Salesforce CEO Says the Future of AI Is Not in LLMs, But Autonomous Agents

• Salesforce CEO, Marc Benioff, gave his comments on AI development at "Future of Everything" on November 23rd
• He shared that we are approaching the limit of what LLMs can do and should not stop there
• Other technologies for AI development can advance it, make it more accessible
• He sees the future of AI in Autonomous Agents and says that this transition is already happening

Salesforce CEO, Marc Benioff gave some comments on the state and direction of the AI industry on The Wall Street Journal's "Future of Everything" podcast on Nov. 23.

He believes that AI should not stop at LLM technology, which dominates now and is starting to show the limits of its capabilities and we must use new technologies to develop the industry.

He sees Autonomous Agents as one of the next key technologies, which should overcome the limitations of LLM and make AI more accessible and even more advanced.

More About His Position and Perspectives on AI

Marc Benioff frames his argument around the fact that he believes LLM has pushed AI much further than it was just a few years ago, but has its limits and we should not stop with this technology.

"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."

Instead, he suggests looking at an agent-based approach to building AI, where many small models decompose a large task into many small ones, relevantly distribute among themselves autonomously, and later produce an aggregate result.

"Let’s replace bureaucracy with an agentic layer that serves people, not politics. Welcome to the future — welcome Agentforce!"

Moreover, since the idea is not new, and there are even some tools to create such an infrastructure as Langchain, he says that such a transition has already started and the very LLM providers, not wanting to lose their power, are starting to move into it.

Salesforce in particular already offers such out-of-the-box solutions in its products:

"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."

Also, other tech giants are already investing in a similar approach, most notably OpenAI is rumored to have done something similar with its Strawberry update, and is rumored to want to do a full-scale launch of a similar product in 2025. At the same time, Nvidia is making its efforts and putting in similar investments and building a proper platform for future AI.

"We're starting to see enterprise adoption of agentic AI really is the latest rage."

Conclusion

We can see an interesting picture regarding AI, where the return on investment is still debatable, but the investment itself continues.

Is this an attempt to delay a new AI winter, a promise that new technology will solve problems and provide a long-awaited return on investment?

Or the fact that LLMs for all their impressive results have not delivered this return, but investment continues to flow in to utilize the new technology may indeed speak volumes about the long-term strategic bet on this entire industry?

Be aware and stay tuned.