It's Already Late. But Better Late than Never: Try AI Now.

By Hira Fernando | September 2025 | Career Strategy

Let's say you're a sceptic. Fair enough. Let's say you think the AI hype has been overdone, that the tools are overrated, that it's a bubble waiting to pop. Fine. I'm not here to argue with you — I'm here to tell you it doesn't matter.

Even if you're right — even if this turns out to be less transformative than the most breathless headlines suggest — the cost of getting a basic handle on the technology now is low. And the cost of being completely unprepared if the optimists are right? That's a much steeper bill.

The models are not standing still

Here's what I keep coming back to: whatever you think about AI today is already out of date. Each new model release isn't a minor patch — it's a meaningful leap. The GPT-4 you might have tried eighteen months ago is not the same as what's available now. The tools that felt clunky and unreliable in 2023 are genuinely different animals today.

If you tried AI and walked away unimpressed, I'd encourage you to try again. Not because I'm an evangelist, but because the product has changed substantially. Even a cursory thirty-minute exploration of what current models can do will recalibrate your sense of where things stand.

Here's what it actually looks like in practice

At Careerly, we've integrated AI into our work almost daily — and I want to be specific about this, because vague claims about "using AI" aren't particularly useful.

We use it to draft session summaries and structured next steps after coaching conversations — the kind of documentation that used to eat up an hour of post-session admin and now takes minutes to review and personalise. We use it for candidate sourcing: building search strings, identifying overlooked talent pools, and stress-testing our criteria before we go to market. We use it for content — drafting LinkedIn posts, articles, and resources for our clients, which we then heavily edit and make our own (like this one, if I'm being transparent with you). And we use it as a thinking partner when we're mapping out career pivots or building job-search strategies with clients — feeding it context and asking it to push back, identify gaps, and suggest angles we might have missed.

None of this has replaced the human judgment, the relationship, or the coaching. But it has made us faster, sharper, and able to deliver a more thorough product to our clients. The session summaries are clearer. The candidate briefs are more comprehensive. The content is more consistent.

You don't need to become an expert. You need to become familiar.

I'm not suggesting you overhaul your entire workflow tomorrow. I'm suggesting you spend an afternoon with one of the tools — ChatGPT, Claude, Gemini — and give it a real task. Something you'd actually do at work. See what happens. Adjust your expectations up or down accordingly. But at least have a first-hand data point.

The professionals I see thriving right now are not the ones who have fully automated their jobs. They're the ones who have figured out where AI makes them better and leaned into it deliberately. That's a much more achievable goal than it sounds.

And given how fast each new model is improving — given that we are, as I wrote about in my last post, on the steep part of an exponential curve — the sooner you start building that familiarity, the better positioned you'll be when the next leap happens.

Because it will. And it will be sooner than you think.