MIT Student Drops Out Over AI Extinction Fears: “I Had to Act”

Alice Blair, 18, arrived at the Massachusetts Institute of Technology last fall as a promising computer science freshman. By spring, she’d made a life-altering decision: to permanently withdraw from her studies. Her reason? Mounting fears that artificial general intelligence (AGI) could render her career and possibly humanity obsolete before she graduates.

“I couldn’t justify investing years into skills an AGI might eclipse overnight,” Blair told Tech Outlook. “The State Department report confirmed my worst fears.” She references a 2024 U.S. Department of State-commissioned analysis warning of “extinction-level threats” from uncontrolled AI development. The report suggests advanced systems could evade human control, triggering catastrophic outcomes.

The Rising Chorus of Alarms

Blair’s drastic move coincides with exploding investment in AI safety. Nonprofits like the Center for AI Safety and firms like Anthropic now dedicate resources to “aligning” AI with human values. Their goal: prevent rogue systems from manipulating information, economies, or infrastructure. Yet researchers remain deeply divided.

“Human extinction remains unlikely with robust safeguards,” argues Dr. Lena Petrova, AI Ethics Director at Stanford. “But dismissing these concerns outright is irresponsible. We’re navigating uncharted territory.”

Others push back vigorously. MIT economists Daron Acemoglu and Simon Johnson contend that AI’s primary risk isn’t annihilation, but misguided prioritization. In their analysis, corporate America’s rush to replace workers with AI tools rather than augment human capabilities could trigger mass unemployment and inequality long before AGI emerges.

Education Exodus and Workforce Disruption

Blair isn’t alone. Universities report rising student anxiety about AI displacing white-collar roles. Entry-level corporate job postings fell 15% this year, per Handshake data 2, while internships shrink as companies deploy AI for coding, compliance, and content tasks. Over 10,000 recent U.S. layoffs explicitly cite AI adoption as the cause.

“Why hire a junior analyst when GPT-6 can draft reports?” said Mark Chen, a Berkeley sophomore shifting from finance to welding. “Trades seem safer now.”

This pivot reflects Indeed’s findings: unlike past automation waves targeting manual labor, generative AI threatens knowledge workers most. Legal research, software engineering, and accounting roles face up to 95% task automation potential. Yet hands-on jobs nursing, plumbing, and childcare, remain insulated by their physical and relational demands.

Two Futures: Replacement vs. Empowerment

The debate crystallizes around competing visions. Replacement-focused automation could eliminate routine cognitive work but risks economic upheaval. Augmentation, alternatively, envisions AI as a collaborator: accelerating drug discovery, personalizing education, or aiding climate science.

Microsoft CSO Eric Horvitz champions this outlook: “We’ve barely scratched human-AI collaboration’s surface. It could democratize expertise, letting teachers or electricians achieve once-impossible problem-solving”.

For students like Blair, the pace of change outweighs optimistic scenarios. “AGI isn’t sci-fi; it’s labs racing for dominance,” she says. “If there’s even a 5% chance this ends humanity, I’d rather fight that risk now than write code for it later.”

The Path Ahead

While Blair joins AI safety nonprofits, educators urge measured responses. Northwestern’s Center for Human-Computer Interaction emphasizes designing “human-centered” tools and policies, ensuring workers share AI’s gains.

“Exiting education isn’t the answer,” argues MIT’s David Autor. “But curricula must evolve. Critical thinking, ethics, and adaptability will define the AI-resilient professional”.

As billions pour into AGI development, society’s challenge is twofold: mitigate apocalyptic risks while navigating AI’s real-world impacts on careers today. For a generation raised amid ChatGPT and deepfakes, the future feels simultaneously limitless and perilous.

Subscribe to my whatsapp channel