The second, opposite position, held by many who adopt the doomerism I described above, is the pessimistic claim that there are certain dynamics in the training process of powerful AI systems that will inevitably lead them to seek power or deceive humans. Thus, once AI systems become intelligent enough and agentic enough, their tendency to maximize... See more

So AI leaders promise that investing trillions into data centers will lead to a cure for cancer and various other downstream benefits, without stopping to think about how these inputs will translate into the desired outputs. It also leads them to support some extreme policies, such as investment in AI science at the expense of human scientists, and... See more

EMBEDDED: Do you believe that the “artificial general intelligence” and “superintelligence” that many AI boosters have warned of actually pose a risk to humanity?

KAREN HAO: These terms are really poorly defined, so I generally loathe using them at all. But if we’re talking about the idea that AI could go rogue and destroy humanity? No. This is a... See more