How do we get to knowing what's going on?
Learning is working towards running your life on a more accurate system, model or mental idea of your ecosystems
=
Societal learning is working towards running society on a more accurate system, model or mental idea of our ecosystems
🦾♥️🧠🤖
HOW MACHINE LEARNING MIGHT HELP US UNDERSTAND HUMAN LEARNING ...and the other way around too.
🤓
CONSIDER SKIPPING THIS NERDERY PART
Recent neuroscience research highlights the role of prediction in human cognitive processes. Predictive coding theories suggest that the brain constructs internal models to anticipate sensory input, and learning occurs by reducing prediction errors between expected and actual input.
Studies show that when sensory expectations are not met, the brain generates error signals to refine its internal models, rather than passively responding to input. This predictive mechanism is foundational in understanding perception, learning, and even consciousness.
🤓
The key difference between machine learning and human learning lies in how they acquire knowledge and make predictions.
🤖🦾
Machine learning relies on learning data (a dataset) to train algorithms. It uses this data to identify patterns and relationships, building a population of predictions based on probabilities. These predictions are applied to new data to make potential predictions within the learned scope. Machine learning is bounded by the quality and size of its training data.
🧠♥️
Human learning integrates information from direct experiences, reasoning, and emotions, often beyond structured learning data. Humans adapt creatively, extrapolating outside an explicit population of predictions. Their potential predictions may consider abstract, context-specific, and moral dimensions.
In essence, machine learning is systematic and data-driven, while human learning is experiential.
🤓🥸
I was recently asked how my partial background in experience economics and experience design relates to learning. Long story, but essentially:
Human learning is experiential ✨
When groups of humans add human-biased filters to machine learning models, we design them towards the experiential. While these filters aim to make models more experiential and human-centered, they risk reinforcing existing societal blind spots.
Quick attempt at an illustrative figure below adjusted from this research paper.
