
Why WPP’s AI Boss Believes Agents Are Still in the ‘Teenage Sex’ Stage of Development
Companies Mentioned
Why It Matters
The gap between expectation and deployment threatens wasted spend and stalled innovation, urging marketers to adopt rigorous testing and focus on genuinely adaptive AI solutions.
Key Takeaways
- •Agentic AI hype outpaces actual deployment in ad agencies
- •80% of effort goes into testing, not building agents
- •Models become obsolete as agents alter consumer behavior
- •Current AI in marketing is rule‑following, not adaptive
- •Marketers should prioritize problem‑focused, differentiating AI solutions
Pulse Analysis
The buzz around agentic AI has surged across the advertising ecosystem, with vendors touting autonomous campaign‑creation platforms and AI‑driven operating systems. WPP’s Daniel Hulm, however, cautions that the excitement is premature; most agencies remain in exploratory pilots rather than full‑scale rollouts. This disconnect mirrors past technology cycles—programmatic buying and data clean rooms—where industry optimism outstripped operational reality, leaving a legacy of under‑utilized tools.
A core challenge lies in the "second‑order gap" Hulm describes: once an AI agent begins making autonomous decisions, it reshapes the very consumer behavior it was trained to predict. The resulting feedback loop can render models obsolete within days, demanding continuous testing and rapid iteration. Hulm estimates that up to 80% of the energy required to deploy agents is spent on validation rather than development, highlighting the need for robust governance frameworks and real‑time performance monitoring.
For marketers, the takeaway is clear: chasing quick wins with superficial AI will not yield sustainable advantage. Companies must shift from rule‑following automation to systems that learn, observe outcomes, and adapt dynamically. Investing in problem‑centric AI, partnering with specialists who can deliver scalable testing infrastructure, and aligning AI initiatives with measurable business objectives will differentiate early adopters from those stuck in the teenage‑sex phase of hype.
Why WPP’s AI boss believes agents are still in the ‘teenage sex’ stage of development
Comments
Want to join the conversation?
Loading comments...