You've probably heard this advice: "Don't use AI to solve problems. Use it to achieve goals you couldn't reach before. That's where the real ROI lives."
I've said versions of this myself. There's truth in it. But it's also missing something important.
The truth in it
Companies that use AI only to make existing work 10% faster usually get 10% returns. The big wins come when AI lets you do something that wasn't possible before.
Netflix didn't use AI to fix a broken process. It used AI to become a different kind of company.
Tesla didn't use AI to build cars more efficiently. It used AI to turn the car into a learning platform — a vehicle that gets better after you drive it home, collects data with every mile, and unlocks new features through software. Tesla isn't solving a car manufacturing problem. It's changing what a car company sells.
That's where transformational ROI lives.
What the advice gets wrong
It forces a false choice.
The best AI strategies don't pick between solving problems and chasing new objectives. They do both — on purpose, in the right order.
The "boring" automation projects are what fund the ambitious ones. They build your data foundation. They train your people. They earn boardroom credibility. Skip them, and your moonshot has no runway.
There's also a quieter issue with the ROI claim. Transformational AI doesn't reliably deliver huge returns. It delivers higher potential returns with much higher risk. Most enterprise AI failures are ambitious bets, not automation projects.
A simpler way to say it
The mistake isn't using AI to solve problems.
The mistake is using AI only to solve today's problems.
Build a portfolio where the small wins fund the big bets. Where every automation project is also building the data, skills, and confidence you'll need when it's time to attempt something truly new.
That's where the real ROI lives — not at either extreme, but in the bridge between them.
In my opinion, the real value of AI is not only in solving current problems. The real value of AI is in helping organizations achieve new objectives that were not possible before with traditional technologies.
Magdy Mansour ↗
Curates Chief AI Officer and contributes perspectives on AI strategy, manufacturing, procurement, cybersecurity, and practical business value.
