The Messenger - Issue #23 “Modelling the Mente”
We have been talking about the various dimensions of wicked problems, but how do we create systems that address these problems? How can we respond to the world appropriately without being determined by it? How can we do so by being fair and just?
There are three levels of understanding (1)Model (2)Represent (3)Implement. Of these, the first feels the “softest” but is perhaps the most crucial step before we get into irreversible decisions.
This issue is about mental models, and how a deeper understanding of various model we can use in different situations and building a vocabulary, can help us deal with the wickedness of a situation.
Till next time!
Afterword
“Should human beings leave complex decision making to computers?” Once famously suggested by Daniel Kahneman and Amos Tversky, stemming from their work in how humans use heuristics to take decisions, and these decisions are full of errors and biases. On the other hand Gerd Gizenger’s work is critical as he and his team show how simple heuristics actually outperform or match with complex regression and computational models.
Simple Heuristics That Make Algorithms Smart - Behavioral Scientist — behavioralscientist.org Although simple heuristics often yield “biased” decisions, they can deliver a better answers. What might this mean for today's complex algorithms?