Scientists have warned about human extinction for centuries, now including AI

In short: Warnings that humanity could be wiped out have a long history, and some researchers now place advanced AI on that list of possible threats.

What's going on

A New York Times opinion piece argues that fears about humanity’s possible destruction are not new. Historians trace early versions of this idea back to the 18th and 19th centuries, when thinkers like David Hume and Buffon discussed the possibility of a world without humans.

After World War II, the topic became more concrete. Nuclear weapons made it possible to imagine a disaster that could end civilization, or even human life. In the 1960s and 1970s, concerns widened to include overpopulation and environmental damage, like pollution and limits on natural resources.

In the 1990s through the 2010s, more public scientific warnings focused on climate change and the loss of plant and animal species. More recently, researchers have used the term “existential risk,” meaning a risk so large it could permanently end human life or permanently damage human society (like a fire that burns down the whole house, not just one room).

The piece also points to AI as part of this newer discussion. It cites reporting that an Anthropic researcher warned that rushing toward “superintelligence” could create an extinction risk. Superintelligence is a hypothetical AI that could outperform humans at most tasks.

What to watch

The key point is that this is a warning tradition, not proof that AI will cause extinction. What matters next is whether more researchers, companies, and governments treat these extreme scenarios as worth planning for, alongside older risks like nuclear war and climate change.

Source: NYTimes