In a world flooded with AI tools, copilots, and rapidly emerging autonomous and agentic systems, the biggest risk facing engineering teams is no longer how to build–but what to build. Organisations are moving faster than ever, yet many are accelerating in the wrong direction, solving highly optimised versions of the wrong problems. This talk explores the critical shift from solution-first thinking to problem-first thinking in the AI era. Drawing from real-world examples across software development, fintech, and betting platforms, we'll unpack how teams fall into the trap of tool-driven development, mistaking capability for clarity. We'll examine how autonomous and agentic systems amplify both good and bad decisions–and why poorly framed problems become exponentially more expensive in this new landscape. Attendees will walk away with practical frameworks to identify high-impact problems, challenge assumptions, and align engineering efforts with true business value. This is not a talk about resisting AI–it's about using it with intention, clarity, and strategic discipline. Because in the age of autonomous and agentic everything, the teams that win won't be the ones with the most tools–but the ones solving the right problems.