The Real Barrier to AI Adoption Isn't Skill. It's Self-Awareness.
What really stands in the way of AI adoption isn't the tools. It's whether people are ready to look at their own thinking.
27 March 2026
I got into a thread recently with Lesia Kalley that went somewhere I wasn't expecting.
Her argument resonated immediately: AI isn't weakening human judgement, it's exposing a muscle that was already atrophied. Humans have been outsourcing thinking for centuries, to the crowd, the boss, the expert on TV, the algorithm. GenAI didn't invent that. It just gave us something new to blame. Her closing line was simple: "The discomfort isn't the tool. It's the mirror."
That framing matched something I'd already been noticing. The value people get from AI tracks the quality of their thinking going in. Prompts, plans, instructions, all of it reflects the clarity of the intent behind it. AI doesn't just generate outputs, it shows you how well you'd actually thought something through before you asked. Strong thinking gets amplified. Weak thinking gets exposed. Both happen faster than before.
Lesia pushed the idea further: is AI a great equaliser, or is it widening the gap between strong and weak thinkers? My instinct is that it's widening it, considerably. The feedback loop is compressed now, so people who invest in how they think compound their advantage, while everyone else feels the distance more visibly and more quickly.
The real insight came next. The gap won't just be about skill, Lesia argued. It'll be about self-awareness, and that's much harder to close, because self-awareness demands something most organisations don't train for: reflection, discomfort, a genuine willingness to confront your own thinking. Not everyone chooses that path.
What this creates in practice is a split that was probably always there. Some people lean into change: they ask questions, experiment, adapt even when they're behind. Others pull back. I'm already seeing it in conversations, curiosity on one side and something close to paralysis on the other whenever AI comes up.
This is where the conversation turned practical. In her own work, Lesia has found that naming the fear openly, not reframing it, not managing it away, is often what helps people move forward. That matches how I've always approached change management. The moment you bring people in early, acknowledge concerns without flinching, and involve them in shaping what's coming, the dynamic shifts. It stops being something happening to them and becomes something they're part of.
Most organisations still frame AI adoption as a skills initiative: tools, training, capability. None of that is wrong, but it's incomplete, because underneath all of it sits something more fundamental: whether people are willing and able to engage at all.
The question isn't only how to train people to use AI. It's how to help people get comfortable enough to engage with it in the first place. That's a different problem, and it needs a different approach.
There will always be people who lean in and people who don't. AI isn't creating that divide, it's making it visible and speeding it up.
For the people who do engage, the upside is real: learning compounds, capability scales, confidence builds. None of it starts with tools. It starts with awareness.
Structure before AI. Human readiness before capability scaling. In that order.
