Marc Benioff has had a lot to say about generative AI recently – from his disparaging comparison of Microsoft’s Copilot to “Clippy 2.0”, to being “blown away” by Google Gemini’s capabilities, the Salesforce CEO has been outspoken when it comes to AI trends and development.
Now, Benioff is turning his attention to large language models (LLMs), stating on The Wall Street Journal’s “Future of Everything” podcast that their capabilities have peaked and may even become obsolete as we enter the next phase in AI evolution.
Are Agents the Future of AI?
Marc Benioff has made a strategic “hard pivot” in Salesforce’s approach to AI this year, moving away from Einstein Copilot to fully embrace AI agents, as exemplified by Agentforce. His recent comments reinforce the argument that the true potential of artificial intelligence lies in the adoption of agent-driven solutions.
On the podcast, Benioff argued that LLMs have been overhyped and should not be seen as the centerpiece of AI advancement. He was critical of the widespread reliance on tools like ChatGPT, claiming it has created “inflated expectations” of what AI can realistically achieve.
“I actually think we’re hitting the upper limits of the LLMs right now. We’ve all got drunk on the ChatGPT Kool-Aid over the past few years, and consumers believe that AI is more powerful than it is and that large language models are key to technological advancement.” Marc Benioff, CEO, Salesforce
Instead, he proposes taking an agent-based approach to AI development going forward:
“Let’s replace bureaucracy with an agentic layer that serves people, not politics. We have incredible tools to augment our productivity, to augment our employees, to prove our margins, to prove our revenues, to make our companies fundamentally better, to have higher fidelity relationships with our customers.”
In this particular instance, Benioff has shown he is ahead of the curve when it comes to this agent-dependent approach, even describing AI agents as “the latest rage” in tech. Ironically, ChatGPT is now taking its first agentic steps, with plans to introduce “Operator” in January of 2025.
Likewise, Nvidia is hopping on the agentic AI bandwagon, with CEO Jensen Huang highlighting the rising trend of businesses adopting AI agents. In a recent earnings call, he hinted that AI-driven “colleagues” might soon become the norm, working side by side with human staff.
Final Thoughts
As Salesforce and other industry leaders set their sights on a future driven by autonomous agents, the conversation around AI is taking an interesting turn. Marc Benioff’s insights shed light on the growing demand for tools that can handle real-world tasks with precision – but how do we balance this ambition with the realities of today’s technology?
Let us know your thoughts in the comments below.