Marc Benioff says he's 'extremely suspect' of those buying into AGI 'hypnosis'

Marc Benioff, CEO of Salesforce, has expressed skepticism regarding the concept of Artificial General Intelligence (AGI), calling it a potential misnomer. During an appearance on the '20VC' podcast, he reacted to comments made by Amazon's head of AGI lab about the limited number of people capable of contributing significantly to AGI development. Benioff remarked, "AGI head — that sounds like an oxymoron," emphasizing his doubts about the validity of those who use the term 'AGI.' He believes the tech community has been swept up in what he describes as 'hypnosis' regarding the imminent capabilities of AI. The conversation around AGI has intensified in Silicon Valley, with industry leaders debating not only its feasibility but also its timeline. For instance, OpenAI's Sam Altman recently stated that their GPT-5 model does not qualify as AGI due to its inability to learn continuously. Meanwhile, Demis Hassabis, CEO of Google DeepMind, pointed to inconsistencies in AI that hinder its progression toward general intelligence. While Benioff acknowledges the potential for AGI in the distant future, he remains grounded about current technological capabilities. He noted, "We've all seen those movies," referencing the influence of futurists like Peter Schwartz, who worked on films such as 'Minority Report.' However, he is clear that today’s technology does not reflect the scenarios depicted in cinematic narratives. Benioff's views align with a growing number of tech leaders who are cautious about the AGI hype. Former Google CEO Eric Schmidt and analyst Selina Xu recently warned that the obsession with AGI is alienating the public and missing vital opportunities to harness existing technologies. Additionally, David Sacks, a prominent figure in cryptocurrency policy, described fears surrounding an AGI 'apocalypse' as exaggerated. Despite his reservations about AGI, Benioff remains enthusiastic about AI advancements, acknowledging large language models as state-of-the-art. However, he cautioned, "It's not a person, and it's not intelligent, and it's not conscious," emphasizing the limitations that still exist in AI technology today.