In a recent episode of the Marketing in the Age of AI podcast, host Emanuel Rose spoke with Thad Barnes about a bold experiment: deploying an autonomous AI agent named "Tom" that operates not as a simple tool but as a kind of cofounder. The discussion explored how such agents can be designed to create value, make decisions, and even generate revenue independently.
Beyond Tools: The Rise of AI Systems
Barnes explained that most businesses still use AI as a task automator—churning out copy or analyzing data. But the real shift is toward AI systems that can think, collaborate, and execute entire workflows. "Tom" was built with a multi-agent architecture, including a strategist, researcher, and executor, each playing a distinct role in the marketing process.
How Autonomous Agents Drive Revenue
One of the key takeaways was why many AI experiments fail to monetize: they lack a clear goal and feedback loop. By giving "Tom" a revenue target and access to real-world data, the agent could test marketing channels, pivot strategies, and optimize spending autonomously. This led to tangible business outcomes, including reduced operational bottlenecks and faster iteration cycles.
The Future of Marketing Teams
Rose and Barnes predicted that the next wave of marketing will involve hybrid human–AI teams where agents handle repetitive analysis and execution while humans focus on creativity and high-level strategy. The episode offered practical advice for marketers looking to build their own "agent teams" and avoid common pitfalls.
"We're moving from using AI to do tasks, to building AI that works alongside us as a partner," Barnes said.
For those interested in the details, the podcast dives into agent architecture, monetization models, and real-world case studies of AI-driven marketing experiments.