May 12, 2025, 3:00 AM ETInterview, Fireside Chat

AI AGENTS DEBATE: These Jobs Won't Exist In 24 Months!

The panel describes AI agents as the creation of a "new species" that has crossed from "highly complicated" to "truly complex," rendering future outcomes fundamentally unpredictable.
The technology is moving from simple request-response models to autonomous agents that can work indefinitely until a goal is achieved or an error occurs.
The speed of this evolution is exponential: the duration an agent can operate autonomously is doubling every seven months, moving from 30 minutes today to potentially days or weeks soon.
The consensus is that society is at an "event horizon" similar to the invention of writing or electricity, but the rate of change is occurring weekly rather than over centuries.

Immediate Displacement: Routine jobs involving text-in/text-out workflows, such as data entry, quality assurance, and basic coding, are projected to be eliminated within the next two years.
High-Skill Vulnerability: High-status professions, including anesthesiologists, accountants, lawyers, and graphic designers, face significant automation risks due to AI's ability to process complex data and images.
Inequality Escalation: The gap between "high agency" individuals (who can orchestrate AI swarms) and "low agency" individuals (who cannot) is predicted to widen dramatically, creating a potential 1,000x variance in productivity rather than the previous 10x.
Global Impact: Outsourcing hubs in countries like India and the Philippines face existential risk as businesses no longer need to send jobs abroad when AI can perform tasks remotely and instantly.
The "Useless Eater" Scenario: There is a specific fear that mass unemployment will lead elites to view the displaced population as "useless eaters," leading to potential social collapse or authoritarian responses rather than humanitarian aid.

The "10x Harm" Ratio: Panelist Brett suggests that while the potential for good is infinite, the potential for harm is ten times greater, citing undetectable deepfakes, scams, and targeted manipulation as immediate threats.
Autonomous Weapons and Surveillance: The deployment of facial recognition and autonomous drones for targeted assassination or citizen subjugation is a primary concern, with existing infrastructure in Iran, London, and the US already enabling these capabilities.
Cognitive Derangement: The ability of AI to generate undetectable deepfakes is causing a crisis of truth, where individuals cannot verify events, leading to societal paralysis or cynicism.
Market Manipulation and Scams: Bad actors can use AI to scale personalized phishing attacks, deepfake fraud (e.g., voice cloning to trick family members), and crypto scams at an industrial scale.
The "Fermi Paradox" Solution: One theory posits that civilizations may invent infinite pleasure simulations (VR/AI) that lock them in solipsistic environments, preventing further exploration or leading to species extinction.

One-on-One Tutoring: AI is identified as the only scalable solution to the "two-sigma" problem, providing every child with high-quality, personalized tutoring previously limited to the wealthy.
Curriculum Overhaul: Traditional education focused on memorization is obsolete; the new curriculum must focus on "generalist" skills, adaptability, prototyping, and distinguishing between "creation" and "consumption."
Loss of Human Agency: There is a concern that AI will remove the need for human agency entirely, as agents can predict needs and execute tasks (e.g., building businesses, managing schedules) without human input, potentially leaving humans without purpose.
Meaning Crisis: The transition to an "age of abundance" may trigger a mental health crisis, as humans lack a mechanism to find meaning without the struggle of traditional labor.

Infinite Leverage for Small Teams: A team of 5-10 people armed with AI agents can now achieve the output of a 1,000-person company, allowing for rapid wealth creation and the launch of businesses with minimal capital.
The Death of Technical Moats: Traditional moats based on engineering talent or capital (e.g., $30 million for a dev team) are dissolving as AI reduces coding costs to minutes and dollars.
New Moats: Value will shift to distribution (audience size), taste, and "human agency" (the ability to identify and direct AI agents).
Rapid Prototyping: The barrier to entry for building software has vanished; examples include building a SaaS company in 15 minutes or creating internal business tools for $1 instead of $10,000.

US-China AI Race: Investment is heavily concentrated in the US and China, creating a competitive dynamic where regulation is viewed as self-sabotage that could hand dominance to the other nation.
Regulatory Disadvantage: Europe's strict AI regulations (e.g., preventing the release of models like OpenAI's O3) have placed it at a competitive disadvantage compared to the US and China.
Inability to Contain: The panel argues that "containment" of AI is impossible due to game theory; if one entity restrains itself, competitors will inevitably advance, making total surveillance or draconian control the only theoretical alternative to an open arms race.
Universal Basic Income (UBI): While some (like Sam Altman) anticipate UBI will be necessary due to job loss, the panel largely disagrees, arguing that human psychology requires purpose and work, and that UBI might lead to parasitic behavior rather than fulfillment.

Demographic Collapse: The combination of AI abundance, virtualized environments, and reduced need for human labor is predicted to further accelerate declining birth rates, potentially leading to a population crash (e.g., South Korea's 0.72 fertility rate).
The "Adaptive Valley": Humanity must traverse a period of severe disruption and uncertainty (the "adaptive valley") to reach a higher state of human flourishing, with no guarantee of survival through the transition.
Lifelong Learning: The concept of a 45-year linear career is dead; the future belongs to "lifelong learners" who must constantly pivot and acquire new skills in 10-36 month cycles.
Human Specialization: The only remaining human advantage may be "agency" (the drive to coordinate and direct) and "emotions," as AI surpasses human cognitive and physical capabilities.
Optimism vs. Realism: While the potential for solving global problems in healthcare and education is "phenomenal," the panel agrees that the immediate future involves significant carnage and requires deliberate societal preparation to avoid catastrophe.