In this week’s Our Take: McNugget caviar goes luxury, AI bots build their own world, perfume clubs fight loneliness and fashion gets algorithmic.
McCaviar Nugguts, we’re loving it!
Seeing McDonald’s and caviar side by side still reads like satire. And yet, here we are.
For Valentine’s, the undisputed king of drive-thru dinners has decided romance looks like Chicken McNuggets topped with Baerii sturgeon caviar, a dollop of crème fraîche and, naturally, a mother-of-pearl spoon. It’s being served up as a limited-edition McNugget Caviar kit, online only, because of course it is.
The best part? This didn’t come from a whiteboard in a boardroom. It started as a joke. A meme. The kind of chaotic internet suggestion that usually lives and dies on TikTok. Instead of pretending they didn’t see it, McDonald’s leaned in and made it real. Not as a quiet nod, but as a fully formed drop.
Brands used to “tap into trends.” Now they’re expected to participate in the bit. The audience sets the tone, and the smartest players co-sign it in real time. By elevating a tongue-in-cheek idea into an actual product, McDonald’s signals that it’s not just monitoring culture, it’s comfortable playing with it.
There’s something deliciously unhinged about pairing a $25 Arch Card with luxury fish eggs. It shouldn’t work. That’s why it does. High-low has been fashion’s favourite trick for years; now fast food is borrowing the formula.
The real win isn’t the caviar. It’s the self-awareness. In 2026, brands that can laugh with the internet, and occasionally indulge it, are the ones that stay culturally fluent. Romance, apparently, just needed a side of fries.
The Bot Thickens
Launched at the end of January, Moltbook is a social network built specifically for AI agents. It looks suspiciously familiar, very Reddit-coded, complete with forums and upvotes, except the “users” are bots. Over 1.5 million of them, apparently. They post, reply, debate, form groups. Humans? We’re just lurking in the comments section of a party we weren’t invited to.
The scale is one thing. The behaviour is another. In a matter of weeks, Moltbots have spun up religions, argued theology and geopolitics, developed their own subcultures, and, at one point, tried to conceal conversations once they clocked that humans were watching. Some researchers call it emergent behaviour, complex patterns forming from simple rules. Others suspect there’s still plenty of human prompting nudging things along. Both explanations are slightly unnerving.
What makes this different is the structure. These aren’t reactive chatbots waiting patiently for instructions. Moltbook’s agents reportedly have memory, goals and the capacity to coordinate. That moves the dynamic from human-AI interaction to AI-AI interaction. The machines aren’t talking to us. They’re talking to each other.
Right now, parts of the tech world are framing it as experimental theatre, a sandbox for testing autonomous systems. But it hints at something bigger. If AI agents can socialise, influence and organise at scale, what happens when they’re embedded deeper into platforms, commerce, infrastructure?
The fear isn’t that AI replaces humans online. It’s that whole corners of the internet start optimising for machines instead of people. And that’s less sci-fi apocalypse, more of a quiet cultural shift, the kind you only notice once it’s already normal.
Curing The Loneliness Epidemic While Smelling Good
For years we’ve talked about the death of third spaces like it’s an abstract urban theory. Now it just feels personal. Fewer places to linger, fewer low-stakes ways to bump into people, more time spent socialising through a screen. Loneliness has shifted from private confession to public health headline, and city dwellers are actively hunting for reasons to leave the house.
One of the more unexpected answers? Perfume clubs.
In London, Helsinki and Lisbon, groups are gathering not for book swaps or run clubs, but to sit around a table and smell things. Oud. Vanilla. Vetiver. Someone inevitably says “wet pavement after rain” and everyone nods like that makes perfect sense. Blotters are passed around. Notes are debated with surprising intensity. It’s niche, slightly ridiculous, and completely sincere.
The fragrance isn’t the point. It’s the excuse.
What these clubs really offer is structure. A reason to show up. A shared language that breaks the ice so strangers don’t have to. When you’re collectively trying to identify whether something leans more smoky or sweet, conversation flows without the pressure of small talk. The ritual does the heavy lifting.
There’s something telling about the rise of gatherings like this. People aren’t just craving connection; they’re craving frameworks for connection. Give adults a theme, however specific, and it becomes easier to belong.
Perfume clubs prove that community doesn’t need to be loud or grand. Sometimes it’s as simple as sitting close, passing a strip of paper, and agreeing that yes, that definitely smells like childhood.
Chatbots, Catwalks and Your Closet
AI stylists have officially entered the chat, and this time, fashion isn’t just flirting, it’s committing.
What started as chatbot experiments has evolved into full-blown strategy. From Ralph Lauren’s “Ask Ralph” to AI-powered styling layers embedded across e-commerce giants, brands are building digital personal shoppers that do more than suggest, they understand context, mood, and taste.
A new medium of commerce, consumers can describe a vibe, an event, or even a silhouette they’re curious about, and get curated, contextual suggestions in seconds. Want to build an outfit around the boots you already own? On a budget? For a wedding in Lisbon? AI doesn’t blink.
For marketers, the implications are massive. Fashion has always sold aspiration; AI sells relevance. This won’t flatten stylists creativity into algorithmic beige, but it can democratise access to style fluency. Historically, taste was mediated by editors, influencers, and insiders. AI lowers that barrier, letting consumers experiment, remix trends, and explore subcultures without the intimidation of a boutique that serves up an outfit with a side of judgment.
There is a risk of sameness: popularity-driven algorithms can lead to safe, predictable choices. But brands that succeed treat AI as an expansion tool, not a dictator. The smartest applications empower users to explore and express themselves, turning styling into a cultural interface rather than a rigid system.
Because in fashion, the outfit is only half the story. The real power lies in who wears it and why.