McKinsey says 57% of US work hours could be automated with today’s AI. Stanford found only 2% of workers want their entire job replaced (well, duh).
But scratch deeper and there’s something useful here. The 57% is technical potential. The 2% is human reality. One measures what AI can do. The other measures what people will actually tolerate. Guess which one matters more for adoption.
We’ve been here before
We’ve all experienced good and bad “digital transformation” projects. The difference? People. How they brought context together, made it relevant for the user, and ensured change actually landed. BAs and PMs were the glue that brought successful strategy together.
‘Agentic AI’, ‘copilots’, ‘AI-powered workflows’—same same. Trust is hard earned and very easily lost. Especially with today’s AI slop.
Visibility, not just oversight
With the volume of questionable AI output out there, the need for supervision is clear. But knowing you need oversight is one thing. Proving it’s working is another.
How do you show your leadership what's actually being adopted, where it's failing, and whether it's worth the investment?
You need visibility. Metrics tied to outcomes. A way to track adoption that isn’t just vibes and anecdotes.
Without that, it's hard to justify the spend—or course correct before it's too late.
The point
Will AI replace jobs? Maybe parts of them. But strategy without successful adoption is just another tool that’s reluctantly used—and they’ll always find a problem with it.
The organisations that win will invest in the people who make change stick—and the visibility to prove it.