Altman Warns of Two Ways AI Could Go ‘Very Wrong’

Sam Altman has warned of two main ways AI could go “very wrong.”

Losing control over the future of AI: This is something Altman has described as “unacceptable.”

He said AI must “always serve people” and that, to guarantee this, safety and alignment techniques must be kept ahead of the development of models’ capabilities.

Concentration of power: Altman also warned of a world where AI power would be broadly concentrated “by one person or company.”

Such a situation, according to him, could create major risks for society and for the way this technology is controlled.

In the end, Altman’s concern is not only about how powerful AI can become, but also about the way it will be controlled and by whom.

He emphasizes that the rapid development of AI must be accompanied by strong safety mechanisms, so that the technology remains in service of people and does not end up as an uncontrolled source of power in the hands of individuals or single companies.