When AI Companies Talk About What Could Go Wrong

You might have heard the chatter about Artificial Intelligence (AI) and its potential risks. One of the big AI development companies, Anthropic, has recently hired an expert whose views on these risks are certainly getting people talking.

This new hire, an economist, has previously discussed the idea of a 33 per cent chance of human extinction from AI as a potentially “acceptable” risk. Now, before anyone panics, it's important to understand this isn't a prediction, but rather part of a complex discussion some experts are having about just how much risk we should tolerate when developing extremely powerful new technologies like advanced AI. Anthropic itself is known for being very cautious and focused on making AI safe.

The real story here isn't necessarily about extinction; it's about how seriously these major AI companies are taking the ethical and safety discussions. They're bringing in different kinds of thinkers – like economists – to explore every angle, from the technical challenges to the broader societal impacts. This can feel a bit abstract, but it shows they're genuinely grappling with the 'what ifs'.

For everyday Australians, whether you run a small business or are just watching this technology evolve, it's a reminder that the people building AI aren't just focused on making it smarter. They're also thinking deeply about how to keep it safe and beneficial for everyone. This kind of robust debate, even about extreme scenarios, is a crucial part of building responsible AI that we can all trust in the long run.