The Future of Hidden AI in Experiences

The Future of Hidden AI in Experiences

Most AI in products today is treated like a main event. Visible, hyped, and frankly a bit exhausting. But the real power of AI doesn't live in flashy features or generating fake animal videos. It lives in what users don't notice.

Hidden AI is about designing systems that remove friction, anticipate needs, and adapt in real time without calling attention to themselves. When it works, experiences feel right. Nobody can quite explain why.

From Screens to Intent

We're moving away from designing interfaces and toward designing for intent. Instead of mapping clicks and scrolls, the focus shifts to understanding what a user actually wants and adjusting the experience around that. Static journeys no longer hold up. People's needs change mid-session, mid-decision, sometimes mid-sentence.

"The best AI makes experiences feel seamless and intuitive. It doesn't ask for attention. It just removes the moments where things could go wrong."

This shows up differently depending on the industry:

  • Luxury. AI quietly anticipates preferences to create more personal experiences and recommendations without adding a single visible layer of complexity.

  • Creative tools. It surfaces the right option at the right moment without breaking concentration.

  • Complex B2B. It cuts through navigation entirely. Instead of forcing users through rigid structures, the system reads where they are, connects the relevant pieces, and gets out of the way.

From Journey to System

We saw this most clearly working on a retail vision for a German car manufacturer. The brief started as a website project. It became something else entirely.

The experience we built doesn't follow a fixed funnel. It watches behaviour and responds to it, quietly removing irrelevant content and sharpening focus as someone moves from curiosity to a decision. Exploration, configuration, and purchase stop being separate stages. They blur into a single guided experience that gets progressively more specific the further in you go.

At the centre of it is an intelligent layer that most users will never consciously register. It connects digital and physical touchpoints, reads where someone is in their thinking, and prepares the next step before they ask for it. It doesn't announce itself. It just makes the whole thing feel surprisingly easy.

"It's not a feature. It's the logic that holds the entire experience together."

Trust is the New UX

The harder problem is trust. As these systems take on more responsibility, making choices on behalf of users, filtering what's shown, deciding what comes next, the question shifts from usability to something closer to legitimacy.

Hidden AI only works when users feel, even unconsciously, that the system won't misuse what it knows. That means:

  • Behaving predictably, even when the system could do more

  • Knowing when not to intervene

  • Always leaving the user with a clear sense that they're still in control

Get that wrong once, and the whole illusion collapses.

Brand by Behavior

Something else changes when AI is woven this deeply into an experience. Brand stops being something you communicate and starts being something you demonstrate.

Visual identity and tone of voice still matter. But increasingly, the brand lives in the decisions the system makes. What it prioritises, when it steps in, and how it handles ambiguity.

"A brand that promises precision needs a system that behaves precisely. One built around warmth needs an experience that actually feels warm, not just one that says so."

That gap between what a brand claims and how its systems actually behave is where a lot of trust gets lost right now. Closing it is one of the more underrated design problems of the moment.

Where This Is Going

None of this is really about adding AI to products. It's about using it to make experiences less complicated. The shift looks something like this:

  • From interfaces to intent

  • From fixed journeys to systems that adapt

  • From brand as something you say to brand as something you do

The products that will stand out won't be the ones with the most visible AI, as we saw with the SORA fall. They'll be the ones where you barely notice it's there, and everything works.

Everything Feels Like a Scam And What Designers Can Do About It

Everything Feels Like a Scam And What Designers Can Do About It

Open SoundCloud and you’ll get a DM from someone promising playlist placement — for a fee. Post on Instagram and a “brand rep” you’ve never heard of offers you free clothes if you just cover the shipping. On TikTok, a stranger invites you into a collab that turns out to be a funnel to their paid Discord. Even LinkedIn has its share of fake recruiters, promising remote dream jobs that somehow require you to buy training first.

These aren’t rare one-offs anymore. They’re daily interactions.

And the emotional residue is always the same: this feels like a scam. Once that suspicion sets in, it spreads. If a phishing page can perfectly clone your bank’s login, then even the real login feels risky. If bots flood Instagram with crypto giveaways, even authentic brands become harder to trust.

The Scam Layer of Digital Life

This isn’t just a vibe — the numbers tell the story:

  • Eight out of ten UK adults report being targeted by a scam in the last three months.

  • Online payment fraud is projected to cost businesses hundreds of billions over the next five years.

  • Only about one in five people feel confident spotting scams when they happen.

Scams have become the background hum of digital life.

How UX Helped Get Us Here

The uncomfortable truth is that legitimate design often mirrors scam tactics: pop-ups demanding immediate action, “limited time only” countdowns that never really end, subscription flows that are easy to enter and impossible to escape.

So when scammers copy those cues, users don’t see the difference. Growth hacks and grifts blur together.

Fighting Back With Design

The irony is that in a world where everything feels manipulative, clarity has become the rarest feature. An experience that feels trustworthy immediately stands out.

Some principles to build towards that:

  • Radical transparency: show costs upfront; make unsubscribing as easy as subscribing.

  • Contextual approvals: don’t just ask “Are you sure?” — show details like location, device, and time.

  • Smart friction: smooth for trusted actions, extra checks for risky ones.

  • Guardrails for email and social: mirrored in-app messages, clear domains, scam-report buttons.

  • Scam-proof support: scheduled in-app calls with session codes, clear “we’ll never ask for…” rules.

  • Human signals: real stories, behind-the-scenes content, and disclosure when AI is used.

Fighting Back Means More Than Defense

Design can’t just be about removing dark patterns and avoiding scammy signals. We also need to create tools that actively expose scams:

  • Interfaces that highlight suspicious behaviors — like unusual domains, recurring scam phrases, or high-risk payment flows.

  • Public scam dashboards that visualize trending fraud attempts in real time, so users see they’re not alone.

  • Built-in scam detectors in messaging apps that flag “this looks like a known scam script” before the user replies.

  • One-click report-and-freeze buttons that both protect the user and feed data back into collective defenses.

Think of it as giving people night-vision goggles for the dark web of scams: making the invisible visible, so scams lose their power in the shadows.

When AI Fights Back

Not all defenses are invisible. Virgin Media O2 recently introduced Daisy, an AI “grandma” who answers scam calls. Instead of hanging up, she rambles about knitting, family stories, and tech confusion — sometimes keeping fraudsters tied up for forty minutes.

It works because Daisy is designed with persona, pacing, and empathy cues that scammers are reluctant to abandon. She doesn’t just block the attack — she flips the experience, wasting the scammer’s time instead of the victim’s.

For designers, Daisy is a reminder that security doesn’t always mean sterile walls of friction. Sometimes it means building resistance into the interaction itself — designing an experience that scammers don’t want to deal with.

SXSW London -  Where creativity, culture, business and technology meet

SXSW London - Where creativity, culture, business and technology meet

Now that Cannes has passed and the focus is now shifting on towards the summer holidays, I wanna go back to the beginning of June and to London, to the hip and cool Shoreditch district where the first SXSW London was held.

SXSW is originating from the Texan capital of Austin and now was held in Europe for the first time. It’s a festival of creativity, startups, technology, film and music all wrapped up together in a beautiful package. As a Creative Director the mix is super close to my heart how I approach creativity and how usually we unlock the best work by mixing all of these together.

As in all conferences now in this day and age AI was in the headlines of many talks. One of the most interesting was by Azeem Azhar talking how AI becomes the default for cognition—“intelligence on tap.” This shift transforms AI from something we summon via prompts to something constantly working in the background. Also how AI will become more personal and proactive acting on our behalf: booking travel, managing finances, handling emails, and even negotiating. I found it one of the few real in depth talks and going beyond being a talk of the AI-bros next video generator tool.

See more from the topic at:
https://www.exponentialview.co/p/ai-in-2030

Digital Art

From a tech meets creative arts perspective one piece really stuck with me: Beeple’s (Mike Winkelmann) “Tree of Knowledge.” It’s a tall, four-sided video sculpture that reacts to real-time data like news, social media, and weather. As the info flows in, the digital tree either grows or decays. It’s constantly shifting, like a living mirror of the world’s noise.

A dial lets you choose between “Signal” and “Noise.” Turn it to Signal, and the piece calms down—the tree grows peacefully. Turn to Noise, and it erupts with chaotic headlines and glitchy media storms.Also at the control console next to the dial was a switch that if pushed 666 times it would permanently delete the piece. Of course this was covered with a locked lid but makes an interesting social experiment, how many would press it and how many would not to keep it not being destroyed.

It’s cool that Beeple isn’t moralizing. He’s just showing us the world we live in, and letting us decide how much of it we really want to take in. In the middle of all the hype and energy of the festival, this piece was a nice breather.


Inclusive design as a creative currency

Idris Elba joined a shared-stage conversation on how creativity can drive meaningful change—especially across the African continent.He laid out plans for an “African Odeon”—a network of local, data-powered cinemas designed to bring first-run films directly to African audiences, and spotlight homegrown talent.

He also introduced Akuna Wallet, a cross-border payment platform built to solve a real problem: helping African creatives actually get paid. He demoed Talking Scripts, an AI tool that converts screenplays to audio, making storytelling more accessible for neurodivergent creatives. His final line summed it all up: “Fame isn’t the prize. The prize is what you do with it.” And this—building infrastructure, not just hype—felt like exactly that.

Also the Shoreditch locations hosted talks from movie stars like Joseph Fiennes and music from many fantastic upcoming artists plus loads of craft coffee and beer.

If there’s a 2026 SXSW London count me in.

5 Take-aways from the Conference

1. AI for Humans, Not Replacements
AI was front and centre—but the focus was on creativity amplification, not automation. Leaders emphasized using AI as a partner in creative work, not a replacement. “The future isn’t AI OR humans. It’s AI AND humans, working together.”

2. Authenticity & Trust Matter More Than Ever
Brands are moving beyond flashy campaigns—what matters is transparency, authentic connections, and ethical operations. Trust is the real differentiator, and true influence now comes from bold, community-rooted storytelling

3. Access Is Innovation
Accessible tech wasn’t siloed—it was embedded across disciplines. From adaptive gaming rigs to voice-driven design systems and neurodivergent-friendly UX, SXSW London made the case that designing for inclusion sparks better ideas for everyone. Accessibility isn’t a checkbox—it’s where true creativity starts.

4. Sustainability Is Getting Smarter
Greenwashing is out—measurable, tech-driven sustainability is in. Startups and big players alike showcased innovations from carbon-literate design tools to blockchain-backed supply transparency. The big idea: sustainability isn’t a separate strategy anymore, it’s a built-in system constraint—and tech is finally catching up to that responsibility.

5. The Internet Is Entering a Feelings Phase
A surprising through line: emotional intelligence in design. From AI mood-matching music to interfaces that respond to mental health signals, the future of tech is becoming more empathic, intuitive, and affective. We’re moving from logic-based systems to ones that read, reflect, and respond to human nuance.

Vibe Coding

Vibe Coding

I started my career 25 years ago as an HTML Developer in one of the oldest ad agencies in Finland. I was hired because one of my hobbies back then was making websites for bands and friends. Building stuff and seeing it work in a browser was fun.

Then came the world of Macromedia Flash and ActionScript, which opened even more creative doors for the web.

After that? I didn’t really code anymore.
Modern front-end development turned into a deep, complex craft—brilliant, but intimidating. Frameworks, build tools, TypeScript, tokens... it all felt like a different world.

Until now.

My colleague Philipp Wornath was using a tool called v0 by Vercel for our Manga Superhero Machine, and he vibe coded most of it into existence. Vibe coding means coding without touching code. You prompt an AI like v0 with an idea or mood, and it builds the thing. No syntax. Just works.

It sounded like hype—until I tried it myself.

I took one client’s website product page, with a complex product portfolio, and simply asked v0 to create a product configurator. Within minutes, I had a working prototype, in brand colors, with mock 3D images and dropdowns—fully interactive. Something that would’ve taken a frontend dev hours. And definitely not something I’d ever try to build in Figma. (Sorry Figma dropdowns—you’re not it.)

Also made this one test of a landing page in 5 seconds

Vercel v0 has lots of community projects visible to give a glimpse of the capabilities.

So basically: it’s GenAI for code.

And it’s not just basic components. It can sketch out full interfaces, design ideas, and functional layouts that are deeply influenced by tone and intent.

Vercel v0 is ideal for:

  • Generating frontend code dynamically with AI

  • Instantly previewing live UIs

  • Prototyping without infrastructure headaches

  • Building experiences that feel right

Does it replace production devs? Not quite.
But for testing, prototyping, and exploring ideas? It’s a game changer.

On Zuckerberg's AI Coding Prediction

Recently, Mark Zuckerberg said he believes that soon “most or all code will be written by AI.”

Honestly? I can see it.

Using v0 was the first time in years that I felt like I could build things again—not as a developer, but as a creative. Not wrestling with code, but expressing an idea. And that’s the shift: AI isn’t here to replace creativity—it’s here to remove friction from it.
If we can go from idea → browser in minutes, why wouldn't we?

Of course, we’ll still need devs. There’s nuance, architecture, performance, and security—things no prompt will fully solve.But for creatives, strategists, and designers who’ve always wanted to make, tools like v0 open the door again.

Sometimes, you just want to test an idea in the browser.
You don’t want to brief a sprint team or wait three weeks.
You want to see it. Feel it. Click it.

And now, you can. Just Vibe it.

Test v0 yourself

Image credit : Unsplash apoorv mittal

Welcome to the Dark Side of UX

Welcome to the Dark Side of UX

As yesterday was Star Wars day May the 4th and today is May the 5th also know as Revenge of the Fifth, The Dark Side Day. As the awesome third Prequel movie is enjoying it’s 20th Anniversary I decided to dive and shine some dark light on the dark side of UX.

Ever tried cancelling a subscription and ended up feeling like you were trapped in an escape room designed by someone who hates you? That’s not a bug. That’s design.

UX has a Dark Side

We all talk about good design. Delight. Joy. Smooth flows and happy faces. But in between all those friendly checkboxes and “accept” buttons, there’s a quieter, more sinister craft going on. The kind that gets users to click “yes” when they meant “no.” That adds things to your cart when you weren’t looking. That hides the cancel button like it owes you money. Design can manipulate just as easily as it can guide. It’s a question of intent.

What does the Dark Side look like?

Some of the classics:

The Roach Motel
Super easy to get in. Seems like impossible to get out. Think: free trials with cancel flows that somehow take 12 clicks and a call to customer service. Examples: LinkedIn’s Premium cancellation flow, HelloFresh subscription cancellation and Amazon: How to cancel Audible subscription?

Confirmshaming
“No thanks, I don’t like saving money.” Or worse: “No thanks, I prefer being cold and alone.” Why is the opt-out always insulting? Example: Wish.com

Sneak into Basket
Suddenly there’s a €9.99 “support plan” in your cart. You didn’t put it there. You just got nudged. Example: Name.com: Look for a "surprise" in your cart

Bait and Switch
You click “Next” thinking you’re skipping a step, but surprise — you just signed up for emails from six partners. Example: Microsoft Windows 10 Upgrade

Misdirection
Cancel? Log out? Delete account? Good luck. You’ll need a map, and possibly emotional support. Example:
Amazon Prime cancelation

And the design rationale?
“Well, it works.”
Of course it works.
Tricks work.
But so does spam, for a while.

Dark patterns spike short-term conversions, sure. But long-term? You’re burning trust for clicks. And once people realise they’ve been tricked, they remember and walk away.It’s like using the Force choke when a firm conversation would’ve worked. Impressive, but maybe not the vibe you want.

Why it matters more than ever

Interfaces are getting smarter, more invisible. AI assistants, voice interactions, subtle nudges. It’s easier than ever to hide the Dark Side in the margins. Which means it’s also easier than ever to excuse it.

“This is how it’s done.”
“This is what performs.”
“This is what the stakeholders want.”

But someone still designs the thing.
Someone still writes the copy.
And someone still decides how hard it is to say “no.”
That’s the moment where we choose our side.

So what now?

It’s May 5th. The revenge of the fifth. But if you’re designing something today — a form, a funnel, a chatbot, anything — maybe stop for a second and ask: Would I want someone to do this to me?

If the answer is no... you know what to do.

For more Dark Patterns head out to: Dark Patterns
Hall of Shame
and Responsible Design series by Symran Bhue & Stuti Mazumdar