They are still getting it wrong on so many levels. The latest push is for everyone to use AI for their jobs.* Figure it out. Where is the common sense, because the communication like this one from Amazon sounds a little insane. Salesforce, Microsoft, Meta, Shopify, have similar stances and the list goes on.
In a June 2025 memo, Amazon CEO Andy Jassy encouraged employees to use and experiment with AI tools to increase efficiency. He made it clear that those who embrace AI would be well-positioned for the future at Amazon, while also warning that the company’s corporate workforce would likely shrink in the coming years due to AI-driven automation.
Through my years at IBM, there were variations on a similar theme. Executive pushes for everyone to do cloud training, design thinking, innovation, etc. From the ground level it was crazy. There were the metrics, managers were called out if their reports did not complete the mandatory requirement on one side and then there was what I called the total disrespect for expertise and experience.
HR staff with masters in their fields after a week in “bootcamp” told engineering we could expedite our MVP, Minimum Viable Product because it’s just like a cupcake without the icing. Or, after 3 days of agile this same group encouraged us to cut down test time, be like Facebook, move fast and break things. Oh, they got it so wrong.
The problem was with any pushback we gave, we were accused of not being agile, innovative, open to change. To a point, this was comical that professionals who were trained in communications, feedback and organizational change were now in our lane without any forethought to use their expertise, but were now emboldened after a few days to dismiss ours.
And no, an MVP is not like a bicycle before the automobile. An MVP is what we call a proof of concept or maybe prototype. So, I’ll stick with Amazon, as a company, they first showed people would buy books online and the strength of rating and recommendations which would provided that “stickiness” that had people return. Now, that has expanded to other things on line.
Such a great success story, we know the before and after and we want to do that to. We like to think there is just one thing we are not doing that can be easily fixed. An app that promises mastery of another language. A fitness membership that extols a new fit body, a spa treatment…OK, I digress or do I? It’s seems to be that same line of thinking that prompts senior executives to believe we can be more profitable, innovative, industry leaders if we jus have our staff do a course.
Let’s train everyone to be innovative, agile, lean, and the list goes on. Some sessions started with, how many uses can you think of for the plastic fork. Or you’ll split up into teams and see who can build the highest tower with marshmallows, dried spaghetti noodles and string. I think tape was included in that. Variations on a theme. Split people into a team and give them a task. But are we training for real life and real problems or party games and ice breakers?
It takes different skill sets to get a product successfully to market. Everyone does not need to be the creative genius. If you’ve done the exercises, you know there is typically the spy, that person that starts to look around at what the other teams are doing. That’s not a bad thing. In product development, there does need to be an awareness of the marketplace. What ideas are out there and in development. There is the leader, this is typically not the creative one, but the organizers that starts to look at how much time is available and this is the person that starts to figure out what has to be done whenever. If there is a presentation involved, they’ve already considered format and started a table of contents. The silent ones. These are ones who know, they are not leaders or says, they get an idea every now and then, but need a little direction and they will give their all. Then there is the skeptic. In a good way, these are ones who listen to a solution and challenge it in a manner that is productive. These are the basic ones. If your team if fortunate, there is a healthy balance of these types.
The issue with a lot of training, these roles, the inherent personality types are dismissed. Sure, you can teach people to sing. It’s a great team building exercise a local business, Musicians Playground does. You split up the group into piano, guitar and singers. They work in vocal, piano or guitar groups and at the end of the exercise, they all come together and perform Don’t Stop Believing. I’ve seen them, the joy, excitement and sense of fulfillment you feel in the room. Cool, cool, cool. However, for productivity and success though, you need to leverage existing and inherent talent. After an afternoon, it’s not likely any of the participants are stadium bound with musical careers. However, it was fun and gave some insight. Some probably followed up with. musical aspirations and others were OK with the experience. This was primarily team building and fun. With an afternoon of some training, no one form this experience would tell a professional singer, pianist or vocalist how to perform. But, it reminded me again, of people with a few days of “training” somehow feeling empowered to tell us, we’re not doing our jobs effectively. With technology training have we confused an overview with execution or do we over estimate our ability?
Consider the people and their epxertise. For example, maybe instead of everyone coming up with a way to use an AI tool, embrace the possibility of listing all the things that cannot be replaced by AI and why. Is there a common pattern, is there an alternative? On the other hand, with the skeptics in their element, they will debate on what can and cannot be done of AI. Give them the opportunity to find the valuable contribution. As for the spy types, let them tell you what others are doing. They likely have acquaintances in other companies doing similar work. Delivery people can tell you delivery hacks, much in the same way, Uber drivers can tell you how they manipulate the system. It may be more useful to understand the hacks and work arounds because let’s be honest, asking people to use AI tools and then saying AI is going to replace jobs, does’t sound very smart. The goal is to decrease cost and increase profit may or may not involve a reduced workforce. In keeping with my occasional theme, use AI, I asked Mildred/ChatGPT
💡 1. Will people use AI tools under this kind of pressure?
Yes, but often with mixed motivation:
Self-preservation: Employees may adopt AI tools to avoid being seen as replaceable. They’ll try to prove they can work with AI rather than be replaced by AI.
Fear-based compliance: Many will use AI tools simply because they’re told to—but may not explore their full potential or use them creatively.
Quiet resistance: Some may resist subtly—downplaying AI benefits, not adopting tools deeply, or even sabotaging implementation—especially if morale is low.🤖 2. Will people try to find ways to use AI for their jobs?
Again, yes, but…
High performers and tech-savvy staff will likely explore AI to gain efficiency, look innovative, or automate grunt work.
Others may feel overwhelmed, anxious, or undertrained—leading to surface-level adoption (e.g., just using ChatGPT for emails or reports) without meaningful integration.
In some cases, employees might use AI to “outsource” parts of their jobs without making it obvious—especially if job loss feels inevitable.
📉 3. What are the risks of a “use AI or be replaced” mandate?
Loss of trust and engagement: People may see AI not as a tool to empower them, but as a threat—damaging morale and loyalty.
Short-term gains, long-term damage: Quick cost-cutting through AI might look good on paper, but risks losing human expertise and tacit knowledge.
Creativity and innovation shrink: Fear-driven cultures rarely encourage experimentation or new ideas—which is ironically where AI could shine most.✅ What works better than fear-based AI edicts?
Collaborative framing: “We’re investing in AI to free up your time for more strategic work” is a very different message from “AI is replacing you.”
Upskilling programs: Companies that invest in AI literacy and reward innovative use cases tend to see better adoption and better morale.
Transparency: Be clear about goals, timelines, and how decisions will be made—especially around workforce impacts.
Bottom line:
People will use AI tools if a CEO mandates it—but whether they do so with enthusiasm, fear, or quiet resistancedepends on how the message is delivered and whether the environment supports trust, learning, and adaptation.
How did this all start? As I changed the battery in the smoke detector, it occurred to me, there are alternative uses for a ladder. I could design and market a “yoga ladder!” This week, think about group dynamics and your role. Are you leader, spy, creator? Does it depend on the situation? Observe, is there an item, an idea, a concept that could be adapted into something new. Or, maybe look at this photo and consider, are there similar products available or why is this not a good idea (hint, the ladder slips : -) Cheers!
*Please note, I am very much for AI. This hypothesis is, with rapid AI adaption, there needs to be a holistic view of the what creates the good customer experiences versus what can be automated. For example, last week, I wrote of the pre popped popcorn. More efficient? Yes, will you sell as much, will it be the same customer experience? I think not. While that may be one of the more obvious examples, it takes knowledge of the process to inform what can be automated versus what should be automated.
Balancing on Unstable Surfaces 