Companies building superhuman AI can’t explain their models’ decisions
Tech firms racing to build the most powerful AI systems admit they don’t fully understand why their machines make certain choices.
The issue started as AI models grew more complex and opaque. Developers hit a wall trying to trace how decisions get made inside these black-box algorithms. 
Soon after, experts flagged the risks of deploying AI without clear reasoning. The lack of transparency sparks concerns over trust, safety, and ethical use.
The launch of ever-larger models only deepens the mystery. Efforts to build “explainable AI” keep falling short as companies push the limits.
This photo from Axios captures the scale of the problem:
No major company has publicly solved this yet. Users and regulators are watching closely as this transparency gap grows wider.
