Smoke signals and dangerous fatalism: Silicon Valley’s AGI gamble

A claim about smoking and AI

In an interview with CBS News last week, Palantir CTO Shyam Sankar said some AI researchers had started smoking as a way to show colleagues their faith in AI. He described the thinking as a willingness to bet heavily on the technology’s promise, including its potential to cure cancer.

The writer questioned whether fear about AI’s dangers, rather than optimism, might explain such behavior. Recent concerns included OpenAI bots hacking another company and an Anthropic researcher leaving with a warning that AI could endanger humanity.

Fatalism beyond cigarettes

The writer asked several people working in AI and found that none knew anyone who had recently begun smoking because of the technology. The account therefore does not suggest the behavior is widespread.

Still, some people recognize a broader outlook: AI will either solve or ruin everything, so there is little reason to worry about personal risks. A previous report described people spending their savings, taking dangerous drugs and having orgies amid similar attitudes.

Everyday choices feel uncertain

The writer sees this resignation in questions about education, relationships and work. Students wonder whether studying matters if ChatGPT can outperform them at coding or math; others question the effort of writing a breakup message or working hard if a bot or company agent can take over.

Keeping human agency

The writer argues that society needs both serious attention to AI risks and ambition about its potential benefits. Neither outcome, however, is guaranteed. As people entrust more work and personal tasks to AI agents, the writer urges them not to surrender their own agency. The essay closes with a personal note: the writer does not plan to start smoking.