AI Can Fix Work—But Are We Solving the Wrong Problem?
New research from Microsoft's WorkLab highlights the degree to which the modern knowledge worker never truly switches off from work, with early morning and late evening emails, workdays full of interruptions, and up to 20% of workers using their weekends to catch up.
This reality that is a core conundrum of modern work will come as no surprise to many workers, and, of course, Microsoft's conclusions to this fact finding heavily push toward the solution being different aspects of AI adoption (both the blog post related to this and research article are linked in the comments).
Yes, AI can address these issues. But the question we should be asking is are these inefficiencies actually the problem that needs to be fixed or are they merely the symptom of other issues, where patching with AI adoption is like taking regular doses of Ibuprofen for a headache but just continuing to regularly headbutt a brick wall.
Yesterday, I had to visit a hospital. They had installed a brand-new food hall, driven by central banks of touchscreen ordering terminals. You ordered your food, got a ticket, and picked it up in a box. No human interaction needed. And the workers now, instead of taking your order, chatting, and being part of the experience, were reduced to productivity-focused waffle makers and sandwich builders. Once we have robots capable of this style of food production, this location is 100% ready as the soul has already been ripped out of the experience. I would bet that the reported job satisfaction of those food service workers has reduced with this change even if it's increased productivity.
In the office, Microsoft's take on how to mitigate this knowledge work creep is to focus on the 20% of work that is most impactful and to "ruthlessly streamline the rest".
Of course, I agree that AI technology gives us a new range of tools and capabilities to make work better. However, I'm not sure a workplace where anything that isn't directly productive is seen as an enemy to cut down and conquer is the right approach to delivering improvement across the board.
Just as there is soul in allowing food service workers to engage with customers, even if objectively an order terminal and constant feverish production is more efficient, there is also soul in the wrinkles of office life. This is where learning is done, and connections are formed, and resilience is grown.
In making work better, we must define what better looks like and which parties measure that better. This, in my mind, is one of the most important aspects of our AI adoption path. The industrial revolution was not just revolutionary because of new technology; it was revolutionary because it fundamentally shifted how we live and work. AI has the same potential to transform everything.
One of the transformations that might make sense is for AI to radically change our work patterns, not just patch the always-on workday around the edges! (See a video in the comments)
🖼️ from Microsoft's blog post
First published on Linkedin on 06/29/2025 -> click here to view