LUNARTECH Fellows Blog
Context: LunarTech Fellows Research
Last Updated: February 2026
Walk into any tech conference today. Listen to the keynotes. Watch the CEOs speak. What do they talk about? They talk about AI models getting larger, automation replacing jobs, robotics taking over manufacturing, the future of work. What do they never talk about? Talent. Specifically: Where does great talent come from? Why are our universities producing fewer and fewer people who can actually build the things these leaders dream about? And what happens to a country when the pipeline breaks? These are uncomfortable questions. But they're the most important ones of our time.
Universities were built for a different era. They offer theoretical knowledge, research methodologies, academic credentials, four or more years of structured learning, and degrees that say you completed this program. Tech companies need something completely different. They need practical engineering skills, the ability to ship real products, experience with production systems, problem-solving under real constraints, and skills that generate revenue from day one. These are not the same thing.
Here's the uncomfortable truth: tech companies invest over $100K into training a single employee. They run internal bootcamps, mentorship programs, and continuous education. A junior engineer at Google or Meta might cost the company $200K–$500K in their first two years, between salary, benefits, training, and mentorship. Meanwhile, students go into debt to pay for their education. The average US student graduates with $30K in debt. In some countries, public universities are underfunded, and the quality shows. The tech employee gets paid to learn. The student pays to learn. No wonder the best talent flows toward tech.
As this gap widens, something predictable happens. Universities output a certain quality of graduate. Tech companies require a different quality of employee. The gap between these two grows wider every year. This is not about degrees being worthless. It is about a structural mismatch. Universities optimize for academic research output, publication counts, peer review, and theoretical advancement. Tech companies optimize for shipping products, user growth, revenue, and engineering excellence. These are opposite optimization targets.
When universities and industry diverge, graduates cannot find jobs. Not because there are not jobs, but because they do not have the skills that matter. In Spain, youth unemployment sits around 27% with underemployment over 40%. Italy sees 24% youth unemployment and 35% underemployment. Greece faces 35% youth unemployment and 45% underemployment. France has 17% youth unemployment and 25% underemployment. The global average sits around 13% youth unemployment and 20% underemployment. Many of these are not idle by choice. They are qualified on paper but cannot contribute to actual work.
And who pays for this mismatch? Governments do. France spends 25% of its government budget on social spending. Germany spends 22%. Sweden spends 28%. The United Kingdom spends 20%. The average across the European Union runs between 20–25%. A significant portion of this goes to unemployment benefits, housing assistance, and retraining programs, all symptoms of a talent pipeline that is not working.
When companies cannot find the talent they need, they do the rational thing. They automate. The shift from 2015 to 2025 shows a clear pattern: traditional hiring gives way to AI and ML engineering, 10 engineers per project becomes 2 engineers plus AI tools, growing headcount becomes growing efficiency, hiring more juniors becomes hiring fewer but paying more, and building teams becomes building systems. This is why you see headlines like AI will replace 300M jobs. It is not because AI is magical. It is because companies could not hire enough qualified people anyway, so they found another way.
This creates a vicious cycle. Universities produce graduates with quality mismatches. Graduates cannot find employment. Government pays for social welfare. Tax revenue goes to welfare instead of education. Universities get less funding. Quality declines further. And then the cycle repeats. Each year, the gap widens. Each year, more talent is wasted. Each year, countries become more dependent on importing talent or outsourcing work.
When talented people cannot find meaningful work at home, they leave. They go where the opportunities are. Romania has a diaspora of 4.2M people, with 18% of the population living abroad. Poland has 20M people in the diaspora, 12% of the population. Greece has 3.5M people abroad, 25%. Armenia has between 7–10M people in the diaspora, an astonishing 70%+. Albania has 1.8M abroad, 40%. These are not people who do not love their countries. They are people who could not find a place to use their skills. The brain drain is not a mystery. It is the logical result of a broken talent pipeline.
When domestic talent is not available, companies outsource. When outsourcing becomes normalized, something strange happens. The country stops building. You can outsource customer support, manufacturing, data entry, even software development. But you cannot outsource core engineering innovation, defense technology, strategic AI research, or national infrastructure. Every country that outsources its brain work eventually finds itself dependent on others for the things that actually matter. You can buy software from anywhere. But you cannot buy national security. You cannot buy economic sovereignty. You cannot buy the ability to shape your own future.
This is where most articles would give you tips. Learn to code. Get certifications. Network more. But that puts the burden on individuals, and individuals cannot solve systemic problems. What we need is a new model.
https://www.lunartech.ai/blog/the-lunartech-apprenticeship-programs-your-6-month-odyssey-from-aspiring-talent-to-ai-trailblazer
%402x.svg){kind=icon}