Skip to main content

Salesforce CEO says AI could destroy us unless his company steps in

Salesforce CEO Marc Benioff walks before delivering his keynote speech at the annual Dreamforce conference in San Francisco on Tuesday. | Source: Justin Katigbak/The Standard

San Franciscans not attending the annual nerd blowout known as Dreamforce are probably most familiar with the way the three-day conference from the cloud software company Salesforce gridlocks the streets near Moscone Center.

But for the 35,000-odd attendees, the feeling is not unlike a pilgrimage to the ordained—if not disdained—center of the tech universe. A handful of Buddhist monks were sprinkled in that number, but for the most part, the uniform consisted of blazers, slacks and baby blue lanyards. Allbirds, it should be noted, appear to have declined in popularity.

As any good pilgrimage must, the 21st iteration of the fest has a mantra: trust in artificial intelligence.

"In Trust We AI" was printed on a fake rock overhang in the conferences' central campground-themed pathway. Salesforce CEO and Chairman Marc Benioff was spotted earlier in the morning monitoring the display, which featured projection mapping of butterflies flapping and characters dancing in the background.

Benioff, the company's sole chief executive after the departure of co-CEO Bret Taylor (who left to start an AI company), mentioned "AI" no fewer than a dozen times during his opening keynote. "Trust" came in a close second.

The keynote itself kicked off with a commercial from Salesforce's celebrity spokesperson, the actor Matthew McConaughey.

"Welcome to the biggest AI event of the year," McConaughey said in his signature Texan drawl as Dave Matthews Band rocked out in the background.

"I love Matthew McConaughey in the morning. It kind of wakes me up," Benioff joked as he walked to center stage.

Salesforce CEO Marc Benioff delivers the keynote speech at the annual Dreamforce conference at Moscone Center in San Francisco on Tuesday. | Source: Justin Katigbak/The Standard

Benioff positioned AI's advancement in four waves, moving from predictive AI to generative AI to autonomous AI agents and eventually to artificial general intelligence. He said technology is firmly in the second wave and heading further soon.

But instead of simply riding that progress, Benioff made the argument for pumping the brakes or at least positioning Salesforce as the bridge to solving what he terms the "AI trust gap."

He spoke somewhat derisively about how public large language models have been built and some of the current problems with them.

"The data they get, they mostly have stolen. They just go out onto the net and get whatever they can get," Benioff said. "The next set of data that they want is your data, the corporate data that will make them even smarter.

A pro-AI Salesforce logo inside the Moscone Center in San Francisco on Tuesday, Sept. 12, 2023. | Source: Justin Katigbak/The Standard

"You get a lot of answers that aren't exactly true. They call them hallucinations. I call them lies," he added.

Benioff laid out what he termed Salesforce's tenets of trusted, ethical and human AI. They include: "Your data isn't our product," and "We prioritize accurate, verifiable results." He described Salesforce's decisions to create a chief trust officer and chief ethical and humane use officer as forward-thinking.

"It's amazing what's happening with AI; it's incredible what's happening. We also recognize that we're in this AI revolution, but it's going to bring us back to our core values," Benioff said. He raised the example of dystopian films like The Terminator and Minority Report as what could happen if AI goes off the rails.

The Dreamforce grounds themselves, sprawling roughly five blocks in Downtown San Francisco, added a bit of a reality-distorting Midjourney-like patina.

Cutesy Albert Einstein characters—a reference to Salesforce's new AI platform—and Disneyland-esque waterfalls provide a backdrop to talks about “creating personalized customer experiences” and “unleashing the power of self-service.”

Ever the showman, Benioff did his fair share of callouts to attendees, including San Francisco Mayor London Breed, Williams-Sonoma CEO Laura Alber and Peter Schwartz, a Salesforce executive who played a role in the production of sci-fi movies like Minority Report.

"Stand up, Japanese!" Benioff said to a cohort of flag-waving Japanese visitors in recognition of Salesforce surpassing Oracle as the country's second-largest software company.

AI even served as a theme for what has become an annual philanthropic tradition for Benioff and Salesforce. The company donated $11 million to Oakland and San Francisco public schools with the goal of helping prepare students for careers in artificial intelligence. The company has donated some $230 million for public education and workforce development over its 20-plus-year history.

READ MORE: Salesforce Gives San Francisco, Oakland Schools $11M for AI Education

"There are now moments where we hear about things like the homeless or we think about public schools or what's going on in our local cities," Benioff said. "We all have a choice. One, we can look at them and go by, or we can look at them and say, 'There by the grace of God goes me, and love thy neighbor as thyself.'"

The overarching theme of Benioff's remarks and Dreamforce writ large is that Salesforce is building the guardrails to protect users from well-publicized AI problems like inaccurate information and toxicity while making them more productive.

In short, companies that are both excited and apprehensive about AI shouldn't worry too much. Just trust Salesforce to handle it.

"We all know and have in our mind what could happen if this AI thing goes really wrong," Benioff said. "Let's keep it going in the right direction."