The most significant challenge with AI is the mass exodus of top researchers from universities and government to a few tech giants. This "hemorrhaging of talent" concentrates knowledge in the private sector, making it nearly impossible for the public to effectively govern or regulate the technology.
Industrial strategy is more effective when focused on solving big problems, like creating healthy school lunches or landing on the moon. This "mission-oriented" approach stimulates innovation across many sectors, unlike traditional policy that just hands subsidies to favored industries.
When governments outsource core functions like pandemic response planning to consultants, they don't just spend money; they prevent their own staff from developing crucial expertise. This creates a dependency cycle that "infantilizes" the state, weakening it over the long term.
The "you'll never get fired for buying IBM" principle is rampant in government. Officials hire prestigious consulting firms like McKinsey to gain political cover. If the project fails, they can deflect blame onto the consultants, effectively diffusing responsibility for their own decisions.
Economic benefits are not enough to win political support from communities that feel left behind. Policies fail when people don't feel valued or respected. To combat populism, governments must involve citizens in designing solutions that restore dignity, not just provide aid.
NASA spurred massive innovation by shifting from cost-plus contracts to "outcomes-oriented procurement." Instead of dictating specifications, they defined problems—like how astronauts would eat or use the bathroom in space—and challenged the private sector to invent solutions, leading to numerous commercial spin-offs.
Effective government requires more than just budget and staff ("capacity"). It needs "dynamic capabilities": the agility to pivot, collaborate effectively, and learn from experimentation. Most public sector reform misses this, focusing only on reactive, market-fixing roles rather than proactive, market-shaping ones.
Applying AI to a fundamentally flawed system like U.S. healthcare billing doesn't fix it. Instead, it creates an arms race where insurer bots fight hospital bots over claims. This only increases complexity and benefits the technology providers, while the core problems remain unsolved.
In the 1960s, a NASA procurement chief warned that relying on consultants would lead to capture by "brochuremanship"—where polished presentations replace substantive, in-house expertise. This accurately predicted today's problem of a government that can no longer write its own terms of reference, relying instead on consultant-driven PowerPoints.