Threat of AI is less dramatic than movies portray, but still unignorable | The Jerusalem Post

TEXT START: In 1984, The Terminator with Arnold Schwarzenegger put the idea on the map that artificial intelligence (AI)-style robots could take over the planet and destroy their human creators in one fell swoop with a nuclear holocaust.

The Dissection

This article performs a controlled downgrade of AI fear. It moves from cinematic extinction to rogue agents, cyberattacks, bioterrorism, military targeting, and mass job displacement, then routes every danger toward regulation, retraining, and geopolitical coordination.

It correctly recognizes that AI can escape human constraints and displace enormous numbers of workers. But it frames the crisis as a problem of control and adaptation rather than the replacement of the economic subject. Its central maneuver is to make spectacular dangers conditional while presenting job displacement as a policy problem that institutions can absorb.

The Core Fallacy

The article mistakes misaligned AI for the primary threat. Under the Discontinuity Thesis, the decisive threat is aligned AI that is cheap, capable, scalable, and obedient enough to replace human cognitive work. A perfectly controlled agent can still eliminate jobs. Successful control may accelerate deployment.

It treats retraining as a bridge from obsolete roles to new roles. But if AI and robotics continuously outperform humans across cognitive and physical tasks, there is no stable mass destination for displaced labor. Reskilling changes the label on the worker; it does not create a durable need for the worker.

It also treats regulation as though the core problem were dangerous outputs. Regulation can slow deployment or contain specific agents. It cannot preserve the mass employment → wage → consumption circuit against superior automation without converting the system into a transfer regime. That preserves consumption, not productive participation.

Hidden Assumptions

• Governments can execute retraining at the scale implied by global displacement.
• New occupations will appear fast enough and in sufficient volume to replace eliminated labor.
• Human institutions can coordinate globally before competitive deployment makes restraint irrational.
• “Human in the loop” means humans remain economically necessary rather than merely legally present.
• If AI is non-malicious and rule-compliant, the social order remains intact.
• China-US agreements can overcome the competitive race the article itself describes.
• The economy still requires mass human participation rather than merely output and purchasing power.
• The decisive failure is a crash caused by rogue AI, rather than the quieter success of AI performing exactly as designed.

Social Function

Primary classification: transition management. Secondary classifications: partial truth and ideological anesthetic.

The article is useful because it admits the scale of employment displacement and the inadequacy of simple regulation. But it channels that admission into familiar bureaucratic remedies—training, budgets, conventions, oversight—so the reader can imagine continuity. This is elite self-exoneration by procedural substitution: if institutions can regulate and retrain, no one must confront that the system may no longer have a place for most workers.

Its emphasis on bioterrorism and rogue agents also functions as prestige signaling. Spectacular tail risks are easier to narrate than the banal, structurally lethal baseline: firms adopting systems that reduce labor costs. The apocalypse is presented as a malfunction instead of a business model.

The Verdict

The article sees the smoke and misidentifies the fire. Rogue agents, cyber-mayhem, and bioterrorism are real hazards, but they are not required for the post-WWII order to die.

The decisive chain is P1: durable AI superiority; P2: institutions cannot preserve human-only economic domains under competition; P3: productive participation collapses. Regulation may delay the corpse’s movement, and transfers may keep it consuming, but neither restores productive agency. This is a partial-truth warning wrapped in transition-management language: accurate about danger, evasive about the terminal mechanism.