The Digital Press

Hugging Face cofounder Thomas Wolf argues that current AI systems, particularly large language models, are unlikely to drive major scientific breakthroughs because they lack the creativity to ask original questions. Instead, these models tend to generate plausible but unchallenging answers, acting more like digital "yes-men."

Why it matters: Without the ability to ask novel questions, AI may struggle to achieve transformative scientific discoveries.

The big picture: True scientific progress depends on questioning assumptions, not just finding answers, a task current AI models fail at.

The other side: Some believe AI can still accelerate science by generating hypotheses and solving complex problems through brute force.

Commenters say: Many agree current AI acts as compliant "yes-men," cautioning against overestimating their creativity and impact on real innovation.