Thank you for your question @0xlide
The fuzzy controlled interest rates would be drastically different from the Morpho Blue dynamic IRM and a novel implementation of a concept widely used in real-world applications but, to our knowledge, not yet introduced in the blockchain space.
While Morpho Blue uses a dynamic PD controller to handle interest rates – which is relatively simple to implement and can suffer from reactivity delays when IR fluctuations are dictated by broad market movements – the fuzzy-based interest rates would be implemented using fuzzy logic. Fuzzy logic is a concept of control theory where control dynamics are expressed using flexible, human-reasoning-like rules evaluated against readout values (the values on which the IR controller defines its behavior).
Fuzzy logic expands on traditional Boolean logic by introducing degrees of truth, which mimic human reasoning. For instance, people tend to describe a room temperature not only as either very cold or very warm, but rather with a wide range of values from very cold to comfortable to reasonably warm to extremely warm. This approach suits interest rate control in a protocol like Aave, where defining such rules mathematically is challenging and computationally intensive.
Without going into too much detail, here is a presentation given by @Emilio on the same approach used for a dynamic borrow rate module for GHO:
