When Yuxiang Zhou, founder of Black Lake Technologies, tested an AI sales coach against the company’s most experienced sales leader, seven of Black Lake’s 10 most senior salespeople selected the machine’s advice.

The experiment grew out of Zhou’s frustration with the way his customer relationship management software summarized a full day of client conversations. The system reduced those discussions to “a very subjective one paragraph, and then the leader sits back and reads this paragraph and analyzes.”

Black Lake, a Shanghai-based startup that supplies AI agents to factories across Asia, responded by giving its salespeople a pin that recorded their conversations and location. The recordings were then processed by an AI agent trained to operate as a sales coach.

“A lot of people use AI as a way to co-pilot the leadership, but I’m trying to explore the way of me co-piloting the AI,” Zhou said at the Fortune Leaders Forum in Macau on Sept. 8. “A lot of my decisions are now made by agents, large language models, and my executives.”

Human salespeople nonetheless remain essential, Zhou said. AI agents are “emotionless,” and sales teams still need motivation—something that, in his view, only a person can provide.

Zhou was one of three panelists at the Forum discussing how the role of corporate executives may change if AI can handle both reasoning and execution.

For Becca Carroll, chief strategy officer at design agency IDEO, speed alone is not enough to establish a durable competitive advantage. Being first to market—or even being the first to follow a new innovation—may not create a lasting moat.

“Once you dream of something, your competitors could roll it out to market the next day, or two weeks later,” Carroll said.

She also described an updated version of IDEO’s well-known “T-shaped person” concept. The model, popularized by former CEO Tim Brown, refers to someone with deep expertise in one discipline. Carroll’s newer “comb-shaped” archetype describes someone who is “almost polymathic, someone who has depth of expertise across a series of disciplines and can fluidly move between them.”

James Arnett, Wipro’s managing director for Southeast Asia, Greater China, and South Korea, framed the change in leadership terms: “The big shift is…from directing to orchestrating.” Wipro, he said, has moved “way past” asking how AI can reduce costs for clients. It is now considering how to redesign its offerings from the ground up.

The panelists agreed that the main constraint is not technology, but people—including employees, customers, and executives. “A lot of the time, we can see that AI can do something—but is the organization ready for it?” Arnett said.

Carroll said organizations must move at the speed of trust. “We can change technology all we want, but we need to move at the speed of trust… It’s not just a question of what technology we can adopt and be done, but what’s the trust journey we are designing in service of our bigger vision.”

Despite his experiments, Zhou said he is still waiting for AI’s decisive inflection point. “I’m still looking for the Google moment for AI,” he said. “I don’t think it has happened yet.”

To help explore that possibility, Black Lake is testing a platform that uses AI to turn a celebrity’s idea into manufactured merchandise. The platform can reduce the minimum order for custom collectibles from 10,000 pieces to 100.

“I think the real Google moment should happen only when you restructure the industry rather than replace human beings,” Zhou said.

This story was originally featured on Fortune.com.