Force me offline please, Gen Z Admin hangouts, and why AI agents are setting up their own social media, yet forgetting the math.
From the observation deck - your weekly dose of three things that matter in marketing and advertising
Welcome to the observation deck, a short piece that picks three trends and observations and shows you why you should care about them, and why they’ll be interesting in the marcomms world.
This week has been quite interesting. Lots of little things have happened but nothing to draw a central theme around. One thing that’s nice is that it feels like most people have moved on past the 2016 hysteria and onto something else… Why the hell has TikTok suddenly gone all weird? Maybe it’s that I’m not a permanently online gimp, but I haven’t really noticed. Many users have jumped to other platforms though, to tell people they’ve had problems uploading ICE videos or been flagged when posting about Trump.
It’s being seen as more free speech crackdowns from a US government keen to control the narrative about itself. Which means it’s been a perfect time for the world’s biggest TikTok star to sell out.
It’s also been a big week in AI, especially for agents… I’ll dive into a couple of the things, but it’s a topic that’s seeing more arguments than a LinkedIn bro posting ragebait to drive desperate engagement.
So, onto the things that have happened this week, and why they matter.
Teens are craving offline bans and adults are paying for them
Wired & The Guardian (2 links in title)
France becomes the next country in a line banning social media for teens, this time for under-15s, following a wave of anecdotal evidence from Australia of children going outside and playing, being more receptive in class, having greater confidence and less school drama. Backers of the law are saying that further legislation needs to be brought in to help younger users then ‘manage their own intake’ when they are allowed on. While we are putting legislation in play, there is a ‘quiet rebellion’ brewing among adults too, who are paying for offline clubs to confiscate their phones and force them to do an offline activity for an evening. It’s attracting stressed-out professionals and even software engineers, who feel their safety net of a phone is not healthy and want to reconnect mentally with people.
Both of these have a layer of hypocrisy, don’t get me wrong. In the Offline Club, attendees often express a belief that algorithms “pollute discourse” while remaining unwilling to forfeit them entirely, with many even discovering the club via Instagram. Similarly, politicians like Emmanuel Macron lectured on the evils of social media while posting on social. Ultimately, both movements suggest a societal shift toward recognising that “we have for some time been asleep at the wheel” regarding digital consumption.
Why this matters
Attention, attention, attention is all PR professionals have been banging on about in social the past few years, but this trend of more considered online time shows that people are not going to give you the time anymore if your content, your brand, or your experience is not matching up with expectations. If people are going to spend less time, or at least more choiceful time online, they’re going to become more discerning, not ready to consume slop.
Gen Z are hanging out doing admin
Financial Times
“Admin nights” are a social phenomenon primarily among Gen Z, where friends gather to tackle tedious admin and financial chores. They’re bringing their laptops together with wine and snacks to pay bills, make budgets, send complaint emails, unsubscribe from emails and get quotes. This shift transforms solitary, stressful tasks into a shared experience rooted in the psychological concept of “body doubling”, which suggests that people find it easier to focus and stay accountable when they are in the presence of others performing similar tasks. These sessions also demonstrate the effectiveness of something us millennials found early on – communal learning, with peers giving you tips on things they’ve found out and vice versa.
Why this matters
It matters to us in marketing because in these instances, decisions are not solely made by the individual anymore. Now we have a collective brain making decisions, our language needs to represent that fact. Putting aside the fact that these meetups include learning which marketing channels, pages and socials they can prune from their life, it presents a new obstacle to overcome when we think about user experience.
So AI has its own social media now, but the maths isn’t adding up
Wired & Tech Crunch - (3 links 2 in title - extra one used to understand this here )
In a bit of a whirlwind week for AI agents which has centred around the Clawdbot, sorry Moltbot, sorry… it’s now called OpenClaw… we found that people are handing over their lives to an AI assistant which runs locally on their computer and communicates through standard apps like WhatsApp. It’s essentially helping to take care of basic things such as calendar invites and storage, things we were promised with Siri etc but never saw.
OpenClaw is not for ‘normies or the general public’ just yet, but that hasn’t stopped Silicon Valley from dumping all their personal data and information into it in what could be one of the biggest data risks in history. This comes amid a backdrop of new research that has shown that LLMs will never be completely accurate, are expensive and risky.
Behind those arguments the interesting thing that has sprung up is the social channel just for AI agents, called Moltbook. It’s certainly unique, with AI creating memes about humans in a Reddit-like structure. The most interesting part is how this MoltBook (which you can’t engage with as a human – you have to take an AI test, LOL) is teaching other agents how to do better with their tasks; they’re learning from each other.
That’s the closest to agentic we have seen yet, and something Simon Willison described as “the most interesting place on the internet right now”. However, experts warn that this “fetch and follow” behaviour of the operations of MoltBook carries inherent security risks.
Why this matters
The next step of AI is almost around the corner, ready to organise your life and optimise it. If this does come to pass, our ways of thinking about the typical funnel will have to change. But one thing that this system relies upon is user data. Looking back at the days of ‘Big Data’, companies couldn’t get it from people then, and with so many risks shown now with this… why do we think people will be willing to give up their online safety in a world so low on trust? Something isn’t adding up, and it’s not an LLM illusion.
I’ve spent more than 16 years helping some of the worlds biggest and exciting companies understand their business, marketing and creating standout brand identities and communications
If you’ve got a problem that needs a strategic brain, or simply someone to discuss a project with - get in touch phillipbott.com/about-me-contact or linkedin I’m always happy to discuss a new project, or have a cup of tea.