Applying AI to existing processes that are human optimized will have a lower ceiling of returns than redesigning the company's org structure and processes from the top down and the bottom up.

The stat that stuck with me isn't the 13%. It's that only 4% have plans to redesign how people work alongside AI.

My company went the other way: gradual rollout, capped plans, ongoing training. The cap looked like a constraint but turned into a filter, because you have to decide where it actually pays off.

"Volume is not value." Agreed.