AI peeps are nailing the comms

In June 2026, Anthropic put up a paper When AI Builds Itself. It suggests that AI is about to design and create newer versions of itself without human input. This is frightening as it brings to mind scenes of robots replicating themselves and taking over humanity - not helped by the virus looking animation the paper comes with. This is the final blog in our series looking at this paper and considers the communication techniques used to scare the pants off us.

Emotion trumps data

It may surprise you that our brains are not so good with cold, hard data & facts, particularly when it comes to remembering something new.  In fact, the best way to have any sort of communication ‘stick’ is to connect with people emotionally - and boy are the AI companies nailing this.

An Example

Take the title of Anthropics’ paper ‘When AI Builds Itself’. What’s the first thing you think of or see in your mind? Some of you might picture the terminator regenerating, or the replicators from Stargate or something equally scary from sci-fi movies or books.

Most importantly, how does it make you feel? (You might also want to click into the paper When AI Builds Itself & check out the graphic - does that change or intensify how you feel?).

In an instant you’ve probably created a whole sense of what the paper is about, maybe even formed some opinions about it and I suspect in a few weeks, months even, you’ll still remember it. Being able to generate an emotional connection from any sort of communication is incredibly powerful and the messages will stick.

Not convinced?

Have a think about the last thing you heard about AI - particularly from one of the big AI companies. I bet that you clearly remember the key points without too much difficulty. Try and consider why you remember it so easily - chances are it’s because it scared the pants off you. With a few choice words & phrases they’ve got their message through.

It’s not right

The fact that AI peeps use of fear to get their message across is succeeding, doesn’t mean it's good communication. Good communication should also be based on reality. When you dig deeper into their messaging, the reality often seems to be a lot less scary. AI is not building itself - Anthropic employees are using Claude to code a lot more (see my previous blog what aren't they telling us); agents haven’t gone rogue - the AI guys left big holes in the supposed ‘secure’ sandbox. They’ve succeeded in scaring us witless, without the evidence to back them up.

Grounded in facts

Towards the end of the When AI Builds Itself article, there is a list of three possible scenarios - not based on evidence, just the author’s imagination (particularly the last two).

Would you be willing to make key decisions for your life or work based on some ideas someone simply dreamt up? I wouldn’t. Good communication is not only about getting your ideas across, but it provides a starting point for making decisions. So, while you want your messages to connect emotionally, they do need to be based on real evidence in the first place and it always pays to provide the proof of what you’re saying.

The Challenge

Try and be aware of the emotions you are experiencing whenever you hear the latest AI headlines. Take a deep breath and ask what the evidence behind the message is (if any) and does it line up with what’s being said.