Palmer Luckey says U.S. colleges are “hollowing out” engineering for China-trained makers

Palmer Luckey does not frame the U.S.-China AI fight as a pure chip-and-code contest. In conversation earlier this year with the Hoover Institution, the Anduril founder argued the real battle is who is training the world's best students, and he claims American companies have helped create the imbalance by “hollowed out” real engineering capacity.

Luckey’s core accusation is blunt: American companies “feed these colleges a whole bill of goods on what they should be teaching people.” He said the result is a mismatch between what universities teach and what companies actually need. “Basically, we're not teaching engineers how to be engineers anymore,” Luckey said. In his telling, China is gaining because it builds a deeper technical bench, including “battery engineers, metallurgists, and optical engineers,” while the U.S. has shifted toward training people he calls “architecture astronauts.”

What Luckey is describing is not just a classroom problem. It is an execution problem that shows up on product roadmaps, supply chains, and who holds the engineering muscle when something has to be built, manufactured, and iterated under real constraints. Luckey argued the U.S. is teaching “designers how to be high-level design shops” that package designs and hand them off to “the real engineers in China” who then figure out how to do the work. The point is not that American companies are incapable. It is that the distribution of responsibilities across borders has changed, and the training pipeline has not kept up.

He used Apple, No. 4 on the Fortune 500, as an example of the broader shift. Luckey said, “Apple used to have to figure out how to actually make their stuff,” but “these days, most of the really hard work is being done by Chinese engineers.” The source details the structural reason this claim matters: Apple designs its products in Cupertino, California, while relying heavily on manufacturing partners in China, with supply chain engineering and manufacturing expertise increasingly shifted overseas. If you are an executive tracking time-to-market or resilience, the relevant question becomes: if hard problem-solving lives abroad, how quickly can your company compress cycles when the situation demands speed?

Luckey also credits the American system for a different kind of advantage: cultivating entrepreneurs willing to pursue unconventional ideas. He contrasted that with China’s output, saying, “[China’s] educational system, it doesn't generate very many queen bees, and it generates a lot of worker bees.” That is a cultural and incentive argument, not an engineering one. And it connects to his own path: Luckey became a billionaire after building prototypes for virtual reality headsets while homeschooled in California, then enrolled as a journalism major at California State University, Long Beach, before dropping out at 19 to focus on Oculus, the virtual reality startup he had been developing as a teenager.

Luckey’s career arc is unusually “U.S.-style” in the way it moves from unorthodox beginnings to outsized capital. He said he was “a 19-year-old kid working a minimum-wage job with no college degree,” living in a 19-foot camper trailer, and that Peter Thiel gave him a million dollars when nobody else would to start Oculus. Luckey sold Oculus to Facebook in 2014 for a deal worth $2 billion, then founded Anduril three years later. Fortune notes he is now worth $5 billion and that Anduril is valued at $61 billion; Fortune also reached out to Anduril for further comment.

If this sounds like a single founder’s view, it is at least echoed by executives outside defense tech. Earlier this year, Pfizer CEO Albert Bourla warned at a Council on Foreign Relations event that Western universities risk falling behind as China accelerates research output. Bourla pointed to the Nature Index, which tracks institutional research output. In 2020, U.S. and European schools dominated the top 10, but today, nine spots are held by Chinese institutions. He said, “Right now, I think they are not at the same level as the U.S., but they are very close,” and added, “But the rate with which they go up predicts that they will be better than us within the end of this decade.”

The “why” behind these warnings is policy. The source says China reportedly eliminated or suspended about 12,200 undergraduate degree programs between 2021 and 2025, mainly in areas like the humanities, foreign languages, and some management disciplines. During the same period, about 10,200 new programs were introduced, largely aligned with industrial priorities such as AI, robotics, and semiconductor engineering, citing Ministry of Education data referenced by Forbes. The push begins even earlier than college: Beijing’s primary and secondary schools offer AI instruction each academic year, exposing students to topics ranging from chatbot use to AI ethics. Donna Morris, Walmart’s chief people officer, previously told Fortune, “Let’s look at China... Five-year-olds are learning DeepSeek, and that says a lot about how they believe in capability building.”

For executives, boards, and investors, the second-order risk is simple: if training pipelines diverge, the talent base that can actually close the gap in AI and advanced manufacturing starts compounding in China’s favor. Luckey’s “hollowing out” argument suggests companies could unintentionally finance the very capability shift that leaves them dependent on external engineering talent. Meanwhile, Bourla’s research warning points to a longer arc, where institutional output growth could translate into more of the discoveries and trained researchers that feed future products. If you run an organization in defense, industrial tech, semiconductors, or any AI-adjacent category, this story is a reminder that competitiveness is built twice: once in factories and labs, and again in the curricula that decide who becomes a builder.