In the bustling corridors of Davos, where global leaders converge to dissect the next technological frontier, Bret Taylor, co-founder and CEO of Sierra and chairman of OpenAI’s board, offered a measured yet bullish take on artificial intelligence’s trajectory. Speaking on CNBC’s Squawk Box (link), Taylor likened the current moment to 1996’s internet rollout: “We’re at the beginning of this curve though… It’s somewhat self-evident. It’s going to have a huge impact.” His optimism stems from Sierra’s rapid ascent, a startup now valued at $10 billion after raising $350 million, as reported by CNBC (link).

Taylor, whose resume includes co-creating Google Maps and co-CEOs at Salesforce, emphasized applied AI solutions over raw model hype. Sierra targets customer service, replacing clunky IVR systems with conversational agents that handle calls autonomously. “Companies don’t necessarily want to buy technology, they want to buy solutions to problems,” Taylor told CNBC hosts. This focus addresses enterprise pain points, where adoption lags potential due to immature off-the-shelf tools.

Early wins are evident: Sierra powers over 1 million outbound sales calls monthly for Rocket Mortgage, showcasing scale in real-world deployment. Taylor highlighted peers like Harvey AI in legal workflows, predicting a surge in startups tackling “boring but important” back-office processes such as KYC checks and vendor onboarding.

Agentic AI’s Enterprise Edge

At the heart of Sierra’s strategy lies agentic AI—autonomous systems that span departments and systems of record. Unlike consumer tools where users flit between ChatGPT and Gemini, enterprise moats are stickier. “The atomic unit of an AI agent is actually a process,” Taylor explained on CNBC, noting how agents disrupt traditional software by shuffling anchor tenants from legacy platforms to autonomous workflows.

This shift raises questions about incumbents like Oracle. Taylor views agents as disruptive rather than directly competitive: “It’s more disruptive than it is competitive.” Agents access untouched systems during tasks like fixing a DirecTV receiver, pivoting digital strategies toward autonomy. Recent partnerships, such as ServiceNow’s collaboration with OpenAI for enterprise-grade agents (SiliconANGLE, link), underscore this momentum.

Switching costs in this domain hinge on process integration, not just data lock-in. As Taylor noted, agents evolve the software stack, potentially creating new systems of record for core operations like customer experience and financial audits.

From Hype to Pragmatic Rollouts

Taylor dismisses bubble fears, echoing his Bloomberg interview where he stated Silicon Valley will weather the storm (link). “When everyone knows that AI is going to have a huge impact… you end up with smart money, dumb money,” he said on CNBC, forecasting correction and consolidation. Sierra’s $4.5 billion valuation earlier, per CNBC (link), reflects investor appetite amid plenty.

OpenAI’s 2026 pivot to practical adoption aligns here, with finance chief Sarah Friar prioritizing enterprise-scale deployment (StartupNews.fyi, link). Taylor, in a recent Yahoo Finance discussion, stressed AI safety and continuous improvement alongside Sierra’s model (link).

Software engineering tools like OpenAI’s Codex and Cursor are already taking off, per Taylor. Customer service leads applied AI, with Sierra’s agents featuring memory for persistent interactions, as detailed in his CNBC appearance on agent data platforms.

Conversational Interfaces Redefine Workflows

Large language models enable computers to converse via chat or voice, evolving from punch cards to touchscreens. This powers sales, legal reviews, and analytics. Harvey’s legal focus exemplifies domain-specific agents, reducing build-from-scratch barriers. Sierra’s platform integrates seamlessly, handling complex queries without model tinkering.

In a Verge podcast, Taylor elaborated: “OpenAI Chairman Bret Taylor on why AI means you won’t ever be stuck on hold again” (link). Posts on X from @bretaylor reinforce this, highlighting agentic memory and enterprise pilots, signaling real-time buzz.

Enterprise adoption accelerates in 2026, with predictions of multi-agent orchestration and reliable agents dominating, per Forbes (link). TechCrunch forecasts pragmatism: new architectures, smaller models, and physical AI (link).

Navigating Incumbents and Consolidation

Legacy players can pivot; agents complement rather than supplant. Reed Hastings’ quip about Netflix competitors like Fortnite illustrates non-obvious threats. Sierra doesn’t knock out Oracle but reorients strategies around autonomous tasks. CMSWire notes Sierra’s $10B rise signals conversational AI’s turning point (link).

Competition fosters innovation: “You can’t get innovation without that kind of messy competition,” Taylor said. Free-market dynamics will sort winners, with startups potentially consolidating layers. Sierra reviews peg pricing at $150k+, positioning it for mid-market to enterprise (ServiceAgent.ai, link).

OpenAI’s structure as a major foundation amplifies its mission, per Taylor amid Elon Musk’s lawsuit. He called it “baseless,” focusing on building amid trial prospects.

2026’s Agentic Horizon

Looking ahead, Taylor envisions agents auditing financials or managing supply chains. Sierra’s memory-enabled agents, launched recently, handle multi-turn interactions. OpenAI’s enterprise push with ServiceNow targets similar ground. As Davos buzz fades, execution defines leaders.

Bret Taylor’s dual role positions him uniquely: OpenAI advances foundational models, Sierra applies them. Enterprise AI matures from experimentation to deployment, promising productivity leaps across industries.