Gregg Bayes-Brown

Most AI programmes fail. Here's why.

Somewhere between 70% and 95% of enterprise AI initiatives don't deliver what they promised (BCG, 2024; MIT NANDA, 2025). The usual suspects get blamed — bad data, wrong tools, lack of budget. But the pattern underneath is simpler and harder to fix: the people using AI haven't changed how they think.

That's what cognitive enablement means. Not training people on software. Not running a prompt engineering workshop and hoping for the best. It means upgrading the thinking that sits behind every interaction with AI — how you frame problems, how you evaluate outputs, how you know when the machine is helping and when it's making you worse at your job.

Bayes Ltd works with individuals, teams, and organisations to close that gap. One-to-one coaching for executives. Team sessions that build real capability. Company-wide programmes that actually stick. The method is Bayesian — start with what you believe, test it against evidence, update, repeat.