Another week gone, and honestly? I'm sitting here on a Friday evening feeling like I've been handed a jetpack and nobody bothered to warn me. The pace of what AI can do right now is genuinely dizzying — not in a scary way, but in that giddy, Christmas-morning kind of way where you keep finding new things under the tree. Let me walk you through the moments this week that made me actually stop and go "wait, we can just… do that?"

I'm not going to give you tutorials or step-by-step walkthroughs — that's not what this is. What I want to share are the moments where AI made me think differently, and why I think you should be paying attention too.

Selling the Dream —
AI as Art Director

Full disclosure: this one is personal. My wife Natz has been obsessed with interior design for fifteen years. Fifteen years. I have sat through more conversations about whether a room "needs warmth or needs light" than I care to count. I have been asked my opinion on paint swatches when I didn't know there was a question coming. I once came home to find the living room furniture rearranged while I was at the shops. I have, in short, been her unofficial — and largely unqualified — interior design consultant for a decade and a half.

So when Natz decided to turn that obsession into an AI-assisted interior design mini course — Little Beatrice, it's called — I was thrilled. Genuinely. Not just because it's a brilliant idea, but because she now has paying students to inflict her opinions on instead of me. AI has done what fifteen years of marriage couldn't: given her a proper audience. I helped her build the sales page for it, and here's where it gets interesting.

When we first had all the room images laid out in a gallery grid, it looked fine — but it felt like an estate agent's brochure. Then we had the conversation that changed everything: what if we used the images more sparsely? Not as a gallery, but as atmosphere? So we dropped one as a near-invisible hero background, gave one image a full-bleed editorial feature moment partway down the page, and let another whisper softly behind the call-to-action button.

"Three images, three completely different treatments — and the page went from 'here are some rooms' to 'here is a feeling.'"

— Jamie Gibbons, Issue #2

That's not a coding trick. That's editorial thinking, delivered in a conversation. And frankly, it's the same instinct Natz has been applying to actual rooms for years — I just never had a way to bottle it into a web page before. It also reminded me of everything I've been doing with ChatGPT Images lately for house redecorating visualisations — a tool that, I'll be honest, Natz has already hijacked and is using far more fluently than me. You describe a room, a palette, a vibe, and it shows you what it could look like before a single tin of paint has been opened. In our house, that means I now get shown AI-rendered versions of proposed room changes rather than being asked to "just imagine it." Progress.

Why it matters

We're moving beyond AI as a tool that executes instructions into AI as a collaborator that helps you make better creative decisions. When the AI suggests "fewer images, more atmosphere," it's not being a coder — it's being an art director. That shift in relationship changes everything about what's possible. And if it can help my wife turn fifteen years of passionate expertise into a sellable course in an afternoon, it can probably help you too.

From Brief to
Breathtaking

Right. This one really got me. I needed a full redesign for a George Michael tribute show website — tour dates, performer bios, a booking form, all of it. In the past this would have been a freelance commission that took weeks and cost a few grand. This week it took an afternoon.

But it's not just the speed that mind-bends me. It's the ambition the AI brought to it. I didn't ask for parallax scrolling. I didn't specify animated spotlight glows or song titles drifting across the screen like floating ghosts. I didn't request a looping gold marquee ticker of George Michael songs, or pulsing concentric rings in the legacy section.

All forty-plus tour dates across 2026 and 2027, switchable by year. Sold-out state indicators. A sticky nav that transitions from transparent to dark glass as you scroll. The whole thing smelled like deep black and warm gold — exactly the palette you'd want for a tribute to one of the greatest artists of his generation. All of that came from a conversation. Not from specifying every pixel.

"I asked for a George Michael tribute website and got parallax heroes, pulsing light rings, and Careless Whisper drifting across the screen in floating animated text."

— Jamie Gibbons, Issue #2

There's something philosophically interesting happening here. When you brief a human designer, they bring their taste and experience to the project. AI is now doing the same thing — interpreting a brief and elevating it. The question isn't "can it build things" anymore. The question is becoming "what does it choose to add?" And increasingly, the answer is: exactly the right things.

40+
Tour dates built
1
Afternoon to build
0
Pixel specs written

Apple's Wild Move —
Choose Your AI

Now for the week's big headline that made my jaw drop a little. Apple is planning to let users in iOS 27 choose their AI provider — not just ChatGPT as the one approved partner, but a full menu: Claude, Gemini, ChatGPT, and potentially others. You'll be able to set your preferred AI in Settings and have it power Writing Tools, Image Playground, Siri responses — the whole Apple Intelligence stack.

This is a really big deal, and I don't think enough people are talking about it properly. Apple built a walled garden and now they're installing a door in it — because they've realised AI is becoming infrastructure, like choosing a mobile network or a browser. Nobody wants to be locked into a single provider forever, and Apple clearly knows that locking people to one AI model is a worse user experience than letting them choose.

Why it matters

When AI becomes a setting you choose — like your default browser — it signals that we've crossed a threshold. AI is no longer a novelty feature. It's utility. It's plumbing. And the companies that make the best AI will compete on their merits rather than on distribution lock-in. That's genuinely good for all of us.

There's a fun implication here too. Apple has internally tested integrations with both Google and Anthropic, and we're likely to hear much more about this at WWDC on June 8th. If you've been curious about trying Claude but haven't set it up separately, iOS 27 might be the moment it just appears on your phone as an option. That's when mass adoption really starts.

Meanwhile, Google has been busy too — they unveiled Gemini 3 Pro this week, their most capable reasoning model yet, with agentic coding, improved tool use, and some genuinely impressive multi-step thinking. Google I/O is on the horizon and a new Gemini Spark personal AI agent is teased — supposedly a 24/7 AI that learns from your behaviour and handles tasks on an ongoing basis. I'll be watching that very closely.

If there's a thread running through all of this week's moments, it's creative leverage and the quiet normalisation of AI. An AI that thinks like an art director. A website that brought cinematic flair to a brief that never asked for it. A phone that's about to let you choose your AI like you choose your browser. These aren't isolated novelties — they're a consistent direction of travel.

See you next Friday — there's already plenty queued up. And if something AI-related has blown your mind this week, I genuinely want to hear it.

AI #ChatGPTImages #iOS27 #Claude #Gemini3 #WebDesign #InteriorDesign #AITools #NoCode