There are non-technical synonyms for the word, or expression, “feedback” but it provides a good term to use in many different situations. More precise terms may need to be used in physics but in social situations it works quite well. I do, or say, something and you respond (or the environment responds) in accordance with what I have done or said. Sometimes the connection between action and response is unclear and sometimes the response, or feedback, is prolonged or delayed but — when it does occur then that it is feedback that occurs.
It can be said that it is the feedback that is the principal desire. If I get something back that I do not like, or is not what I expected, then that tells me that I said, or did, something poorly. I need to change my actions to obtain the desired feedback. The actions, in themselves, don’t actually do anything. I can smile broadly in the middle of a forest and, without others around to react, I am basically just exercising the facial muscles needed to perform the smile. I can sing at the seaside with no one near and it will bother nobody (I used to be a good singer but, without practice and over time, I am not sure I can make that claim anymore). Or within the realm of mechanics (which is not a focus of this newsletter), pressing on an accelerator does nothing unless it is connected to something that will increase, or lessen, the amount of fuel entering the motor system. Pushing down on the accelerator has no use in itself — it is what it triggers that is of importance.
Feedback is an intrinsic part of learning. During Machine Learning or Generative AI training, feedback is a constant, vital, aspect of the process. The system has no inherent knowledge of good/bad, true/false, beneficial/destructive. But, as it receives feedback about the results created from input, it can learn. It can also learn incorrectly. If it is told that something is true when it is actually false then, once trained, it will not be reliable to give correct responses. This is the primary reason why any AI system has to be trained by people — and why it is mandatory that any input is validated for accuracy and facts.
Inaccurate feedback can create alternate worlds — and not just for Machine Learning or Generative AI. The novel “1984” by George Orwell shows what can happen when history becomes malleable — changeable at will. Once again, this indicates the need for feedback based on accuracy and facts. (“Truth” is more a part of philosophy as it depends on history, point-of-view, and access to information. “Facts” are observable and recordable — though such documentation must be precise and detailed.)
Feedback is a vital aspect of education in general. During the recent pandemic, many colleges and universities who were not already experienced in remote learning panicked and tried this or that peculiar approach to remote learning. Many failed. The mathematics department of a local community college back in Maryland decided that their biggest concern during remote education was cheating. Their answer? Don’t give tests back, don’t review tests for right and wrong answers — eliminate feedback. One of my sons had to take a course three times before passing — because, without feedback, it is impossible to learn. He was the one who had to repeat the class but the community college was the entity that FAILED in its duty to teach. This failure on the part of the college carried over when presented to a university that had the illogical premise that it is the student’s fault if they have to take a course over. It would have been better to have given our children a sabbatical from school until post-pandemic times.
Feedback can take many forms. It may be a physical reaction or a verbal one. It may take place in the absence of approval — implying disapproval. Note that explicit feedback is always to be favored as there will be times when most people will not be able to interpret implied feedback correctly — and some (especially those with poor social skills and abilities) cannot interpret implied feedback at all. Some justify implied negative feedback by saying it is “nicer” and “more polite” but, if interpreted correctly, there is no difference to the recipient. And, if not interpreted correctly by the recipient, it causes more confusion and problems. The only advantages about the implied negative feedback is that it allows the giver an attempt to deflect any discomfort for giving the feedback.
At base, feedback enables potential directed change. Without feedback, change can only be random. Feedback is needed for education — with education being defined as adding new verifiable knowledge and the ability to add new verifiable knowledge. Feedback is needed for healthy social relationships. Feedback is also needed for physical/mechanical actions.
I am trying to move my blogs onto the substack newsletter community. If you are interested in these thoughts, then please look at, and subscribe to, my substack newsletter. Please click here, enter your email address and subscribe (free or paid). Thank you.
