2015–2026 / TEN+ YEARS OF EXPLORING CREATIVE FRONTIERS
2015–2026 / TEN+ YEARS OF EXPLORING CREATIVE FRONTIERS
17
You might have seen those videos (real or not) where some intrepid AI ranger pairs up a couple of chatbots for a conversation. Within a short time the two machines decide human language is just too slow. So they switch to — or invent on the spot — a more machine-aligned method of communication. To my ear, it’s indecipherable, and I doubt even the most adroit polyglot among us would be able to catch up.
Perhaps the two bots were plotting our demise or simply sharing funny stories about the higher primates who built them. The net result is the same: they created an environment where simple monkeys like us could no longer participate. Maybe we could’ve asked a third AI to translate, but in the end we were out of the loop.
Such a scenario clearly engenders all sorts of existential dilemmas…but the music hasn’t stopped yet.
Because, for the moment, the internet still belongs to us.
For the last 30+ years we’ve designed it that way.
Any designer who has spent decades spinning on this rock has certainly witnessed the best of times, the worst of times, and everything in between. We designed enough tri-fold brochures and magazine ads to choke landfills before it all eventually coalesced into one main bucket: online content.
But the one thing you could always count on was that the audience was human.
They responded well to all the squishy bits we use to ply our trade: storytelling, aesthetics, and emotion.
A popular AI platform (sorry, Google Search) told me that in 2026 there are ~1.4 billion people in the world who, natively or as a second language, speak English. That puts it well ahead of Mandarin, Hindi, and Spanish. Great for me — a product of the Georgia public school system whose two years of German never prepared me to deploy such gems as rindfleischetikettierungsüberwachungsaufgabenübertragungsgesetz, much less remember how to ask where the Bibliothekis.
So the concept of becoming competent in BotSpeak is a non-starter.
As I peer over the fortress wall, I see many thousands of AI agents steadily gathering on the field. Follow the trend line and eventually the math gets us to billions. That’s a pretty big cohort of “users” who will no doubt prefer words, visuals, and symbols that are as inscrutable to me as whale song.
And unlike us, they won’t care about the squishy bits.
These days, if you need to book a flight or hotel, you probably don’t engage with another human. You jump on Google or one of the several travel sites. And you’re met with splashy landing pages, sexy photos of poolside bars, and meticulously presented text showing price and availability.
All of it carefully designed to influence your emotions and convert you into a paying customer.
But soon, in many cases, you won’t be the one doing the browsing.
When we say “you,” we’ll actually be talking about a small army of agents operating on your behalf — scraping the internet for the best deal at the best price. Each person, multiple agents. Exponentially increasing their footprint. Eventually, the majority of traffic may not be human at all.
For a casual observer, it might look like nothing has changed. Humans will still be at their desks, mouse in hand, or nose deep in their phones.
But search will change.
Social media will change.
The browser itself will change.
Because these agents couldn’t care less how big the logo is or whether the photography on the landing page is perfectly on brand.
Maybe that’s a little liberating for us designers.
But it also points to a bigger shift: much of a brand’s future engagement may not happen through a graphical interface at all—at least not one designed in the traditional sense.
Rather, the agents will take an end run around all the UX and UI we currently pour our hearts and souls into. They’ll simply pull the data directly through APIs. And once they have it, the way they present it back to their users will be ad hoc layouts and procedurally generated content on the fly.
Power to the user. Love it.
It’s possible we’re headed toward a split. Not a new internet — but two layers of the same one.
The Human Web
Persuasive design. Brand storytelling. Carefully crafted experiences built for emotion, attention, and conversion.
The Agent Web
Structured data. APIs. Machine-to-machine negotiation happening far below the surface.
Humans will browse the first. Machines will increasingly operate in the second. And those two layers won’t always care about the same things.
Who recalls sitting in meetings hour after hour reviewing site traffic and watching mobile usage steadily supplant desktop? Responsive design eventually became de rigueur, but not without a few awkward years along the way.
We adapted.
We always do.
But what happens when the “user” is no longer concerned with things like “nothing important should fall below the fold” or “three clicks to conversion”?
Do we try to game the system again?
Maybe. But agents don’t behave like humans. They don’t fatigue. They don’t skim. They don’t bounce. They don’t feel.
They will happily visit all 366,258 pages.
Because humans optimize for attention scarcity.
Machines optimize for compute abundance.
I’ve seen robots doing wushu routines, assembling cars, folding laundry. They have two legs and five fingers because the world is built for us — stairs, doorknobs, dishwashers.
Robots adapt to human environments. But the internet isn’t a kitchen.
It’s not fixed. It’s not physical. It doesn’t need to stay human-shaped. Which means agents may not adapt to our interfaces. They may simply route around them.
And this is where things get uncomfortable.
Because if agents are just consuming structured data and generating outputs on the fly…what happens to design?
What happens to aesthetics?
What happens to brand?
I’ve seen enough rounded-corner, purple-gradient dashboards generated by AI to wonder if we’re drifting toward what I think of as the Creative Singularity — a point where everything looks the same, sounds the same, and feels the same.
A world where differentiation collapses into pure utility.
Price. Availability. Speed.
Nothing else.
Most of my career has revolved around three pillars.
Aesthetics.
Make it beautiful. Make it desirable. Make it feel like something.
Efficiency.
Remove friction. Shorten the path. Optimize the funnel.
Loyalty.
Create emotional connection. Build preference. Reward return behavior.
These things matter deeply to humans.
But agents don’t care how big the logo is.
They don’t care if the headline pops.
They don’t care about your brand story.
They care about data.
Maybe the answer isn’t to fight it.
Maybe agents still access the underlying data directly — but we give them a way to understand brand as well. Not visually, but structurally. Guidelines. Voice. Constraints. Signals.
Something that allows them to preserve differentiation — even when the interface is gone.
Because if we don’t…
We may end up with a perfectly efficient system that produces perfectly identical outcomes.
Either way, it appears we’re headed for two internets:
The Human Web.
And the Agent Web.
So where does that leave us?
Is there a way for creatives to operate in both?
Or are we destined to keep refining the surface layer — perfecting the squishy bits — while the real decisions move somewhere deeper, quieter, and entirely out of reach?