FFD Primer

A Personal Letter to the Uninitiated

(No special background is needed for this.)

The brain processes information in a way that is extremely powerful, fast, and efficient.  I was apparently the first to discover both the brain's particular kind of logic and how connections between brain cells are organized to generate it.  These discoveries were presented in the Related Articles

Electrical components connected in the same way can emulate the brain's method of processing information.  Virtually all of the method's advantages can translate to electronic computational systems.

Fuzzy vs. classical logic

The brain uses a special form of fuzzy logic, and most electronic computational systems use classical logic.

In classical logic, a statement is either true or false.  These two truth values are typically represented by 1 and 0.  In fuzzy logic, a statement can have any degree of truth from absolutely false to absolutely true.  The truth value is typically represented by a number in the interval from 0 to 1. 

We make hundreds or even thousands of decisions every day using fuzzy logic.  While driving a car, if the car in front of you is slowing rapidly and is not very far ahead, you step on the brake hard.  Is the car ahead slowing rapidly?  It could be maintaining a constant speed, or it could run into a brick wall, or it could have any degree of deceleration between these two extremes.  The same "fuzziness" applies to "The car is not very far ahead" and "You step on the brake hard."  You make a decision and step on the brake with the appropriate force, probably adjusting as you go with more fuzzy logic.  Note that a robotic car must make the same kinds of decisions.  

An FFD takes the truth values of several statements as inputs, and gives the truth values of the possible conjunctions connected by "and" and "not" as outputs.  In the simple example above, the inputs would be the truth values of "The car in front of you is slowing rapidly" and "It is very far ahead."  The output that could control the brake is the truth value of "The car in front of you is slowing rapidly AND it is NOT very far ahead."  Of course more than these two inputs would be needed to control the brake.  The input values themselves could be determined by other FFDs with many inputs from various sensors - eyes and ears for the brain, or video camera etc. for a robotic car.  

Analog vs. digital computing

The brain uses analog computing, and most electronic computational systems use digital.

Digital computing means discrete signals are used, usually only two that are designated 0 and 1.  Because of this, digital computing usually uses classical logic.  The main advantage of digital computing over analog is accuracy.  This is the reason most computers use digital computing.  

Analog computing means the signals can have any value in some interval, usually designated 0 to 1.  Because of this, and because fuzzy logic applications typically don't require extreme accuracy, analog computing and fuzzy logic are a natural marriage.  The advantage of analog computing is the large amount of information that can be conveyed in the signal.  Another advantage is that the brain's form of fuzzy logic can be implemented best with analog computing.  The architecture is simple, and information processing is extremely fast.  Because of the newness of the discovery of the brain's logic, this advantage has yet to be widely recognized.


Electronic FFDs could have more success than existing digital systems in tackling computationally intensive problems.  As fast as modern processors are, there are still many problems that can't be solved fast enough.  Typical examples are robotics, speech comprehension, and simulations of large systems such as climates and economies.  Robotics and speech comprehension require processing massive amounts of data in real time, and large system simulations can currently take six months to a year on the fastest super computer. 

Please comment on this page here.  It's a work in progress, like the rest of the website.