99% of AI Layoffs Had Nothing to Do With AI

RYAN: Here's a number that should make every CEO in America sweat. One in 50. That's how many AI investments actually deliver transformational value. One in 50. Yet companies are laying off people right now, banking on AI gains that don't exist yet. Let me ask this: are we building the future of work, or are we gutting the present one?

DANIEL: That is a crazy statistic. So you're telling me that companies are firing people based on a 2% success rate. That's got to be the most expensive game of make believe in corporate history.

RYAN: That's today's episode. The AI hype hangover is here, and it's hitting your team, your culture, and your bottom line. Let's talk about it.

[INTRO] Welcome to the Make Work Not Suck podcast.

THE BIG WHY

ressive AI growth targets for: 2026

So here's the framing. This isn't about whether AI is good or bad. It's about the gap between what leaders are promising and what workers are actually experiencing.

DANIEL: I'm hearing about this in different organizations around the country. The story kind of goes like this. Imagine your boss comes to you and says the company's investing millions of dollars in AI. And you're like, great, that sounds exciting. But then a couple months later, your boss tells you you've got to cut your team size by 30%. Fast forward six months, the AI tools you invested in still can't do what your team did. And so you're kind of stuck, right?

That's not really innovation, although I'm sure somebody's gonna call it that. It really feels more like organizational malpractice to let people go before you've got a viable plan in place to replace the work.

And over half of chief HR executives don't even know how to prepare for this AI transformation. So if the people responsible for doing the work are gone, how in the world are organizations going to be able to move forward? I think it's kind of crazy.

BY THE NUMBERS

RYAN: All right, so let's talk about some of the numbers. Let's go back to the first one. Only one in 50 AI investments delivers transformational value. That's a Gartner survey from this year that was just published. Gosh. One out of 50 is a 2% success rate.

If I was a baseball player and my batting average was .020, I would not be a professional baseball player much longer. But these are the kind of decisions that professional managers and executives are making. And they've got a 2% success rate. That's craziness.

layoffs in the first half of: 2025

DANIEL: Only 1% of layoffs were caused by actual productivity gains. 99% of them — 99% of AI-related layoffs was AI not working.

RYAN: No, it was because they were assuming it was going to, but it hadn't done it yet.

DANIEL: That's just... you laid off —

RYAN: Hope is not a strategy.

DANIEL: Hope is not a strategy. You laid off 99% of your workforce — or 99% of the layoffs were because we think AI is going to win.

RYAN: Okay, so here's another one. Over 50% of chief human resource officers say they don't know how to prepare for an AI-driven transformation.

DANIEL: Yeah, that's terrifying. Because if the leaders don't know — and honestly, nobody knows right now, so that's not a slam on them — but if they don't know what the future is going to look like, why are they making massive headcount decisions? Why are they making layoffs? And how in the world are the people in the middle and at the front lines of an organization going to succeed when the people at the top have no idea what they're doing? And they admit they have no idea what they're doing.

RYAN: It's like you said. Hope is not a strategy, and they hope AI is going to pay off. It seems good, we're getting these early productivity gains.

DANIEL: Right.

RYAN: But just as it said there, the people in charge of the workforce don't have a plan. What does that tell us about everything else?

SPICY TAKES

RYAN: So here's where it gets spicy. All right, ready for this? Should companies cut headcount now in anticipation of AI gains, or is that corporate malpractice?

DANIEL: Ooh, I've got a strong opinion on this. What do you think?

RYAN: So one of us has to play the bad guy. I'll play the bad guy this round.

DANIEL: Yeah, all right.

RYAN: I'm going to go cut now.

DANIEL: Why?

RYAN: Companies that wait to restructure will bloat when AI matures. And I think that's one of those — it's going to force you to adopt AI faster. So if you want to, it's burn the boats.

DANIEL: Burn the boats. Cut now, burn the boats. Okay.

I think this is reckless. I think you're firing people based on a 2% success rate and calling it strategy. I don't think that's bold, decisive leadership. I think that's just rolling the dice.

The problem I have with your cuts is: who do you cut? Who is it that you're going to lay off? And the fact that you can't answer that question proves that it's reckless. Because cutting headcount to save money sounds easy, but I've been a part of a lot of organizational RIFs, layoffs, whatever you want to call them. And it's terrifying how poorly they are done in some organizations, and how they're letting go people who are actually the only people who know how to do certain things, because the top and the bottom are so disconnected from each other.

And so if these decisions are coming from the corner offices, I have very little confidence that they actually know who's doing what. And if they start taking people out of the equation, how is that system going to operate? You're rolling the dice. It is not going to go well for those people who make that decision.

RYAN: Well, and I got another one for you.

DANIEL: Okay.

RYAN: Employees that start getting paid to train their digital... dang. Their digital doppel— if I said that right. Their digital twin.

DANIEL: Doppelgänger.

RYAN: Doppelgänger. Good Lord, I can't speak today. Your digital twin, I'm just going to go with that. An AI clone of your best work. What does that mean for compensation and ownership? Do I own that, because it's a clone of me? And should I be paid my same compensation if I can get the same amount of work done in half the time? Which, if you remember, that was the original point behind a salary. It was about the amount of work, and you get paid whether you get it done in 20 hours or 60 hours.

DANIEL: I have a friend of a friend — so I don't know him personally, but I was catching up with this buddy of mine and he says that this guy's negotiating a contract right now where he gets paid, I think it's like a three or five year deal. He's in a technology role, but as part of his role, he is building his AI doppelgänger, and he is negotiating residuals from his AI after his three-to-five-year contract's up.

That's craziness. He's going to come in, sit in that role — I think it's like a CTO role — build his digital avatar, and then he's going to leave. But you're still going to pay me for my digital avatar. That's me.

RYAN: It's crazy, but that's very much in line with how movie contracts work, right? Like, I'm going to take salary today, I'm going to perform in this movie, and then I get royalties through end of time.

DANIEL: Yeah. Music, movies, TV. Not everybody gets that deal, but the really talented people do. And I wonder if that's coming for knowledge workers who can really establish some sort of a unique value prop.

RYAN: Yeah, that's a whole other episode. I'm cloning my brain and I'm going to put royalties behind it.

WORK RECEIPT

RYAN: Okay, let's see what this looks like in real life. Let's bring this home into real life stories here.

So I've worked in technical product management for a long time. I know a product manager who's really been trying to lean into AI — transcribing all their internal meetings, building some sort of a knowledge management system to keep track of all this, and was trying to use those conversations to build a product. So they captured all the information, ran it through the AI, was trying to build user stories, got really nice documents, looked pretty, handed them to the engineers... and the engineers just threw up all over it.

It was like workslop. I wouldn't say it was bad, but I would say it was insufficient for the engineers to actually do their job.

And I think this is one of the big problems that people miss. We're seeing all these productivity gains in individual work, right? Like, I do my thing and AI can help me do my thing faster. Great. But your thing is just one thing in a long sequence of things. And if you can't translate your thing to the next person's thing, the whole system falls apart.

And that's where there's a lot of this human interaction, because most organizations don't have those processes well defined. And so when I do my thing and I hand it to the next person and they look at it and they're like, "What the heck is this? I can't do anything with this" — suddenly those AI productivity gains are gone. And you're actually becoming less productive than you would have been if you'd just done it the hard way the first time.

DANIEL: Yep. I think you said it well. It's workslop. Hearing you talk about that backlog build using AI just kind of makes me develop on the other side, because I know AI over-engineered it, or put extra things in there. And it probably sounded good to the product manager and to the stakeholders. But as soon as you got to engineering, it was like, what the hell are we building here? We just need a bicycle and we're trying to build a Titanic. This project is going to sink, because —

RYAN: Why build a Ferrari when you need a go-kart?

DANIEL: You're right. AI is not going to — I mean, unless you're really good at prompting it and you know exactly how to ask that question. And even still, you might disagree. The engineer and the product manager might disagree on how sophisticated the solution needs to be.

RYAN: The practicality here is that AI didn't really help that product manager become more effective, right? If anything, it might've hurt their reputation internally, because they delivered something they thought was really good and then everybody else realized it wasn't. And so they can't do their jobs now. And that's going to reflect poorly on the product manager.

DANIEL: Everybody's got an idea and they give it to AI to produce a result.

RYAN: Right.

DANIEL: The problem is, if you don't know how to use the AI at scale, you're not going to have cultural alignment. Granted, it's AI, so it doesn't have core values. But there's still value behind human core values, because it drives the way we look at things. So if you haven't set those values or principles or boundaries into the AI model, then it's going to be a miss.

And the other thing too is, it's going to assume the journey. Again, unless you can feed it the documentation and feed it the processes and feed it all that stuff. But now we're going back to our original question of who owns the IP of unloading their brain into the AI.

RYAN: And who do you know that has that documentation, those processes documented and up to date? Even the places I've seen that have it documented are like, "Yeah, we don't do it this way anymore." It's like, okay. So even if I fed it into the AI, it's still not going to be true.

DANIEL: Right.

RYAN: And it's that kind of undocumented connective tissue, especially in larger organizations. I don't know how you articulate that to an AI.

THE MONDAY MORNING FIX

RYAN: So what do we actually do about all this? I think we've got three options to throw out here.

Number one: audit before you cut.

DANIEL: Yes.

RYAN: Simple, right? Before making any AI-driven headcount decisions, demand proof of value. Not a projection, not a vendor demo. An actual business result. If the AI can't prove it's doing the job better, cheaper, or faster than real data or a real human, the headcount stays. Plain and simple.

DANIEL: I think workslop is probably the second big fix that we need to be cognizant of. And I really like the idea of bringing back the wheel. I think maybe that becomes the framework that you use to train your AI in, right? Here's the vision, here's the journey, here's the culture, here's the results, help me get through this. Because if you're just going from vision to results, you're going to get crappy output from the AI.

And whether you realize it's crappy or not, somebody down the line will know. So you've got to force the AI to go through the process in the right way, because the point is not just, is this faster? The point is, is this better? Are you proud to put your name on it? Or did you just throw this into ChatGPT and hand it off?

RYAN: Now let's ask the uncomfortable question.

DANIEL: Uh oh.

RYAN: In your next team meeting, ask: if we fast forward to December and we fall short of all of our goals, or our major goals, which of our AI bets would you say got in our way? If nobody can answer that, you're flying blind.

DANIEL: Oh, I like that. It's almost like a pre-mortem of, like, if we get there and we haven't won, why? What broke? What didn't turn out the way we were expecting it to? And that assumes you were making very conscious, intentional bets. And I think that's probably not true of most organizations. I think they're just winging it.

I might take that to some of my clients next week. That's a good question.

THE LAST WORD

RYAN: Here's what we want you to take away from this. Here's the one thing I want everyone to take away: AI isn't the enemy. Hype without honesty is the enemy. The companies that win this era won't be the ones that adopt AI the fastest. They'll be the ones that adopt it the smartest. And that starts with treating your workforce like the competitive advantage it actually is today.

DANIEL: Value your people, produce AI, be smart about it. You'll be the company that survives this.

My challenge to you, to me, and to anybody who's listening: this week, go find the process person. Go find the person on your team who really knows the process better than anybody else. The person who actually knows how to make the work work. And ask them how AI is impacting their job. And listen to what they have to say.

Because Gartner says that those process people — not the tech prodigies, but the process people — are the ones who are going to unlock the real value. And if you let them go, you are in a world of trouble. So it's good to not only keep them, but make sure they understand how valuable and appreciated they are.

RYAN: All right. If this episode hit home for you, share it with the leader that needs to hear it. Drop your take in the comments below. And if you haven't subscribed yet, well, you know what to do. And we'll see you on the next episode. Thanks.

[OUTRO] Make Work Not Suck. Our podcast that talks about exactly that: our process, vision, journey, culture, and results. We present real world business solutions that make the difference. Our goal is to make work not suck. Hosted by Ryan Hodges, co-host Daniel Steer. Join us each episode, and make work not suck.