from the FAA and the Department of Transportation, and he pointed to the federal eVTOL Integration Pilot Program as a sign that public agencies want the US to lead this category. He also said states and manufacturers could use those pilot programs to begin pre-certification operations in places such as Texas, Florida, and New York.
His last point pushed beyond aircraft. McGill said the long-term opportunity is multimodal. Ground-based autonomous vehicles, urban hubs, airports, resort links, and eVTOL routes could work as one system instead of separate modes competing for attention.
Cinnamon Sipper on physical AI and the bottleneck in engineering
Sipper focused on a gap that many AI discussions skip. Today’s best-known models are trained on text and language. They can summarize, generate, and answer. But, as she put it, AI trained on text data can’t build the physical world.
That problem gets real fast when you move past chat and into science or engineering. A useful physical model has to predict how heat moves, how fluids flow, and how structures break. Those behaviors are the grammar of physics. Without them, AI can talk about reality without being able to reason through it.
Sipper argued that this is why physical AI matters. If machine intelligence can reason from first principles about matter, energy, and dynamics, then a new class of work speeds up. She pointed to hard domains such as protein folding, turbulence modeling, and energy systems, all of which depend on accurate physical behavior rather than word prediction.
Werner compared that ambition to AlphaFold’s impact on life sciences, and the parallel makes sense. A milestone in physics-based reasoning wouldn’t be flashy in the same way as a consumer chatbot. Yet it could change how products get designed, tested, and improved.
Sipper gave a practical example. Every hardware product, whether it’s a phone, a battery system, or an industrial machine, depends on engineers answering the same kinds of questions. She said a physics-aware AI system could handle that reasoning thousands of times faster than tools available today.
“12 months of engineering in 12 minutes.”
That slogan from Godella captures the ambition. The target is not novelty for its own sake. The target is a shorter path between idea, simulation, test, and revision.
Sipper said the biggest gains would show up where experimentation and R&D are expensive, slow, or limited to a small set of specialists. If those barriers fall, the effect on daily life could be plain: cheaper energy, faster cures, and safer products. Most people would never see the model behind it. They would feel the result in price, speed, and reliability.
The trillion-dollar debate, and the personal bets behind it
When Werner asked where the next trillion-dollar opportunity might come from, Graylin gave the most useful answer of the day by resisting the premise. He said the future is “AI plus everything.” In his view, AI will blend into nearly every sector. Yet he also warned that chasing the next trillion-dollar company can pull attention away from the harder question: what value does a technology create for society?
That matters because tech often acts as a deflationary force. Better tools can make goods and services cheaper. If that happens, the social value can rise even when the headline market doesn’t map neatly onto one giant company.
A recent Dallas Fed analysis on AI, productivity, and living standards echoes that tension. Higher productivity can lift living standards over time. However, the road there can still be rough for workers whose tasks change faster than institutions do.
Graylin’s personal response to that risk was striking. He said he’s stepping away from commercial work to focus on policy. His concern is not abstract. America, he argued, is not prepared for the labor shift AI may trigger. The country struggled when blue-collar jobs moved or disappeared, and he thinks the next disruption may hit a much larger group. White-collar work touches a far larger share of the population, and he believes the US needs a stronger social safety net before that pressure peaks.
Sipper’s answer was more personal and more intimate. Outside company-building, she said she’s putting more of her free time into pursuits that AI is less likely to replace quickly, including piano, reading, and spending time with ideas. That choice wasn’t framed as retreat. It sounded more like balance. If tools keep climbing the ladder of abstraction, human curiosity and taste matter more, not less.
McGill’s biggest move was public-facing. He said part of his job now is showing people that the aircraft are real, and then building smart partnerships around that reality. He talked about working with autonomous vehicle companies and planning city hubs where ground transport and air transport meet.
Together, those answers gave the panel a sharper finish. The future isn’t only about new machines. It’s also about policy, habits, trust, and the daily systems that let new tools fit into ordinary life.
Final thoughts
The Apple Newton failed, but the idea behind it didn’t. The Star Trek communicator was fiction, until pieces of it became normal. That was the thread running through this conversation.
The next big tech markets may come from AI inside everything, short-hop electric flight, and software that understands physics instead of only language. Yet the harder message was even more important: invention alone won’t carry the future. Research, infrastructure, education, immigration, and human judgment still decide whether a breakthrough becomes part of everyday life.
