ChatGPT creator Sam Altman continued his whirlwind international tour at Dreamforce 2023 in San Francisco, simultaneously touting artificial intelligence's potential and terrifying the masses with his prophetic words about its real-world applications.
"The level of enthusiasm, hopefulness and excitement around the world—of course, balanced with wanting to successfully address the downsides [of AI]—was really extraordinary to see," Altman said. "I thought maybe this was just a tech, Silicon Valley phenomenon, but seeing what everyone around the world is doing with the technology, and how they've incorporated it into their lives, was really cool."
Set against a leafy green backdrop on the Yerba Buena Theater stage, Altman met with Salesforce CEO Marc Benioff on Tuesday afternoon, speaking to a packed audience about everything from generative AI to his favorite sci-fi movie, Her.
"I like Her. The things Her got right—like the whole interaction models of how people use AI—that was incredibly prophetic," Altman said.
Spike Jonze's 2013 movie Her stars Joaquin Phoenix as a heartbroken man going through a divorce who falls for an AI virtual assistant named Samantha, voiced by Scarlett Johansson.
Altman, OpenAI's CEO, maintained a prophetic yet pragmatic tone, poking fun at the yearslong process of releasing his famous GPT text generators.
"The biggest surprise is just that it's all working," Altman said, to laughter in the audience. "We want to have the smartest, most capable, most customizable models out there."
ChatGPT and artificial intelligence exploded onto the San Francisco tech scene this year after OpenAI released publicly accessible text and image generators in 2022. The excitement surrounding these technologies has lured investors and tech founders back to the Bay Area, creating a startup frenzy in San Francisco.
But as text generators like ChatGPT or Anthropic's Claude AI grow in popularity, tech leadership and government officials are increasingly flagging concerns about security, misinformation and intellectual property.
Altman has testified about generative AI in Congress and called for more robust regulatory oversight of the very technology his company creates. The entrepreneur came short of admitting an interest in entering politics on Tuesday, telling Benioff he "wouldn't be a good politician at all."
Altman suggested a new governing agency might be necessary in order to create a framework to "deal with short-term and long-term challenges" posed by artificial intelligence.
"Getting something going—even if just focused on insight and not oversight—I think would be great," Altman said.
Trust is now the name of the game. The 21st iteration of Dreamforce has a mantra, "In Trust We AI," reflecting Salesforce's efforts to position itself as a responsible tech company that experiments with new technology yet establishes guardrails and oversight at the same time.
For Altman, the goals of technological advancement and oversight are one and the same. As OpenAI scales up its efforts to develop GPT's capabilities, Altman says the very thing "helping us make these systems" will answer the question of alignment, such as making sure AI reflects human intent.
Still, the ChatGPT creator admitted he hadn't yet heard of a good solution for the surveillance state issues AI introduces.
"I don’t see a world where—if AI is as powerful as we think and people can do significant harm with it—I don’t see a world where we don’t have less surveillance," Altman said. "I don’t think that's a good thing."
Nevertheless, Altman's hourlong talk struck a future-oriented tone, befitting of artificial intelligence's most prolific—and arguably most influential—leader.
"In terms of 'how we make progress' and 'are we surprised'—not really. We are empiricists; we know we are going to be surprised," Altman said. "We're never surprised; we just try to meet reality where it is and follow the technology where it can go."
Altman, though an AI evangelist, is also a self-described doomsday prepper and survivalist, who once admitted to owning guns, antibiotics and a "patch of land in Big Sur" to escape to, should disaster strike.
Though bullish on the future of his technology, the OpenAI CEO admitted on Tuesday that artificial intelligence won't save him in a real-world crisis.
"I have no delusions that any of that is going to protect us or anyone if AI goes wrong," Altman said. "I think it's silly that's what people assume it's for. It was just like a boyhood dream that stuck."
Dreamforce is an annual three-day technology conference in San Francisco, hosted by cloud management company Salesforce. The company says Dreamforce is the “largest software conference in the world,” boasting appearances from high-profile politicians, A-list celebrities and prominent tech bosses.
This year is the 21st iteration of Dreamforce. Around 40,000 people attended the conference last year, much lower than previous years, where attendance reached, on average, roughly 200,000. Dreamforce is expected to draw between 30,000 to 35,000 attendees to Moscone Center this year.