We sat down with our Digital Director, Neil, to chat all things AI, craft and cognitive surrender.

The impact of AI on craft and creativity is a critical one. As AI continues to develop, and integrates itself more and more into everyday life, we must consider how we can keep craft, the human touch, and creative sparks alive in our work.

Are we thinking less because AI is thinking for us? 

Perhaps optimistically, I don’t think so. AI is a tool and, like any tool, it can be used skilfully, or poorly. I remember when carbon fibre tennis racquets came out. All of a sudden you could hit the ball further, faster and harder – but that didn’t translate to more shots in court. In my case, I still lost the game, just faster. It’s the same with AI.

If you surrender your thinking to the machine, you’ll get somewhere fast, but there’s a good chance it won’t be where you want or need to be. So where does that leave us?

Lean into good practice. Keep hitting the ball against the wall. Engage your brain. Read. Write by hand. I’m only half-joking when I say the most important tools of the AI era are your notebook and a good pen. Because using those sets you up to think. The words that form on the page are different from the words you clatter together on the keyboard, or that an LLM predictively unspools. Make handwriting part of your daily practice.    

Is craft under threat, or just changing shape? 

If you’re not serious about craft, then, sure, it can be vaporised. If you want to, you can generate almost anything with AI. I dabble in songwriting and it is now possible to generate entire songs, in any genre, with AI. But the results are inauthentic and mediocre, a kind of summation of averageness. Where is the value in that? What is the point? If you’re serious people, craft prevails. In fact, there is craft in using AI effectively and intentionally, not to replace, but to complement and augment our beautifully human efforts.   

When does useful offloading become dependency? 

The excellent Azeem Azhar discusses cognitive offloading vs. cognitive surrender in his newsletter Exponential View:  

I used to know dozens of phone numbers by heart. Now I know two, and I don’t miss the rest. That’s cognitive offloading: a strategic delegation that costs nothing. Cognitive surrender is something different; an uncritical abdication of reasoning itself.  

As Azeem goes on to note, AI is uniquely charismatic as a technology. Almost without realising, we could find ourselves drifting into uncritical dependency. I think the early signs can be identified in our habits – the notebook left unopened, the unread primary docs piled into a folder for AI to read instead. So, conversely, we should cultivate good habits. Anthropic’s co-founder Jack Clark outlined a set of good habits in a recent lecture:  

  • Read the primary material directly 
  • Form an opinion before asking the model 
  • Maintain independent practices — reading, music, sport, craft — where it is “you versus the world,” minimally mediated by algorithmic systems 
  • Do not defer entirely to AI systems even when they are usually right 

In a world where AI can produce good enough content at volume, what does excellent actually look like? 

I don’t think the look of excellence has changed. To me, it looks like effort, craft, and thoughtfulness applied at scale. Achieved via the age-old process. Try again. Fail again. Fail better.  

Zero-click is now the default. What does a communications strategy look like when there’s no page to land on? 

There is still a destination, it’s just at the level of the aggregator or synthesiser, rather than the originating page. Side note: we’re still seeing this play out, but I think ultimately we will need to find a way to reward publishers. Simply harvesting their content is not sustainable and if the system fails, we all lose out.

That aside, as communications professionals, we need to understand how the AI platforms generate their answers. What is the ecosystem of sources? This is what led us to develop Signal, our GEO programme. For any organisation, it maps whether you’re named, how prominently, in what tone, and which sources shape the answer. Once you can see that, the strategy follows.