Elon Musk outlines three qualities AI must have to avoid becoming a threat

Elon Musk said in a podcast conversation with Indian entrepreneur Nikhil Kamath that safe AI development is only possible if systems are built around three core principles: truth, beauty, and curiosity.

He argued that ignoring these foundations increases the risk that AI systems will produce dangerous or faulty conclusions and drift away from objective reality.

• Musk noted that AI inevitably encounters false information online, which can disrupt its reasoning mechanisms and lead to absurd or destructive outcomes.
• As an example, Musk pointed to hallucinations—instances where AI systems produce incorrect answers. He recalled how, in early 2025, Apple Intelligence on the iPhone issued a false notification claiming that British darts player Luke Littler had already won the PDC championship before the actual final took place.
• Musk also said that AI must not only pursue factual accuracy but demonstrate an aesthetic sense and curiosity about the nature of reality and the future of humanity. Without this, he warned, the technology risks becoming a tool for destruction rather than progress.
• His position aligns with warnings from AI pioneer Geoffrey Hinton, who has estimated a 10–20% chance that AI could lead to humanity’s extinction. Among short-term risks, Hinton has highlighted hallucinations and the threat of large-scale automation of low-skill jobs. He maintains that with sufficient research and resources, developers may still build