Bret Taylor, co-founder of Sierra and chairman of OpenAI, says the quiet part out loud. AI is coming for programmers. “The thing I self-identify with [being a computer programmer] is, like, being obviated by this technology,” he admitted on a recent podcast with The Information. For someone who built Google Maps, led Salesforce as co-CEO, and still calls himself a coder, Taylor’s warning lands with weight. And it frames a reality that the man now pitching AI agents to Fortune 500s is also predicting the extinction of his own profession.
Taylor argues that the enterprise software market moves like a pendulum between “best of breed” products and “best of platform” incumbents. Right now, he says, the pendulum is firmly on the side of startups building agents that solve specific problems more effectively than larger competitors.
Sierra, founded in 2023, sits squarely in this trend. The company builds customer service agents that handle millions of calls and chats, offering businesses measurable efficiency gains. Taylor framed Sierra’s model as “pay per outcome”: companies are only billed when an agent successfully completes a task. “If our agent solves the problem or makes the sale, we get paid. If not, it’s free,” he said. That approach, he added, directly aligns Sierra’s incentives with its customers.
“In enterprise software, there’s an old debate between best of breed and best of platform,” Taylor explained. The reason, he noted, is straightforward. The financial gains from working AI agents dwarf the licensing costs of software. “If you run a large call center and can cut 75% of calls versus 50%, the business impact is enormous compared to the licensing cost,” he said. Startups, Taylor argued, are surging not because they’re cheaper, but because their products work better.
Yet Taylor is under no illusion that the market will stay this way. As he put it, “the market is like a pendulum.” Once agents are more mature and functional reliability is no longer a differentiator, he expects procurement efficiency to push enterprises back toward platform consolidation. Sierra’s challenge, then, is to become one of those platforms before the pendulum swings back. “Being a first mover matters, but being consistently great over time is what really matters.”
Taylor’s historical framing comes from direct experience. He lived through the dotcom boom and bust as a college student and later joined Google in the early 2000s. He saw how early leaders often gave way to companies that combined innovation with business model durability. The same pattern, he suggested, played out again with the mobile era, when companies like Uber, WhatsApp, and DoorDash rose to dominance.
Now, with large language models enabling applied AI, Taylor sees a similar opening. “These technological shifts change the balance between incumbents and insurgents,” he said. “The opportunity for us as entrepreneurs is to become the next incumbents, which means driving customer success.”
At Sierra, Taylor pointed to real traction. The company is already handling customer service operations for clients including ADT Home Security and Ramp, powering “hundreds of millions” of interactions this year. But he emphasized that the privilege of becoming a lasting platform requires sustained execution. “First-mover advantage matters, but being consistently great over time is what really matters,” he said.
If Taylor’s market analysis is rooted in history, his personal reflections on AI strike at something deeper. Despite decades as an executive, his résumé includes co-CEO of Salesforce and CTO of Facebook, he has long preferred to identify as a programmer. Yet in recent months he has admitted that AI now threatens that very identity.
“The thing I self-identify with [being a computer programmer] is, like, being obviated by this technology,” Taylor said on the Acquired podcast. He likened AI to a productivity booster so powerful it feels like “an Iron Man suit for all of us as individuals.”
It highlights the irony of a programmer predicting the end of programming. The same AI technologies he is deploying at Sierra could make the work of programmers far less central or even unnecessary. “You’re going to have this period of transition where it’s saying, ‘How I’ve come to identify my own worth, either as a person or as an employee, has been disrupted,’” he said. “That’s very uncomfortable. And that transition isn’t always easy.”
The irony is not lost on him. Having built his career through programming, Taylor now sees the role being transformed by the tools he champions. He compared the shift to the obsolescence of “computers”, the human job title for people who performed calculations before machines replaced them.
For Taylor, then, the AI transition is playing out on two levels. Professionally, he is working to establish Sierra as a “best of breed” product that can one day grow into a platform. Personally, he is confronting the fact that AI may erase the craft that defined his career.
Taylor doesn’t linger on what’s being lost. He talks instead about the opportunities. For companies, applied AI is about whether the technology can actually solve a customer’s problem, close a sale, or run a service line at scale. For workers, he describes it as an amplifier of human ability, even if it reshapes the roles people once took for granted.
That dual perspective of being part business strategist, part coder with disruption captures why Taylor remains a closely watched voice in the AI agent space. “AltaVista didn’t write the history books—Google did,” he says.
