Years ago, I was photographing one of the exhibitions we designed in our early days as a brand consultancy. While I diligently followed the light meter to ensure perfectly exposed photos, I later realised that human error had run its course. I had messed up the camera settings, leaving one of the images notably underexposed.
But, despite this, the photo became my favourite of the set.
I learned so much from this. What was clearly a mistake on paper had so much value in it, and being human is being able to read and judge that value. Call it "taste."
That unpredictability is a mark of human authorship. It's the spark that makes the Guggenheim Museum Bilbao so iconic—a lateral, completely unexpected counter to the norm. It's also what allowed Nike to make the raw, chaotic and anti-polished commercial "Nothing Beats a Londoner."
Light meters work on averages to calculate the most logical outcome, much like AI. If we continue to unthinkingly use software such as ChatGPT, Claude or Gemini to drive creative output, that spark will be hard to find by 2030.
Creative ideas are cyclical, but that's OK, because every cycle is recontextualised with meaning. My fear is that Al will get stuck in a feedback loop if we don't rethink the way we use it and reintroduce unpredictability.
Read the full article here
