Why AI Won't Replace IT Professionals Anytime Soon: 9 Key Arguments
IT company leaders are under pressure to implement AI for workforce reduction, but data and history show that the technology is far from fully replacing specialists. Experts like Geoffrey Hinton and Sergey Brin have made bold promises that haven't materialized. Over 10 years, AI hasn't displaced radiologists or made autonomous vehicles mainstream.
Skepticism Toward Big Tech Claims
Leaders at Anthropic and OpenAI warn of mass unemployment, but internal research from their companies contradicts these risks. The theoretical potential of AI in finance and architecture is enormous, yet the actual "observed AI coverage" is less than 5%. The gap between capabilities and implementation makes predictions more hype for investors than reality.
AI models excel at narrow tasks but struggle with integration:
- Hallucinations and errors: Even in strong areas, AI makes absurd mistakes.
- Tasks ≠ jobs: Automating a fragment doesn't replace an entire workflow.
- Visual analysis: Problems with interpreting diagrams, schematics, and blueprints.
AI Limitations in Multimodal Tasks
Modern LLMs handle text well, but office work often requires interpreting visuals—charts, maps, blueprints. Remote work indexes show that less than 4.5% of professions are fully automatable by AI agents. In customer support, chatbots disappoint due to contextual errors.
For senior developers, this means:
- AI speeds up routine tasks (code generation, debugging) but not system architecture.
- Requires human oversight for verification.
- Doesn't replace domain knowledge in integrations and optimizations.
Physical Labor Remains Out of Reach
AI won't affect hands-on professions: plumbers, mechanics, nurses. Robotics lags behind—even in data centers, replacing technicians is unlikely without breakthroughs in embodied AI.
Layoffs: AI as a Smokescreen
Mass layoffs at Block and Klarna are attributed to AI, but the real causes are financial failures and overhiring. Klarna claimed in 2024 to automate 700 FTEs, but by 2025, it returned to hiring humans for complex cases.
Modest ROI from AI
Companies spend billions, but productivity increases by only 1–5%. Without AGI-level breakthroughs, expected in 10+ years, there are no radical leaps.
Key Takeaways
- Historical skepticism: Predictions by Hinton (2016) and Brin (2012) didn't come true.
- Theory/practice gap: <5% real-world AI task coverage.
- Partial automation: AI complements but doesn't replace due to errors.
- Klarna effect: Temporary cuts give way to rehiring.
- Focus on augmentation: Use AI to boost your current team.
IT professionals should focus on AI as a tool: fine-tuning models, prompt engineering, hybrid workflows. This enhances value without the risk of displacement.
— Editorial Team
No comments yet.