The Real Reason Your CEO Wants to Replace You With AI

Relax. Your CEO isn’t going to replace you with an AI chatbot. Not most of you, anyway. Not now.

Maybe some CEOs haven’t received the ChatGPT-generated memo yet, but as the AI sales cycle cools on generative AI and starts warming to AI automation and agents, most CEOs have figured out that a large language model without a business expert guiding it is just like having a dictionary they’ll need to yell at until it reconciles their quarterly results. 

Yelling at inanimate objects only works in rare instances. Any software developer will tell you that. 

That said, the cycle shift to AI automation and agentic AI — think of it as AI that “does” things instead of just “says” things — opens the door for CEOs to come after a new cohort of employee, the employee who uses data to make decisions, and then executes those decisions. 

But are CEOs going to repeat the same mistakes they made when they implemented generative AI — buying into not-ready-for-prime-time tech, firing everyone, and then sorting it out later?

I don’t know the answer to that question. So let’s ask some CEOs. 

Buying Into the Transformative Story

“I don’t know the specifics of our ROI on AI. It’s probably around here somewhere.”

That quote comes, somewhat facetiously, from “Mike,” a notoriously laid-back and self-aware CEO of a cutting-edge tech company. Mike isn’t a chaser of new tech trends, but he does enjoy the benefits once that tech becomes useful. Thus, he didn’t mandate the use of AI across his company, but instead relied on his tech team to guide the tech side of the business. 

That’s a method that’s been working for him for years, since before AI, and he’s not about to change it now.

“No,” he said, much more seriously. “We’re not buying into the … transformative story of AI. We didn’t then and I don’t see any reason to now. My team has been on top of the advancements as they’ve been happening. I have a high-level understanding of what we’re doing, and that’s all I need.”

I would agree. Especially after talking to “Kurt,” another CEO who reached out to me a few months back, struggling with a lack of return on a hefty investment into “vibe coding.”

“Honestly, I let it get away from me,” he admitted. “There’s a ton of pressure [from his board] to do more with less. We ended up with a lot more, but no one can tell me why it’s not showing up in top-line growth.”

He paused and genuinely looked hurt. “It’s like we’re making stuff just to make stuff. It’s not additive,” he said. “I’m not doing that again.”

Pity the Poor CEO

If you’re like me, you feel for Kurt.

But whom am I kidding? In this regard, I know for sure that a lot of people don’t have empathy for CEOs, especially those who totally bought into the chatbot-as-replacement — whether for customer service reps, HR screeners, product managers, or expert coders.

“What keeps me up at night,” said Tanya, a tech startup founder, “is — now that we’ve stabilized the burn, how do we get back to growth? My team is maxed out, even with 100 percent AI adoption, and nothing on the roadmap I believe in as a growth catalyst.”

Last year, Tanya’s investors made it very clear that they weren’t investing more until she could demonstrate both profitability and a growth rate well into the “mid-double digits.” 

“I hear it every month,” she said. “We’re not growing quickly enough — where are all those AI productivity gains?”

The solution is implied, not direct, she told me. She needs to keep cutting people, extend the runway, and buy time to hit that growth rate.

“I want to joke,” she laughs. “Or at least be a little sarcastic. Do you guys expect Claude to make sales calls and close deals? But I’m afraid of what the answer will be.”

What CEOs Are Being Told

One of my IBM insiders — OK, a guy I know who works at IBM — pointed me to this bit of CEO guidance, published back in April, full of gorgeous advice for how to navigate the choppy waters of AI-driven change.

I showed it to Mike first, on a video call, and enjoyed his skimming, his chuckling, and his repeating of his “favorite” phrases. 

“Oh, good,” he said. “I don’t have to be clairvoyant. Good to know. I was worried there.”

Then later:

“’Transformation’ … ‘doubling down’ … ‘disruption’ … ‘transformational impact’ … ‘transform faster,’” he said. “Please tell your readers I said ‘transformative’ before reading this.” 

Kurt gave the study a quiet five minutes and just shook his head. “There’s so much double-speak in here, and I’m not surprised,” he said. “‘Stay safe by taking on more risk.’ Oh, ‘creative destruction.’ I keep hearing that from my board. That phrase exactly.”

And Tanya, while she didn’t hear Kurt’s point, nevertheless drove it home. 

“This is exactly what I hear from my board every month,” she said. “It’s like a … dare. Like a game of … is it chicken? I’m not succeeding with AI because I’m not brave enough to ‘embrace’ it. It’s … yeah, it’s insulting. But what should I do?”

Good question, Tanya.

The Problem Is Still Profit and Loss

There’s a playbook that boards and investors have been following since all that post-pandemic money dried up and we went from cheap money to hyperinflation. We’re still trying to fix that. The demands for profitability were quieted at the kneecapping of future growth. 

Now that the books have been balanced, growth is stagnant and inflation is still an issue. This is the paradox that needs to be solved, and in some sense, AI was always the magic wand that was going to make it all go away.

CEOs and company leaders weren’t ready to wave that wand. So they just kept repeatedly smacking employees over the head with it.

“It was capitulation,” sighed Kurt. “Try anything, because the mistakes were already made and the hole was too deep. In leaderspeak, ‘try anything’ becomes ‘do whatever it takes.’”

So CEOs leaned way into generative AI, fingers crossed, and hoped for the best. As long as the jury was still out on the timeline for ROI on that AI implementation, there was another story to tell to get them through another month.

“That sounds so depressing,” said Mike. “Thank God we’re bootstrapped.” He laughed, “None of what we’re doing fits into a magic quadrant.”

We’re starting to get the real story on AI productivity, and it’s not a horror story. But frankly, I wonder if the resentment that festered the first time around is going to cause good use cases to be thrown out with bad.

Mike has the right AI strategy. He let his tech team lead on tech. He doesn’t mandate AI adoption. He doesn’t chase an “AI-first culture.” He trusts the people who understand the technology to use it when it makes sense. 

So here’s my advice to CEOs being pressured to replace people with AI:

Don’t.

Not because AI can’t do the work. But because if you don’t have people around you who know specifically which work AI should be doing — and how — you’re not implementing AI. You’re capitulating to pressure and crossing your fingers that things will work out by the end of the quarter.

I’m trying to untangle the knots of tech with a lot of experience and a little bit of humor. If that appeals to you, please join the tens of thousands of tech professionals on my email list.