Unlocking the Future: OpenAI Chairman Bret Taylor’s Compelling Case for AI Agents

During a recent chat at the Mobile World Congress in Barcelona, Bret Taylor, the OpenAI board chair, didn’t clearly define what an AI agent is. He dodged a question about how “agentic AI” differs from chatbots, mentioning that people appreciate the friendly responses AI agents provide, even if they have mixed feelings about the term itself.

As the founder of a new startup focused on customer service AI, Taylor expressed his enthusiasm for current AI technology. “I am more excited about large language models than any tech I’ve seen since I discovered the internet as a teenager,” he shared with the audience.

The key difference between today’s AI service agents and older chatbots is their increased capabilities. These AIs can provide instant assistance and speak multiple languages, enhancing customer interactions.

Taylor pointed out the real-world impact of these AI agents. He mentioned working with companies like SiriusXM and ADT Security. “Now, if your alarm malfunctions, an AI can help fix it without you waiting for a technician,” he explained. “And what’s fascinating is that people really like these AI agents.”

Enhancing Customer Experiences

These advanced AI bots are making customer service cheaper for businesses. “This will elevate the consumer experience across many brands,” Taylor noted.

However, he acknowledged that highly capable bots can sometimes create new problems. For instance, some AI agents have inaccurately generated refund policies in sensitive situations like customer bereavement.

Building appropriate guidelines for AI agents is crucial for safe implementation, Taylor said. Still, he remains optimistic that as AI becomes more tailored to specific brands, these challenges will diminish.

“My approach is not to wait for technology to be flawless. It may never be. Instead, focus on a specific area to make tough problems more manageable,” he advised. He mentioned successful AI applications like Cursor and Harvey, which work within defined fields.

Looking ahead, Taylor believes AI agents will be just as essential as websites and mobile apps for businesses. “In five to ten years, I wouldn’t be surprised if an AI agent becomes a brand’s primary digital experience,” he said.

He also suggested that the way people interact with AI agents is likely to evolve. User interfaces may become less prominent as technology makes it easier to use these tools.

“I hope that the trend of everyone glued to their screens fades away. With conversational AI that understands us better, devices might become less visible, and that would be exciting,” he said, adding that he hopes his children won’t have to spend their lives staring at screens.

Addressing Workforce Changes

What about the impact of AI on jobs in customer service?

Taylor acknowledged the concern but remained positive, saying that while some jobs may vanish, new ones will emerge. He stressed that developers must engage in this conversation and not simply push out technology.

He warned that if job changes outpace society’s ability to reskill, it could lead to significant disruption. “Ideal solutions require collaboration between public and private sectors,” he added.

When the moderator raised the topic of OpenAI transitioning from nonprofit to for-profit, Taylor emphasized that the organization’s mission to create beneficial AI solutions for humanity remains unchanged. He mentioned the high costs associated with AI development as a factor in their strategies.

“We want to strengthen our mission, and that’s the standard we hold ourselves to,” he concluded, assuring that their goals would not shift even if their structure might evolve.