Summary
Google DeepMind CEO Demis Hassabis warned that some dangers from artificial intelligence (AI), such as attacks on important infrastructures like energy and water systems, are already real. He discussed these concerns at the AI+ Summit in San Francisco, emphasizing the need for strong cybersecurity to protect against such risks.
Key Facts
- Google DeepMind CEO Demis Hassabis spoke at the AI+ Summit in San Francisco.
- Hassabis highlighted that AI can be misused, leading to risks like attacks on critical infrastructure.
- He mentioned that these threats might occur even without very advanced AI.
- Google is focused on improving cybersecurity to defend against AI-related threats.
- Hassabis predicted that AI could match or exceed human abilities, called artificial general intelligence, by 2030.
- A concept referred to in the discussion was "p(doom)," indicating the chance of AI causing a catastrophe.
- Hassabis believes the probability of an AI-related catastrophe is significant and should be addressed seriously.
