There is a curious contradiction at the heart of the debate over artificial intelligence. The people who understand the technology best are now warning us about where it could lead, while the rest of us still debate whether it is acceptable to use it to write an email.

You might worry that the local Facebook page is full of AI-generated posters for car boot sales; meanwhile experts like Evan Hubinger, alignment science lead at Anthropic, warned this week that the chance AI could “kill all humans” in the next decade was greater than 10 per cent.

That has focused minds somewhat. But there is still a gap between what most of us see of AI in our day-to-day lives and what it is truly capable of. That gap should worry us.

The biggest mistake is to assume that if we don’t understand exactly how AI works, we don’t need to worry about it. Most of us couldn’t explain how a nuclear reactor works, but that would not stop us listening if experts told us a meltdown was imminent. So why,........

© iNews