Deep Learning in Training (Part 2)

Part 2: Deep Learning, AI and the Brain

I have no doubt that you’ve been on training courses and attended presentations. As a member of the audience, consider for a moment the difference between those that engaged you and those that really didn’t. Those that engaged you will have no doubt been more memorable; and you might even remember some of the things you learnt!

This could be the difference between ‘deep learning’ and ‘flat learning’.


Deep Learning and The Brain

Simply put:

  1. You have a brain… a massive network of neurons that is forever updating and forging new connections.
  2. As we learn, new associations are made across the brain where neurons wire up to other neurons. If two neurons (or sets of neurons) are triggered at the same time (or in close sequence), the brain will fuse those two together (known as Hebb’s Rule).
  3. Where we have knowledge, experience or expertise in a subject, there will be denser ‘hubs’ of connections within the neural network.
  4. Information we receive that doesn’t connect with a ‘hub’ is soon forgotten (i.e. it is not learnt) unless it is associated with a significant trauma.

Point 4 above is the key to ‘deep’ verses ‘flat’ learning. Presumably, we don’t usually want to create a trauma in an audience (unless it is a massively positive awakening/ inspiring/ transformational moment), so in order to create deep learning, we need to ‘plug in’ to the existing neural networks in our audience.


Deep Learning and AI (Artificial Intelligence)

To add another layer here, deep learning is used as a model in AI theory. Of course, you are already familiar with Artificial Intelligence and you will probably have your own opinions as to its pros and cons. AI often uses the analogy of a biological neural network, so here, we are borrowing the analogy back!

Deep learning in AI is about creating layers of processing and understanding, starting with raw data and building up to something more meaningful. For example, starting with lines and edges and then shapes and then patterns of shapes and then faces and then movement of faces to reading and understanding facial expressions.

Deep learning, as an approach to training, presenting and facilitating, works from the same principle of starting with the raw data (i.e. the audience’s direct experiences) and then building up from there to identifying other examples then patterns then more general concepts, models and approaches. As we ‘layer up’ we create more meaningful and useful information, hints, tips and techniques that people can take away and utilise.

Using the Deep Learning analogy from AI and understanding how the brain learns, we have our second rule of the Imaginarium Deep engagement approach to Learning (IDeaL):

IDeaL Rule #2   Start with the specific sensory experience of the audience (e.g. what they see, hear, feel, remember) and build on this to create new meaningful learning.

In part 3, we will explore an example of Deep Learning…

By Joe Cheal

The Model Presenter by Joe Cheal and Melody Cheal

For lots more information on how to engage your audience, see The Model Presenter. To order your copy click here.