Skip to content

Harnessing visual learning in digital signage

I discuss how Tesla Autopilot has been trained through camera input, drawing a comparison with the way humans learn visually from the world around them. I also connect this to digital signage, where visual information can influence understanding and behaviour in ways we may not always consciously recognise.

Key takeaways

  • Tesla Autopilot is trained largely through visual camera data rather than written or manually fed information.
  • Human learning is heavily visual, especially in early childhood before language and reading skills develop.
  • Digital signage can work in a similar way, with visual cues shaping subconscious cognition and behaviour.

Topics

  • AI & Machine Learning
  • Neuroscience
  • Cognition
  • User Experience

Transcript

AI is a super hot topic right now, and all of those new large language models are being trained on data publicly available from the internet. However, one of the more mature AI models is actually Tesla's Autopilot. This is the model that helps auto-correct and drive the car. Now, this model was trained exclusively on camera input, so it was visually trained. Elon explains why. I mean, humans are photons in, controls out. Like, the vast majority of information reaching our brain is from our eyes. And you say, well, what's the output? The output is our motor signals to our sort of fingers and mouth in order to communicate. So what Elon is getting at here is that as humans, we learn a huge amount through our eyes. Think of a small child. Before they can speak, before they can read, they are learning very visually and picking up a whole sense of what the world is really all about, and then seeing how people are reacting to that. So in Tesla's case, by filming drivers, just having live inputs of film as they're on the road, we can begin to learn that's a bicycle, that's a potential hazard, we need to slow down now. All of this was done visually without having to feed data into the model. Digital signage kind of works pretty similar, right? It's visual, going into our subconscious cognition in the right environment. We learn so much more than we realize through our eyes than necessarily by what we're told to read and think about.