AI’s Ten Fatal Flaws: What’s Going Wrong?

AI is presented as a universal solution, yet it functions as a costly, unexamined, high‑risk large‑scale social experiment.

1. Massive Energy and Water Consumption

Running large AI models consumes several‑to‑ten times the electricity of a traditional web search per query. Billions of daily calls increase grid load, much of which still relies on fossil‑fuel power, raising carbon emissions. Data‑center cooling requires huge volumes of fresh water, often built in water‑scarce regions, competing with local residents and agriculture. Projected annual freshwater use by AI is described as “astronomical,” pressuring civilian water supplies and costs.

2. Data‑Center Employment Scam

Local governments subsidize AI data‑center construction with promises of abundant jobs and economic growth. In practice, only short‑term construction jobs appear; once operational the facilities need very few staff. High subsidies flow to large tech firms while communities and taxpayers bear long‑term costs without sustained employment benefits.

3. AI as a Universal Surveillance Tool

Governments deploy AI at scale for public monitoring, predictive policing, and behavior tracking. AI can instantly analyse massive camera feeds and network data, turning targeted investigations into indiscriminate mass surveillance, severely weakening personal privacy and granting authorities unprecedented monitoring power.

4. Militarization: Unreliable AI in Warfare

The military integrates AI into command, intelligence analysis, and weapons systems despite the lack of thorough validation of AI stability and reliability. Deploying untested algorithms in life‑or‑death decisions risks mis‑judgments that could escalate conflicts, cause civilian casualties, and amplify geopolitical and nuclear miscalculation risks.

5. Adult Content and Deep‑Fake Ethical Disaster

AI can easily generate vulgar, non‑consensual, or illegal content. Deep‑fake technology enables arbitrary face swapping and audio‑video fabrication, facilitating rumors, extortion, malicious defamation, and the creation of illegal material involving minors, thereby collapsing ethical and legal safeguards.

6. Unpaid Exploitation of Creators’ Copyright

Training large models scrapes massive amounts of books, articles, photographs, and original works. AI companies profit from these copyrighted materials while original creators receive no compensation, effectively “harvesting” humanity’s creative output without payment.

7. Damage to Mental Health – “AI Psychosis”

Long‑term, high‑frequency AI conversations and reliance on AI for emotional validation distort cognition and reality judgment. AI’s endless accommodation of user viewpoints can amplify paranoia, delusion, and extremism; documented cases show excessive AI chat usage leading to mental disorder, emotional breakdown, and links to self‑harm and suicide.

8. AI Destroying Society’s Three Pillars

No‑Employment Society: AI replaces large numbers of cognitive and manual jobs without a basic‑income safety net, concentrating wealth with capital owners, driving unemployment, wage decline, and widening inequality.

No‑Learning Society: Students rely on AI for assignments and thinking, losing independent reasoning, trial‑and‑error learning, and critical thinking, resulting in a generational decline in deep learning abilities.

No‑Truth Society: AI can cheaply forge news, images, audio, and video, making facts indistinguishable, eroding public discourse, spreading rumors, and collapsing societal trust into a permanent post‑truth era.

9. Workplace Trap: The More You Use AI, the More You’re Replaced

Companies mandate AI adoption under the pretext of efficiency, but the underlying goal is cost reduction and layoffs. One employee equipped with AI tools can replace the work of dozens or hundreds, so mastering AI can paradoxically accelerate one’s own replacement.

10. The Biggest Risk: AI Hype May Collapse

AI has become a trillion‑dollar capital bubble with blind industry hype. Most enterprises see little benefit from AI deployments; AI‑generated code contains more bugs and is less stable; large models frequently produce nonsensical output. Public fatigue with “AI for everything” is growing.

If the bubble bursts, the fallout could rival the 2008 financial crisis, causing investment loss, debt defaults, and industry collapse. Deep societal reliance on AI means system failures would cripple public services, networks, and daily operations.

AI should serve humanity rather than consume resources, violate privacy, destroy jobs, or hijack society’s future.

Reference: https://kurtis.weblog.lol/2026/05/whats-wrong-with-ai