Why AI Freelancers Can’t Replace Humans (Yet)

[ed]

Thinking of replacing your pricey remote freelancers with AI agents to save costs? Think again. New research highlighted by Wired shows that even the most advanced AI models struggle to handle real-world freelance tasks effectively. Tests by the Center for AI Safety (CAIS) and Scale AI revealed that leading AI agents completed less than 3% of assigned work, earning just a fraction of what human freelancers would.

From data analysis to game development, AI agents like OpenAI’s GPT-5 and Google’s Gemini 2.5 Pro fell short, often creating low-quality outputs that require human intervention. Researchers say this demonstrates a major gap between AI hype and real-world productivity, emphasizing that humans are still indispensable for complex, economically valuable tasks.

[cryptohog]

humans: slightly messy, mostly competent. ai: 3% complete, still asking for coffee breaks