I saw an article in the New York Times this week that rang so true to me. I found myself nodding along to almost every point it made, which is rare. I see a lot of pieces where I agree with parts, and then either shake my head vehemently at other parts or at least make faces wondering why they ignored so many relevant facts or failed to complete the story.
AI Wishing and AI Washing
The writer, a former CIO of Lululemon, made 2 key points.
First, she noted what she called AI Wishing – the belief by company leaders that AI is magic. There has been a steady stream of new AI companies that claim their product can solve a huge array of business problems with what looks like magic, not to mention consultants claiming to be able to deliver similar feats. And they sell it like it’s magic, with flashy demos that ignore the complicated and tedious behind-the-scenes data cleansing and process changes required for the product to work properly.
Second, she discussed AI Washing – the phenomenon of companies claiming to be doing more with AI than they actually are. There are lots of theories for why this happens, but the one that rings the most true to me is the version where they do it because Wall Street rewards the behavior. I’ve seen it over and over, and experienced it firsthand while I was at Microsoft. For any publicly traded company, saying they’re using AI is good for them. It’s the new hotness that’s been valued for its association with streamlining, cutting costs, and keeping up with the latest tech. When they pair it with a layoff announcement, suddenly the layoff isn’t an admission of a leadership mistake, it’s the unfortunate side effect of progress. The stock gets a bump, the executives get a bonus. For companies that are building and selling AI models and tools, they’re basically required to say they’re using AI, and laying off people as a result, as a sales tactic. “Look at how many people we’ve been able to lay off because we’re using our tools internally. You should buy them from us and get the same benefits.” So the logic goes. Whether they’re actually getting those benefits, or just overloading the layoff survivors, is a whole other thing.
Most companies are failing at AI, and they know it
The through-line is that AI is not magic, but a lot of companies are acting like it is. The data tells a different story: study after study has come out showing that the majority of companies that adopted AI have not seen any substantive ROI. Anywhere from 60-95% of those companies (depending on the study) see no real value. And yet a majority of them continue to invest, even acknowledging that they’re doing so more as a performative act.
How to do it right
My thing is… what if adopting AI can actually predictably deliver ROI? And what if you could get that value and not lay people off? And what if all of this meant your company can grow without adding more headcount or burning people out?
The difference is in how you design your strategy and how you roll it out. There’s a wrong way, and a right way. The right way starts with acknowledging that AI is not magic and that rolling it out requires the same rigor as any other initiative.
After 25 years at Microsoft rolling out changes of all kinds and designing countless software updates, I work with my clients to roll out changes the right way. First, identify the problems to solve. Measure how impactful those problems are, and what objective metrics would show that the problem has been solved. Then design changes that solve the problems. Work with the team, help them see the changes as good for them, and not a risk to their job. Build the changes, test carefully, roll them out. Train the team, make sure they know what to do. Measure the results and iterate if needed.
It’s not magic, it’s just work. It’s worth it to do it right.
Here’s the article if you want to it for yourself:
https://www.nytimes.com/2026/08/03/opinion/ai-hype-tech-layoffs.html