Despite the hype, despite the clarion call to action, despite the mandates, roughly half of the individual contributors in the U.S. aren’t using AI very much, and about half of those folks aren’t using AI at all.
Half. No or not much AI. So sayeth Gallup.
The article mentions some predictable signals of lack of adoption, like role differences or not having a manager who’s supportive. But this makes it sound like if you just pulled the right organizational levers, there would magically be humans and robots working in harmony.
Those are indeed important factors, and I just finished an entire series about it. There’s more to the story, though.
If you’re a white-collar worker who’s paying attention — any attention at all, positive or negative — you know where this AI thing is heading. You’re also quite aware of what you need to do if you’d like to keep receiving a paycheck. That’s a pretty big motivator, right?
Yet it’s apparently not stopping a good chunk of the workforce from trying to wait it out or even to sabotage the effort.
I get it. I’ve spent a good chunk of my career in the creative services, which have just taken another hit from Google’s nobody-wanted-this AI makeover of web search. I’m also scrupulous about not using AI to write stuff like this blog post, or anything else where it’s important for my authentic voice to come through. Not the trained-on-your-writing-style voice. The actual me.
And … my actual voice matters not a whit for the business documents I prepare in my day job. By design, these documents shouldn’t have a personality, but — and this is the important part — they also shouldn’t be a generic robot regurgitation. So when I use the tools at my disposal to save time as either a starting point or sanity checker, I’m still very much there in the idea/storyline, the prompts, the de-slopping, and the editing.
It took me awhile to get to that level of comfort, though. Wordsmithing is a core part of my identity (hello! blog!).
Two sides to everything
The Gallup report digs deeper into the real reasons behind the resistance, and the three biggest ones are (and I’m exaggerating to make a point): 1/ I prefer my job the way it is, thank you. 2/ AI is unethical and a scourge on society. 3/ Did y’all see the giant privacy and security holes this opens up??
I’m surprised we even needed to have a survey to confirm this. OF COURSE people are hesitant to rush out and adopt a technology that could destroy their livelihood and, hell, all of society whenever Skynet becomes self-aware.
So when it seems like the folks in charge don’t care or, worse, are actively doling out punishment, well, then, a bit of rebellion is to be expected. I’m not sanctioning that, but I also think leaders shouldn’t be surprised by it.
What I’m more worried about, however, is the acceleration of a reckoning that makes Gallup’s organizational “let’s simply redesign your role!” levers look like child’s play.
On one side of the AI spectrum, we’ve got the non-users, the wave of data-center bans, and the rampant disparagement at high-profile graduation ceremonies where a bunch of new grads know they’re about to enter a job market that places more value on robots than their skills and education.
On the other side, we’ve got the tech bros, CEOs, founders, and Claude Code fanatics who are eating, breathing, sleeping AI, and telling everyone who will listen that if you don’t join them, you’ll be cast into the ether of invisibility faster than a social media post that doesn’t match the algorithm.
I consider myself squarely in the middle (pragmatic optimist), and it’s getting rather lonely here.
The power struggle ramps up
Workers and bosses can’t be at open war with each other and expect business to carry on as usual. This is the Queen Mother of organizational dysfunction, and yes, there are a lot of societal factors feeding into it, too. I’m not a fan of the Us vs. Them that permeates everything these days, but this is a workplace blog, so … back on topic …
Another recent survey shows that almost every executive out there expects to win the war by laying folks off in the next two years. This sort of talk, and the number of layoffs that have already happened, only prompt fear and resentment in the people who remain. Some think the cowing of the workforce is deliberate, and there’s likely some truth in that. Capitalists gonna capitalist.
However, one of the assumptions the execs are making is that AI agents either already can or soon will be able to replace humans for a majority of tasks. This skips over the inconvenient truth that using AI at this scale is frickin’ expensive. Cheap access disguised it for a long time, but it’s starting to crack.
Lemme say that again for the folks in the back: AI is frickin’ expensive.
In the news recently was the cautionary tale of a company that spent a half-billion dollars in a single month on Claude. They forgot to put limits on tokens. Even Microsoft pulled back its Claude Code licenses and is steering employees to its homegrown Copilot to save money. And Meta and Amazon took down their internal AI leaderboards after folks started gaming the system and running up the $’s with frivolous tokenmaxxing prompts.
Play this out a bit: Companies are telling everyone to get on the AI train; they grumble a bit but they do; hundreds and thousands of agents start doing work, and … oops! The productivity savings don’t come anywhere close to offsetting the cost of all the compute power, which is already in short supply because everyone else is deploying massive agent armies, too.
Then, here come the bosses, hat in hand, back to the humans who are cheaper to hire than agents. Except now the humans have seen behind the curtain at just how expendable they’re considered to be.
It’s the economy, sunshine
AI backlash is on the upswing right now, but it’s not driving the recent moderation of breathless new-era prognostications. The balance sheets are, as is the tolerance for long-term risk. One of the reasons for the lackluster productivity savings is because there hasn’t been enough time yet to overhaul the systems for agents, which in theory will get resolved. In theory.
So despite the jobpocalypse threats, I predict we’ll get a breather in the open war between workers and bosses. The cold war will persist, though, which is a formidable culture challenge. Companies will always be looking for ways to cut costs, and certain groups will be disrupted no matter what. In particular, software engineering, creative services, and customer support are the low-hanging fruit of automation.
If you’re caught in the middle, whether as a manager or a pragmatic optimist, it might seem like you’re walking a thin line between preserving the continuity of your paycheck and betraying your colleagues and/or the world at large. It’s a tough spot to be in, but you don’t have to pick sides. It’s still anyone’s guess as to how it will turn out, and how your personal value prop will play a role. This might be the time to activate that new career direction you’ve been mulling over, inside or outside of corporate life. Or it can be an opportunity to boost both your hard skills — the technology itself — and soft skills like empathy and relationship-building.
You’ll need both to navigate whatever’s ahead. They’re the skills that make you uniquely human, and uniquely well-positioned to control your own destiny.
All opinions here are my own. All text is my own, too, including the em dashes. I welcome constructive comments and discussion on LinkedIn and Bluesky.