The real AI revolution isn’t software. It’s farms, mines, and trucks. | Qasar Younis

Applied Intuition is a $15 billion company that adds AI to physical vehicles—cars, tractors, mining rigs, planes, submarines, and construction equipment—making them semi-autonomous or fully autonomous. Co-founder and CEO Qasar Younis has spent a decade building the company largely in silence, and today 18 of the top 20 automakers are customers, along with major construction, mining, trucking companies, and the Department of Defense. The company is essentially “Waymo or Tesla without the hardware”—they provide the intelligence layer, not the vehicles themselves.

The real AI revolution in the next 5–10 years will be physical, not software. While most attention goes to coding assistants and chatbots, Younis argues the biggest impact will be in farming, mining, construction, and trucking—industries where autonomy addresses urgent labor shortages and safety crises. The average American farmer is 58 years old. Commercial trucking jobs go unfilled not because people don’t exist, but because the trade-offs (weeks away from family) are no longer acceptable when alternatives like Uber and DoorDash exist. These are jobs where people die, and nobody is clamoring to fill them.

AI is coming “just in time” to save us from demographic decline. Populations are aging and shrinking in key economies. Rather than stealing jobs, physical AI fills gaps that humans are already abandoning. Younis frames this as the next Industrial Revolution: painful transitions in the short term, but a massive net reduction in suffering and increase in abundance over decades—just as electrification, heating, and communication access eventually became universal.

Anxiety about AI stems from misunderstanding, and the antidote is education. Younis argues that fear of AI is rooted in not understanding the technology’s limitations. People see a nunchuck-wielding humanoid robot (which costs $15 million and is pre-programmed) and imagine sentience, but they don’t fear car factory welding robots that have been doing advanced autonomous work for 25 years. The gap between what something looks like and how it actually works creates irrational fear. His advice: spend time learning the technology, see its edges, and then actively steer it toward good uses.

The recent AI stock sell-off reflects hedge fund misunderstanding, not societal risk. Younis separates two distinct anxieties: individual fear of AI’s impact on society, and public market sell-offs driven by investors. The sell-off happened because hedge fund managers hired AI consultancies to replicate products like Figma in weeks, concluded that AI makes billion-dollar software companies obsolete, and priced in that risk. But these investors don’t understand the depth, integrations, and complexity of the real products. The sell-off is a market misread, not a signal that AI will destroy these companies.

Comparing Chinese AI companies to American ones is a category error. Chinese firms like Huawei are not profit-driven companies competing in free markets—they are effectively extensions of the state, with a quarter of their workforce being Communist Party members. When Americans compare Huawei to Apple or DeepSeek to OpenAI, they’re comparing a company to a government. Chinese EVs look amazing in part because they don’t need to be profitable; an American equivalent like Rivian loses money and gets punished by investors. If America combined all its EV efforts into one state-backed entity unconcerned by profits, it could field equally impressive products. The competition is real, but it’s not apples-to-apples.

Self-driving cars are already far safer than human drivers, and the moral case for autonomy is overwhelming. Over 30,000 people die annually in U.S. car accidents. Younis argues that in 25–30 years, humans driving cars will look as morally indefensible as child labor does today. Self-driving technology, whether Tesla’s cheaper camera-based L2++ approach or Waymo’s sensor-heavy L4 approach, will become ubiquitous globally within 5–7 years. As it becomes standard in every car, injuries and deaths will drop, and applications we can’t yet imagine will emerge—just as Instagram was inconceivable before the iPhone’s app store, cameras, and social network adoption converged.

Physical AI will follow the mobile revolution’s playbook: intelligence added to existing machines first, new form factors later. The near-term wins come from putting AI into machines that already exist—mining vehicles, tractors, trucks—because the engineering for those machines is already mature. You’re just adding intelligence. Humanoid robots capture our imagination because we’re primates, but pragmatically, the biggest impact comes from making the machines around us smarter. Within 5–7 years, every new car will have some level of autonomy, and that ubiquity will unlock new applications the way CarPlay and Android Auto did once navigation became standard.

Applied Intuition’s core values are radical pragmatism, speed, and doing the best work alone and quietly. Younis intentionally stayed under the radar for a decade, inspired more by Berkshire Hathaway than Silicon Valley darlings. He believes every minute spent on podcasts, tweets, or public promotion is a minute not spent on customers and product. The company’s values include “speed above everything,” “never disappoint the customer,” “technical mastery,” “high output matters,” “half the work is follow-up,” and “laugh a lot.” These aren’t abstract—they’re used to assess, compensate, and promote managers.

The company cleans its own office and has never spent a dollar of raised capital. Applied Intuition runs a weekly “cleaning zen” where every employee cleans their own area, inspired by Japanese schools. Younis draws a direct line between this ethos of maintenance, humility, and craft and the company’s financial discipline: despite being nearly 10 years old with over 1,000 engineers, they’ve never spent any of the capital they’ve raised. The company is self-sustaining. He connects this to automotive engineering, which is fundamentally an exercise in quality, safety, and maintenance at scale—making a car every 30 seconds, extremely cheap, globally competitive.

Successful companies show traction early; if the market isn’t informing a specific path after two years, reset. Drawing on his experience as YC COO, Younis observes that good companies tend to have early traction and sustain it for a decade. If after two years the market isn’t giving you increasingly specific signals about what to build, the foundation may be wrong—co-founders, market choice, or life circumstances. He encourages founders to treat their first startup as practice: “The first table you built was wobbly. You wouldn’t go work at Crate & Barrel—you’d keep building tables.” Being a founder is a muscle, and it’s not random that his third company is his most successful.

Great founders develop taste through broad exposure to history, society, and unfamiliar domains. Younis reads voraciously and deliberately—old books, not new ones, because time filters out noise. He recommends reading outside your expertise: if you know nothing about Roman history, find the best book on it. He believes reading Malcolm X’s autobiography makes you a better founder, not because of any direct lesson, but because diverse inputs make your understanding of the world richer and more nuanced. He also emphasizes the importance of having been an employee at the bottom of large organizations (he worked at GM and Bosch) to understand bureaucracy, bad tools, and what it feels like to be powerless—experience that informs better leadership and policy decisions.

The best ideas win when companies create cultures that surface dissent and remove emotion from decisions. Younis operationalizes open debate by encouraging every person in a room—regardless of seniority—to speak up, especially if they disagree. He references Google’s failure to compete with Facebook: Google was the apex predator of Silicon Valley, but its momentum and culture were so oriented toward search that it couldn’t pivot to social media. The “wall of sound” from internal momentum drowned out signals that the market was changing. At Applied Intuition, the goal is to find the best idea regardless of where it comes from, then commit to it decisively. He uses a heuristic: if multiple people independently make the same decision, you’ve successfully removed emotional filters.

Decisiveness matters as much as openness—and founders must be right, not just visionary. Younis holds two conflicting ideas in tension: be humble enough to listen, then be confident enough to act. Speed is a core value, specifically worded “Move fast, move safe.” Once a decision is made, leaders must commit fully. But he emphasizes that founders love to take credit, and the real test is not having a vision—it’s being right. The evidence is whether the company becomes a sustainable standalone business. Many founders fail not because they lacked ambition, but because they were 5 degrees off course and too proud to correct.

Most Silicon Valley CEOs lack taste because their life experience is too narrow. Younis observes that many founders grow up in Cupertino, go to Berkeley, start a company immediately, and never experience being an employee, working in different industries, or living in different cultures. Taste—the ability to discern what’s good—comes from broad exposure to humans and life. He believes someone who backpacked the world for years will be a better founder, not because of any specific skill, but because they understand people and can exercise judgment. He’s careful to note he doesn’t claim to have great taste himself—he’s part of the same system he’s critiquing.