At Salesforce’s Dreamforce conference in San Francisco, attendees were talking about AI risk — but not in the apocalyptic way that dominates online debates.
People at the event said they were more concerned about labor disruption, mental health, privacy, and the environmental toll of data centers than about an AI takeover. Others raised worries about electricity prices, deepfakes in schools, and the use of AI in military technologies.
One attendee, Dean Levita, said he was mostly unconcerned about an AI-driven extinction. Henry Sandles said he trusts that smart people will shut down any AI before it causes real-world havoc, and he added, 'I just can't see computers being smarter than humans.' Sara Jamison said she is far more concerned about the environmental impacts of data centers, while Ashleigh Kaye pointed to electricity prices and the difficulty of getting away from AI as it spreads through more applications.
Deep Karia, who was in town representing his Indian software development company, said his worries cluster around privacy and mental health. He said a disaster is more likely if someone becomes dependent on the tech and has a psychotic break. Soumya Nimmu, a law school student passing by the conference, said the here-and-now issues deserve more attention than extinction fears.
The broader debate is not disappearing. Valerie Sizemore, a co-leader of Stop AI, said her group fears 'human extinction within a matter of years' and wants a ban on developing superintelligent AI. She said it has been 'really nice' to see AI risk get more attention, and she spent the day handing out flyers with the message: 'AI - Either we stop it, or it stops us!'
For founders, the signal here is simple: AI anxiety is moving from science fiction toward operational risk. The questions that matter most are the ones that hit cost, trust, and control — not just what AI can do, but what it does to labor, power bills, and the systems customers rely on.
That is where the real playbook starts. If AI raises electricity use, expands compliance risk, or creates new ways to mislead users, the bill eventually lands on businesses and consumers, not on abstract doomsday forecasts. The companies that win will be the ones that treat AI as an infrastructure and governance problem, not just a product feature.
