Two years ago I wrote a LinkedIn post telling businesses to develop AI agility. Build your awareness. Get people who understand this stuff close to the decision-making. And get ready to pivot, because the pivots were going to keep coming.
It was the right call. But it didn't land. Most of the businesses I speak to today still aren't built to move at the pace that's now required, and that pace has only accelerated since.
The tools I rely on most heavily today were either non-existent or materially weaker six months ago. Workflows I'd carefully built and optimised have already been superseded. Not because they were wrong. Because something better arrived. This keeps happening, and it's accelerating.
I think about agility differently now. Not as a strategy concept. As the foundational operating skill for anyone working with AI. Your ability to adapt, genuinely and quickly, is probably the most valuable thing you can develop right now. More valuable than any specific tool. More valuable than any particular workflow. Because the tools and workflows keep changing. The people who compound are the ones who move with them.
There are two dimensions to what's changing, and it helps to keep them separate.
The first is acceleration. Things you already know how to do, done dramatically faster. I work with a range of tools across my businesses, and I regularly discover that a task I've been doing manually has quietly developed an AI-native path. Something that used to take real configuration effort now largely handles itself, with me in the director's chair rather than the engine room. The 20 to 30 percent gains come from this constantly. Occasionally something lands that doubles the speed entirely. These aren't one-off improvements. They compound, because each one lands on a base that's already grown.
The second dimension is expansion. AI is collapsing the gap between what you know and what you can do. I build full applications now. Front-end, back-end, the AI engineering underneath, deployment, hosting, security, management. I use specialised AI agents across my business for financial management, privacy compliance, marketing strategy. I still have an accountant. I still consult specialists when I need to. But I can now get 80% of the way there independently, which means when I do engage a specialist, I arrive with better questions and clearer context. The expensive part of working across disciplines was never the expertise itself. It was starting from zero every time.
That second dimension is quietly enormous. In the old world, extending into a new discipline meant hiring someone, briefing them, waiting, reviewing, iterating across meetings and emails. The slow, expensive friction of coordinating across multiple specialists with different availability and different priorities. A lot of that friction is gone. You collapse the communication overhead entirely when the expertise sits alongside you in real time.
Both dimensions matter. But neither delivers value to people who aren't willing to keep moving.
Here's a pattern I see constantly. People treat AI fluency like a qualification. Do a course. Complete the training. Tick the box. Organisations do it too. Put the team through an AI programme and declare the job done. But AI fluency isn't a certificate you earn once. It's a practice you maintain. The landscape six months from now will look materially different from today, which means the skills and workflows you learned in that course are already on a countdown. The course gets you started. Agility is what keeps you current. And the difference between the two is the difference between knowing how to use the tools you have and recognising when those tools have been surpassed.
What actually works is staying close to the frontier. Not passively. Actively. I'm constantly absorbing what's coming out. New capabilities, new releases, new approaches. And I'm pattern-matching them against friction I already know exists in my own work. I know where my workflows are slow. I know where the manual steps are. I know where things could be smoother. So when a new capability lands, I'm not evaluating it in the abstract. I'm asking a very specific question: does this solve a problem I already have?
Sometimes the answer is yes, and the impact is immediate. A new capability arrives, I test it against a real workflow, and it dramatically accelerates something I was already doing. Other times the answer is not yet, but I've filed it away and I'm watching. The key is that I'm not waiting for someone to tell me what's relevant. I'm maintaining my own map of where the friction lives and actively scanning for the tools that eliminate it.
This is the real agility. Not reacting to change after it's obvious. Positioning yourself to absorb it the moment it arrives.
If you're building AI fluency, build this underneath it all. Not just skill with specific tools. The willingness to let go of tools when something better arrives. The habit of staying close enough to the frontier that you recognise the better thing when it shows up. The instinct to stay at the front of the change rather than consolidate behind it.
The acceleration isn't slowing down. The people moving fastest aren't necessarily the most technically skilled. They're the most adaptable. They treat their current approach as provisional, always. And that keeps the compounding running in their favour.
Your move, human.
Damien Healy is the founder of Qanara, an Australian AI consultancy helping businesses accelerate from strategy to impact. He writes about AI-native workflows, frontier AI capabilities, and practical transformation.
My LinkedIn articles are available via my post history and here: LinkedIn Articles | Damien Healy