AI Experts Can't Agree on How to Warn You About AI's Dangers
The way scientists talk about AI risk could change how you vote and what you fear.
There's a big argument among people who study how we talk about huge risks, like climate change or AI wiping out humanity. Some say we should shout about the worst-case scenario because it's true and fear makes people act. Others say that just makes people feel hopeless and ignore smaller, more likely problems. This fight has been going on for years, and now it's happening in AI safety.
The author says we're stuck in this debate and need a new approach. They point out that we've now had "warning shots" – real incidents where AI caused harm, like AI-powered cyberattacks or deepfakes. These concrete examples can help make the big, scary risk of human extinction feel more real and relatable. Instead of just arguing about whether to use fear, we should test how to talk about AI risk in ways that actually motivate people.
For example, when studies show that people care more about everyday risks, it might be because those risks are easy to picture. Now that AI has caused real damage, we can use those stories to explain why we need to prevent the ultimate catastrophe. The author gives tips for having these conversations, like focusing on concrete examples and avoiding jargon.
This matters because how we talk about AI risk affects public opinion, which influences laws and funding for safety research. If we scare people too much, they might tune out; if we're too calm, they might not take it seriously. The goal is to find a way to communicate that leads to action, not panic or apathy.
- There's a debate among experts about whether to use scary language when talking about AI's risk of ending humanity.
- Recent real-world AI harms, like cyberattacks, can make the big risk feel more concrete and personal.
- How we talk about AI risk affects public policy and funding for safety measures.
Why It Matters
How we talk about AI risks shapes laws, funding, and your safety.