Leadership1 publisher3 min readPublished
Dreamforce attendees answered an AI extinction question with worries about jobs, mental health and the environment, not doom
Business Insider asked seven named people at and around Salesforce's conference where they put the odds of an AI catastrophe. Six treated it as remote, and the specifics they offered instead were electricity prices, privacy and school deepfakes.
The Board Room · Leadership desk
What happened
- A Business Insider reporter spent Tuesday at Salesforce's Dreamforce conference asking attendees where they put the odds of an AI catastrophe, and was told about labor, mental health and the environment.
- Two of the six people quoted at the conference raised data centers, one on environmental impact and one on electricity prices, and a third put his concern on privacy and mental health.
- Valerie Sizemore of the nonprofit Stop AI, met near the venue, said her group fears human extinction within a matter of years and wants development of superintelligent AI banned.
Compiled by The Board RoomSomething wrong?How this is made
Why it matters
- constraint The two questions this sample actually asked, who pays for the power and what dependence on the tools does to the people using them, sit outside an internal AI program scoped around model-safety pledges, with nobody assigned to them.
- contradiction Two-thirds of Americans grant AI a moderate chance of destroying humanity, while six of the seven people quoted called it remote. The polling measures worry and the interviews measure urgency. Seven interviews cannot say which one turns up at your next all-hands.
- exposure One interviewee gave only the name she uses with friends because she feared retaliation from her employer. A company running its own listening exercise should expect the sharpest views to arrive anonymously or not at all.
- precedent A senator's proposed ban until a federal regulator writes safety rules gives the extinction argument a legislative route. That exposure develops on a timeline unrelated to whatever staff rank first.
The labor worry is the least supported part of this sample. The reporter's summary line says attendees raised impacts on labor, mental health and the environment [2]. The published quotes cover data centers, electricity prices, privacy, dependence, deepfakes in schools, military use and the trauma of workers labeling horrific training content [11][12][14][16][17]. The nearest thing to a comment about work came from Sara Jamison, who said she is happy to see AI automating boring work, though she does not want it used for art [11].
Seven people at one conference on one Tuesday is not a survey [1]. It sits beside a figure pointing the other way: a Politico poll taken this week found around two-thirds of Americans think there is at least a moderate chance AI will destroy humanity [3]. Dean Levita, a local tech worker at the conference, gave the line that squares the two. "It's cautionary, it's not imminent," he said [7], comparing the discussion to the earthquake warnings he has heard across a lifetime in the Bay Area [6].
Dreamforce is a self-selecting room. People who turn up at a Salesforce conference have already decided the software is worth their week, and Henry Sandles and Jamison were there to learn more about a product their nonprofit already uses to preserve a historic railroad [8]. It also makes them worth asking. These are the people who put the tools into organizations, and their list is the list that arrives in an internal Q&A. Deep Karia, at the conference representing an Indian software development company, said he is "not at all" worried about extinction. He puts his concern on privacy and mental health, with someone growing dependent on the tech and suffering a psychotic break as his disaster case [14].
The extinction argument was present, just outside. Valerie Sizemore, a co-leader of the nonprofit Stop AI, spent the day handing out flyers next to a drawing of the Grim Reaper. Her group fears human extinction within a matter of years and wants superintelligent AI development banned [18]. She said it has been "really nice" to see AI risk discussed on the news [20]. The week that produced those broadcasts also produced Anthropic researcher Jacob Coxon's resignation, in which he said the labs are "gambling with our lives", and long posts on new safety measures from the heads of Anthropic and OpenAI [4][5].
For anyone planning the next two quarters, the two agendas run on different timescales. A federal moratorium proposal and the labs' safety posts are an argument about what gets built over years [19][5]. Electricity prices and dependence are on this quarter's list, and they land on procurement, facilities and HR. Ashleigh Kaye gave only the name she uses with friends because she feared retaliation from her employer. She travelled from Oklahoma and raised data centers and power prices, then said the harder problem is how much of daily life now runs through the technology [12]. "It's hard to get away from AI," she said. "Everything's AI-driven." [13]
What to watch
- Whether Sanders' proposed moratorium picks up co-sponsors or a hearing date.
- Whether a properly sampled employee survey backs what these seven interviews suggest about power costs and mental health outranking extinction.
- Whether more researchers follow Coxon out of the labs with public statements about safety.