Marc Benioff Dismisses AGI Hype, Says AI Still Far from Human-Level Intelligence

Salesforce CEO Marc Benioff has joined the growing list of industry leaders skeptical about the hype surrounding artificial general intelligence (AGI). Speaking on the “20VC” podcast with Harry Stebbings, Benioff openly questioned claims that AGI is near, pointing out that technology today is still a long way from reaching human-level intelligence.

When Stebbings referenced Amazon’s AGI research team in San Francisco, which recently suggested that fewer than 1,000 people worldwide possess the expertise to build the next generation of AI models, Benioff responded with sarcasm. He remarked that even the title “AGI head” sounded contradictory, adding that much of the AGI narrative feels like “hypnosis” about what artificial intelligence might achieve.

Benioff acknowledged that large language models (LLMs) have made impressive advances in natural language processing and automation but firmly stated that these tools should not be confused with genuine human intelligence. “It is not a person, it does not feel pain, and it has no compassion,” he said, underscoring the deep gap between machines and human beings.

Industry Divided Over AGI

The debate around AGI has split Silicon Valley leaders. OpenAI CEO Sam Altman recently clarified that GPT-5 is not AGI, as it cannot learn continuously. While Altman has spoken about the economic potential of AGI, he admitted the current technology is far from reaching such milestones.

Google DeepMind’s Demis Hassabis has also emphasized the challenge of consistency in AI models, noting this as one of the biggest hurdles in pursuing AGI. Meanwhile, former Google CEO Eric Schmidt warned in an opinion piece that Silicon Valley’s fixation on AGI risks distracting from the tangible benefits today’s AI tools already provide across industries.

Fiction vs. Reality

Despite his skepticism, Benioff did not completely rule out the possibility of AGI in the future. He referred to Hollywood films like Minority Report and WarGames—projects on which Salesforce’s chief futurist Peter Schwartz served as a consultant—as examples of how people often blur the line between science fiction and real-world technology. For now, however, he stressed that such visions remain fictional rather than reflections of the current state of AI.

His perspective resonates with other leaders who have urged caution against exaggerated fears of an AI-driven doomsday. Recently, David Sacks, White House crypto adviser, remarked that predictions of an “AGI apocalypse” are largely overblown.

A Measured View

Benioff’s comments reflect a more measured approach to AI—recognizing the significant progress made while firmly reminding audiences that machines cannot replicate the essence of human intelligence. While some in the tech industry continue to push bold timelines for AGI, Benioff’s take serves as a reminder that the technology is still far from achieving the human-like abilities often portrayed in movies or imagined by futurists.