You know those moments when you’re scrolling through Instagram or YouTube and suddenly an ad pops up that feels a little too accurate? Not in a creepy sci-fi way, but in a “wait, how did it know I was thinking about this? ” kind of way. I still catch myself raising an eyebrow at those. And that’s coming from someone who literally works in social media and sees the backside of how all this stuff works. Even then, it can feel strangely personal.
What’s wild is how quietly this shift happened. There was no big day where the internet collectively decided, “Alright, everyone, from now on, your feed will practically read your mind.” Instead, it happened slowly — quietly slipping into the corners of how every app functions. AI isn’t just looking at clicks anymore. That was years ago. Now it watches all these tiny, half-unconscious things we do without noticing. How long do you hover on a cooking reel? The video you rewatch because you zoned out the first time. Even the kind of thumbnails your eyes naturally drift toward. None of these moments seems important on their own, but AI picks them up like puzzle pieces.
Put enough of those pieces together, and suddenly your feed feels almost custom-built for you. Not intentionally personal, just… familiar. Like it knows the version of you that exists when you’re not trying. And honestly, most of us don’t question it anymore. We’ve sort of made peace with the idea that our phones know more about our habits than we do.
But if you rewind to a few years back, marketing wasn’t anything like this. Campaigns used to be these big, all-or-nothing ideas. You’d come up with a hero concept, package it nicely, and send it out into the world, hoping it connected with as many people as possible. It was broader, more general. There wasn’t really a way to personalise things unless you wanted to spend weeks creating separate versions manually—and nobody had the time for that.
Today, that whole system feels ancient. Now you could take one idea and turn it into fifty different versions without blinking. AI tweaks the visuals, rewrites the captions, and changes the tone — and it can do this for micro-groups of people. Sometimes even individuals. It’s not that a human sits and crafts each version. It’s more like the machine says, “For this type of viewer, warm colours work. For this one, a short caption. For that one, maybe a slightly emotional tone.” And boom—you have ten variations running at the same time.
What’s funny is that even within the industry, this didn’t feel like a sudden jump. It was like a slow update that kept happening in the background. A new tool here, a smarter analytics dashboard there, and before we realised it, AI had quietly started holding up half the workload. Ads are running right now that no human ever manually designed — they exist because a machine stitched together what it believes is the “best possible combination”. That still blows my mind a little.
Even so, the creative part hasn’t disappeared. If anything, it’s become more important. AI is great at calculations, patterns, and spotting the tiny signals we overlook — but it doesn’t know why something feels funny or touching or comforting. It doesn’t understand culture. It can’t sense when humour is too dry or when a message might unintentionally offend someone. That’s the stuff humans still handle. And thankfully so, because that’s the part that makes marketing actually feel alive.
Most regular users never see any of this complexity. From the outside, the ads look random. You watch three workout videos, and the next day your feed suddenly decides you’re on some fitness transformation journey. Or you save one travel reel, and suddenly, every platform wants you to plan a holiday. It feels like coincidence, but it’s not. The system pays attention to the tiniest shifts in your behaviour. We leave little breadcrumbs everywhere — and AI is very good at sweeping them up.
But even though AI has gotten scarily efficient, it still lacks something crucial: judgement. It doesn’t know what’s ethically right or where the line should be drawn. It can’t tell when something feels invasive. It just follows patterns. Which is why humans still have to decide how far things should go. Privacy, transparency, consent — those aren’t things AI “understands”. They are things we have to enforce.
And trust matters. If people feel like their data is being used in a sketchy way, they pull away. The responsibility for keeping that trust isn’t on the machine. It’s on the people running it.
So where does that leave us now? Somewhere in the middle. AI handles speed, scale, and the heavy technical work. Humans handle everything that makes communication feel like communication — emotion, storytelling, timing, humour, empathy, and judgement. Honestly, the best results we’re seeing these days come from the collaboration between the two. Not from AI alone. Definitely not from humans alone. But from this weird partnership we’ve ended up with.
And the truth is, even with machines quietly running in the background of almost every digital platform, the final experience — what someone actually feels when they see a message — still depends on human choices. The warmth of a line, the little joke in the copy, the way an image feels familiar — none of that comes from an algorithm.
AI might help you reach the right person, but only people can make something worth being reached.
