I love it when people come up to me after my talks, I get a real insight to what’s happening out in the “real-world”. And last week was no exception, A CEO approached me, with a slightly conspiratorial tone, like he’s about to share something clever. “I think you’ll like this,” he said. “We’ve announced we’re AI-first.”
It was said in a way that someone might say they’ve just bought an electric car (looking at my co-founder Helena here) or taken up cold plunges. Like they’d just made some bold, progressive choice.
He was waiting for applause. So I asked the stupid question. “What does that mean?”
Long pause. Then: “We’ve rolled out Copilot licences across the business.”
Right. OK. That’s… one thing you can do. “And strategically? What’s the actual plan?”
Another pause. Longer this time.
It turns out the plan was that “50% of work will be done with or by AI by the end of the year.” It was unlikely that anyone was actually sure what that meant or what the final result would actually look like. There was no training budget, no redesign of workflows, no thinking about what capabilities the organisation actually needed. Just, “People will figure it out,” he said. “They’re smart.”
When I gently pushed on how any of this actually built long-term advantage, he just repeated the slogan. “We’re AI-first and our competitors aren’t.”
It was like meeting someone wearing a Motörhead T-shirt who’s never actually listened to the music (you really should!) And as I wished him luck and walked away I felt two things: deeply sorry for his employees, and mildly stunned that someone could be so confidently shallow about something so consequential.
“AI-first” isn’t a strategy. It’s a confession.
Declaring you’re AI-first sounds bold. It isn’t.
It’s answering the question before you’ve even asked it. It’s walking into a pharmacy and asking for their very latest drugs before you’ve been diagnosed.
Real strategy is about trade-offs. It clarifies what you’re going to prioritise and what you’re not. It tells you when not to do something. “AI-first” clarifies nothing. It doesn’t tell you when not to automate, what capabilities to build or what makes you distinctive.
It’s pure theatre. It’s AI virtue-signalling.
But, let’s be frank, it’s also a tell. Because the actual choice most companies are making isn’t AI-first versus human-first. It’s short-term margin extraction versus long-term capability building. And if you’re being honest with yourself, you probably know which one you picked.
AI doesn’t remove the need for capability. It simply reveals whether you have any to begin with.
Harvard Business Review Review recently looked at the difference between automation and augmentation where automation replaces labour while augmentation amplifies it.
The research is pretty clear: automation gives you early wins. Augmentation compounds. But there’s something more fundamental happening. AI doesn’t remove the need for capability. It multiplies it.
When you can generate a thousand options an hour, judgment becomes more important, not less. If you can’t spot good thinking, you’re just faster at being wrong. It’s the same with briefing, prompting isn’t magic, it’s structured thinking. And most organisations have been historically terrible at structured thinking.
Same with domain expertise. If the knowledge isn’t in your head, you can’t evaluate what the tool gives you. Outsourcing cognition doesn’t build advantages, it undermines them.
Same with how fast your people learn. If employees feel replaceable, they stop investing extra effort, they self-protect, they stop taking risks. And culture – I mean real culture, not the town hall slide deck – becomes immediately visible.
Because when leaders signal “we’re AI-first,” a lot of people in the corridors hear something different:
“You are not a priority for us.”
The best people don’t usually stick around to hear the rest. Because good people tend to seek opportunities that are worth the effort, not opportunities to reduce their effort.
The seduction of the shortcut
Automation looks brilliant in the early months. Headcount drops. Costs shrink. The spreadsheets glow approvingly. The CFO is happy. The board nods. Everyone feels clever.
Economist Erik Brynjolfsson calls it the “productivity J-curve” – the lag between adopting new technology and actually seeing real gains. Meaningful transformation requires rewiring the organisation and developing skills, and that costs many times more than the technology itself.
Licences are cheap.
Building capability is expensive.
So shortcut leadership goes for the visible win.
But there’s a predictable spiral when AI becomes a replacement strategy rather than an augmentation strategy. Morale dips. Engagement becomes shallow. “Use the tool” becomes compliance instead of curiosity. You get what researchers call “workslop” – low-quality AI output everywhere. Your high performers quietly update their LinkedIn profiles. Institutional knowledge thins out.
On paper, efficiency is improved.
Underneath the hood however, capability is being eroded.
The real fork in the road
HBR’s research shows that employee perception determines whether AI adoption becomes shallow automation or something that actually compounds. But perception is shaped by leadership intent.
When boards obsess over quarterly cost compression, AI becomes a weapon for margin engineering.
When leaders invest in capability – training judgment, developing better briefing skills, building domain depth, creating AI literacy – the same technology becomes leverage. This isn’t sentimental thinking, it’s strategic. Strong organisations get stronger because AI amplifies what’s already there. Weak ones just get exposed faster. You can’t automate your way to intelligence.
So here’s the thing
That CEO genuinely believed they were being progressive. They weren’t, they were taking a shortcut and calling it strategy.
Slogans don’t build an advantage. Capability does.
If you want 50% of work done “with or by AI,” fine. But what kind of work? With what judgment? Evaluated by whom? Built on what expertise? Sustained by what culture?
“AI-first” is not a bold stance. It’s often just a shortcut dressed up as strategy. And shortcuts don’t compound.
The leaders winning this decade won’t be the ones replacing people fastest. They’ll be the ones building the most capable humans and using AI tools to help them reach higher and do better.
AI isn’t a test of technological sophistication. It’s a test of whether you believe people are costs to be minimised or potential to be amplified.
And my final words to that CEO – come and talk to us at thegenaiacademy.com We can help you reach your AI goals without sacrificing the motivations and talents of your team. Because we believe the best way to handle this wave of technology is to be people-first. Focus your efforts in the right place and profits will follow.
