Bait the hook! Fishing in the global talent pool

Global Talent Competition and Immigration Policy

• The economic argument for immigration: Immigrants contribute disproportionately to innovation, accounting for approximately 36% of U.S. innovation despite making up only 12% of the population, as shown in a Harvard study where immigrant departures reduced team productivity twice as much as native-born departures.
• Policy failures in major economies:
  • China: National security priorities create a hostile environment that limits foreign talent acquisition.
  • Britain: Recent government advice discourages hiring foreign engineers to create jobs for locals, a strategy economists argue will fail to increase employment.
  • United States: The immigration system is rated second-worst globally (after Iraq), with a 75% rejection rate for H-1B tech visas and a 134-year average wait time for green cards for applicants from India due to the 7% per-country cap.
• Successful models:
  • United Arab Emirates: Offers one-week visas for high earners, "golden visas" allowing job mobility and family sponsorship, and rapid bureaucratic processing (ID, banking) within a week of arrival.
  • Portugal: Transformed its status as a digital nomad hub in a decade; if all interested graduates could move there, the graduate population would have risen 140% compared to a 1% rise a decade ago.
• Future projections: If university graduates had free movement, the U.S. graduate population would rise by 7%, while Canada, Australia, and Switzerland would see rises of 2.5%, and New Zealand would quadruple its population.
• Donald Trump's proposal: Suggested automatically granting green cards to anyone graduating from an American university, though the feasibility and intent remain unclear.

The Economics of "Tipflation" in the United States

• Current trends: The standard tip has shifted from 10% in the 1950s to 15% in the 1980s–90s, with 20% now serving as the expected default for most service interactions.
• Structural drivers: Unlike other nations, U.S. restaurants legally pay a "tipped minimum wage" (e.g., $2.50 hourly) rather than a standard living wage, making tips a mandatory component of employee salary rather than a voluntary reward.
• Political and fiscal proposals: Both Donald Trump and Kamala Harris have proposed exempting tips from taxation; while politically popular with the Hispanic food industry workforce, economists estimate this could increase tax losses from $10 billion to $500 billion over a decade and accelerate tip inflation.
• Regulatory shifts toward "peak tipflation": Several states are moving to eliminate the tipped minimum wage, forcing restaurants to raise upfront menu prices; this is expected to gradually reduce the percentage of tips while increasing the total cost of service.
• Consumer sentiment: Surveys indicate approximately one-third of Americans are dissatisfied with the current tipping culture, which has expanded beyond restaurants to include baristas, bellhops, and service providers.

Regulatory Divergence on Facial Recognition AI

• The EU approach: The EU AI Act strictly bans live public facial recognition by police to protect freedom of association and protest, citing risks of retrospective identification and persecution.
• The U.S. approach: No federal ban exists; technology is used widely by law enforcement, though restricted in specific jurisdictions (e.g., San Francisco) and allowed in federal zones (airports).
• Global misuse cases:
  • Russia: Post-funeral identification of dissidents at Alexei Navalny's funeral using apps like "FindFace" (70% accuracy on metro passengers) demonstrated the tool's use for political suppression.
  • China: The country leads in deployment for surveillance, including cash withdrawals and airport check-ins, with systems allegedly targeting specific ethnic groups like the Uyghurs.
• Algorithmic bias: Systems trained on non-diverse datasets (e.g., predominantly white faces) show significantly lower accuracy for minorities, leading to wrongful arrests and historical misidentifications (e.g., Google Photos labeling Black individuals as gorillas).
• Specific U.S. controversy: The startup "Clearview AI," which scraped billions of social media images, has been banned in several countries but remains in use by U.S. law enforcement despite lawsuits and wrongful arrests linked to its inaccuracy.

The Economist's AI Schools Briefs

• Content focus: The publication's summer series provides beginner guides to complex AI topics, with an episode specifically addressing listener questions on facial recognition and law enforcement.
• Expert consensus: Experts agree that while existential risks (e.g., super-intelligent computers) are distant concerns, the immediate dangers of facial recognition involve privacy erosion, the chilling of protest movements, and systemic bias.