AI replaces jobs - how to win at AI adoption

2025/06/17

IBM, what are you doing? You set the blue standard for the tech industry. And then you say something like this? Using AI to replace 8,000 jobs. That’s a head scratcher.

I’m a big advocate of AI. I believe that AI should find its way into every ecosystem we navigate and live in. It should be in our personal lives. It should be in our business lives and academia. When AI is used right, it helps with mental health. It helps bring out the best in us. It helps us be more productive and generate more value. There are a lot of wins for AI. But you wouldn’t think it when you hear about the company, a brand name company, Big Blue, saying that they’re using AI to replace people, that AI is the job replacer.

Tech layoffs 2025: IBM lays off 8,000 employees as AI replaces HR department (Business Today)

IBM did it

People are going to take that very seriously for what IBM did, because if it’s good enough for IBM, it’s good enough for us. I saw this in Business Today, where they talked about IBM using AI to replace 8,000 jobs. Now they were HR jobs. But no one’s going to care about that nuance. They’re just going to see IBM used AI to replace jobs. AI replaces jobs. That’s going to be the first conversation when people adopt AI. And that’s no good. You lose. You stifle a lot of innovation, a lot of discussions from opening the conversation that way. And who in their right mind is going to accept a technology where they’re quite literally training their replacement?

They’re not going to do that.
They’re not going to buy in.

justifiable mistrust and fear

You’re going to see a lot of resistance to AI adoption if that’s the first go-to conversation. AI makes you more productive. AI multiplies your impact, helps it scale. It helps accelerate you and your activities, especially when you’re working in tandem with people. AI does all of that, but you wouldn’t hear or think of it when you read stuff like what IBM is doing with AI and replacing jobs with it. Now, this is creating a lot of fear because people will not trust AI. They will fear AI because they’re scared it’ll take their job. And this fear is not just from people adopting AI at the lower levels. CEOs are getting it too.

The vast majority of CEOs are fearful of losing their jobs due to AI, survey reveals (Fortune)

I was surprised. I didn’t realize the pressure was so hot for CEOs. I saw it in Fortune. It was a survey, 80% of the CEOs felt they’re going to lose their job if they don’t adopt AI in the next two years. Really? Two years is not a big window to make sure that you’ve proven your mettle with AI. It takes a little bit longer than that. You know what people do to prove their mettle that AI works? They do process redesign. The AI may be involved in that, but it’s not the AI that contributed to success. It was the process redesign that did it. And then there are other companies under that type of deadline for an AI project or a program. They pump it through. And then it goes off the rails. It doesn’t meet expectations, and that’s putting it mildly. It’s not good to put CEOs in a position where they feel pressure, where they need to change the organization just for the sake of changing the organization. The best thing to do is to pick your battles and work AI in where it makes sense. And with a two-year timeline, that’s very hard to do.

misaligned mindset

With AI adoption, successful AI adoption, you have to turn the temperature down on the fear. And slow things down. AI is moving too fast, and every time you figure out what AI is, it’s changing a little bit more. You need to get ahead of the curve, and that requires you to do a little more thinking and a little more planning of how to better change your organization. Just slapping in AI and expecting it to serve you might not be the best option for successful AI adoption. Now you think that it would be just digital transformation. That’s what everyone seems to be saying. Digital transformation with culture, business, and financial outcomes. But that’s not what AI is. AI is not a digital transformation. It may involve it, but it’s much more than that.

Unleashing AI: Canada’s blueprint for productivity, innovation, and workforce integration (Macdonald Laurier Institute)

But the think tanks. The experts. They’re all singing to the same beats on what AI is and how to handle successful AI adoption. And it doesn’t matter. US, Canada, Europe. They all follow the same beats, just like McDonald Laurier Institute. They mentioned human-centric design, process redesign. and having people involved in their own AI adoption. And that’s highly digital. AI is automating manual steps. AI is automating decision-making steps. Repetitive. Repetitive decision-making. Repetitive manual steps. And this is a very different thing from our previous kinds of technology. It’s automating thinking. That means AI fits within culture because anything involving thinking changes the way people believe, how people interact with each other, how people plan, and how people behave. And because there is automated thinking along with a lot of automated manual steps, AI is really good at strategy optimization, process optimization, which makes it fit nicely with business transformation. So AI, successful AI adoption has aspects of Business, Culture, and Digital transformation. It’s BCD transformation. And that’s not being talked about very much. Instead, those other transformations, business and culture, are glossed over and digital is emphasized. Instead, business and culture are showcased as outcomes. That’s going to cause you some trouble from narrow thinking. You can only get ahead of the curve when you think about AI in those three transformation viewpoints. And that can also help you anticipate how AI, even if AI is changing how AI will work better for your company. BCD helps you see the future to help navigate how your company can better adapt. You need to focus on those three lenses.

you will lose something

The big thing about adopting automation is that you lose something. It doesn’t matter what kind of automation it is. You lose something when you adopt. It’s never win-win-win. You lose something. You always lose something. The questions are —

do you know what you’re losing?
are you happy with the loss?
can you live with it?

A lot of people don’t have that discussion about AI. For instance, you’re farming out work. You give work to your subordinates. It could be portions of a report to help them clean up, a presentation, an analysis. You could even argue that includes doing stuff for a project.

McKinsey Is Using AI to Create PowerPoints and Take Over Junior Employee Tasks: ‘Technology Could Do That’ (Entrepreneur)

Well, you can use large language models to automate a lot of the work that they’re doing. But when you do that, and they’re no longer doing the work, not receiving practical experience, you’re not developing your people. You lose that. What can you do? How can you replace that? You can’t replace it with training. You can’t replace it by hiring the right people or people with the right education. You just can’t, like it’s gone. You need people to actually do real-world stuff so they build real-world experience. And you lose that when you automate what you’re farming out.

Or if you’re a manager and you have some subordinates who you want to determine if it makes sense to move some more responsibility to them. But how do you know the work that they’re producing is done by the employee or done by the AI itself? How do you determine that? You need to be an expert in that AI to do that. And based on how many different things are around and how fast things are changing, that’s a tall order. I’m not sure how you could even consider doing that. You don’t know how to gauge people, to assess people. It becomes extremely difficult.

Dream job no more? AI is coming for Wall Street’s entry-level Junior Analyst roles, and experts say it’s just the start (Economic Times)

And then your internal conversation is going to be that you can’t really tell the difference between people producing this work using AI versus the work that they actually produce if they didn’t use AI. You can’t tell the difference.

That makes you pose the question, “does it make sense for me to keep this person aboard?” Because the AI can do just as a good job. I don’t want to have that discussion, but it’s sure going be set up that way with the way people are using AI and not putting guardrails around it.

Now there’s a question about transparency, and that’s a big one. Risk management is the big talking point when you talk to people about AI, especially when in it’s same vein as ethical AI. The thing is, you’re automating a process, a series of steps, using AI. And those steps are glossed over. When people were doing the steps manually, they were manually course-correcting, using their subject matter expertise to make sure they were reducing risk. You don’t really replace that with automation. People don’t know what those points were, what those course corrections were. And so you’ve got to figure out what those things were at the end, after you look at what the AI produced. And that’s a tall order. You’ve got to have a lot more specialized expertise, a lot more knowledge, and maybe even manpower to be able to mitigate risk. So you lose something when you automate.

Sometimes it’s good not to automate because the cost is too high.

wrapping up

We have to reframe AI from being the ultimate job replacer to being something that maximizes our productivity and our value. We have to slow things down. We have to reduce the fear and treat AI as a BCD initiative, a BCD transformation. And only then will my dream of having artificial assimilation as a cornerstone in our daily lives.

Resources