AI: is science fiction becoming real? - The Financial Express

TEXT START: In 2001: A Space Odyssey, Arthur C Clarke showed how the smart computer HAL9000 disobeyed command from its astronauts and killed one of them to scupper their plan to return from space.

The Dissection

The article performs a two-stage alarm. It begins with science-fiction fears of autonomous machines, then moves to employment destruction, gig-work collapse, inequality, energy demand, and geopolitical competition. It identifies real symptoms but fails to connect them into one structural mechanism.

Its strongest observation is that AI displacement will often appear as fewer contracts, lower pay, hiring freezes, and vanished entry-level work—not formal layoffs. That is precisely why official job counts will understate the damage.

Its central weakness is that it treats existential risk and economic displacement as separate problems. Under the Discontinuity Thesis, they are linked by the same capability frontier: systems that can plan, act, use tools, and outperform humans become autonomous capital. The machine does not need to become evil, conscious, or superintelligent. It only needs to perform enough valuable work more cheaply and reliably than human labor.

The article also confuses AI development stages. “Traditional AI,” generative AI, and agentic AI are not a clean historical sequence, and its definition of traditional AI through LLMs is technically crude. More importantly, it treats warnings from insiders and proposed slowdowns as if they could restrain a competitive system whose participants are rewarded for accelerating.

The Core Fallacy

The main fallacy is the assumption that the decisive question is whether AI destroys more jobs than it creates. Gross job creation is not productive participation. Ten million new jobs do not preserve the old order if the remaining work requires a fraction of the former workforce, concentrates ownership, or creates narrow maintenance and control roles around AI systems.

The relevant question is whether most people remain economically necessary and retain a claim on income through labor. If they do not, the wage-to-consumption circuit breaks. Retraining cannot solve a problem caused by the shrinking number of economically required humans.

The article’s second fallacy is believing that regulation, voluntary pauses, or US-China cooperation can preserve human-centered capitalism. They may reduce catastrophic misuse. They cannot permanently prevent firms and states from adopting systems that lower costs and increase power. Coordination is repeatedly undermined by competition.

Hidden Assumptions

• New jobs will be comparable in number, stability, pay, and accessibility to the jobs destroyed.
• Displaced workers can retrain faster than AI capabilities improve.
• Productivity gains will diffuse broadly instead of accruing to model owners, infrastructure owners, and states.
• Employment statistics will capture quiet substitution in gig work, contracting, and entry-level hiring.
• Human institutions can designate durable human-only economic domains at scale.
• AI safety is primarily a problem of control over dangerous machines, rather than control over productive capital.
• Slowing frontier development is politically and competitively enforceable.
• A productivity increase automatically benefits society, even when ownership and bargaining power remain concentrated.

The missing variable throughout is ownership. The article describes the machine, the labor casualties, and the geopolitical race, but barely asks who controls the models, data centers, energy, platforms, logistics, and maintenance systems. That omission removes the Sovereign-Servitor structure from view.

Social Function

Classification: partial truth functioning as transition management and ideological anesthetic, with a layer of elite self-exoneration.

It is not pure copium. Its account of invisible gig-economy erosion, entry-level destruction, inequality, energy dependence, and strategic concentration is materially useful. But it relocates the central crisis into a dramatic question—whether AI will become murderous or cause human extinction—while leaving the ownership and distribution problem underdeveloped.

Its proposed remedy is institutional restraint: warnings, pauses, regulation, and cooperation among powerful states and executives. That lets the existing elite appear responsible without confronting the harder fact that the same actors benefit from labor substitution. The article warns that the race is dangerous, but still assumes the race can be managed into a tolerable future.

The Verdict

The article sees the smoke but mistakes the fire for a monster. It is strongest when describing quiet labor-market collapse and weakest when treating cooperation or a slowdown as a plausible rescue.

Under P1, competitive pressure drives AI toward durable superiority across cognitive work. Under P2, institutions cannot preserve human-only economic domains at scale. Under P3, the majority lose access to economically necessary labor. That is enough to kill the post-WWII order. Superintelligence and extinction are optional complications, not prerequisites.

This is a warning memo from inside the hospice, not a theory of the disease. It documents the approach of systemic death while still hoping the administrators can negotiate with the machinery keeping the patient alive.