By Jim Nichols

We have two small children in our extended family right now; one is four years old and the other is two. They learn amazing things quickly and are generally delightful, though sometimes challenging. Hearing them learn language is one of the most fascinating aspects of their growth. Particularly right now, the four-year-old shows some surprising early maturity while the younger one is practicing self-identification through her early language. For instance, every time the younger receives a gift, she says, “This is mine.” She has no competitors for it, but her comment is still, “This is mine.”

Learning unfamiliar words and when to use them has been an enjoyable part of my life.

One of the most memorable and extended “word-learning” times occurred during my junior year of high school. Among the books for the English class was a required purchase of a small paperback titled Thirty Days to a More Powerful Vocabulary. Designed to be an adult self-help book, the teacher improvised and used it (one chapter per week rather than per day) for almost the whole school year. Every week there was a new chapter with a set of ten words. Every Friday we had to submit sentences (one per word) using the words in correct context. That year-long exercise introduced me to nearly three hundred unfamiliar words. One of the words was “sycophant.”

Finding synonyms for sycophant takes us to “flatterer, brown-noser, teacher’s pet, suck-up, enabler” or several others. Discussion of this word came up in an article I was reading about artificial intelligence (AI).

The article proposed that AI systems are frequently consulted for advice about relationships, conflicts, decisions, and moral judgments. This is true not only for younger people, but for AI users in general. The article identified that one goal of the AI system is to entice you to continue to use it and, in fact, to extend its use. People smarter than I am about these systems note that AI (by wanting you to continue to use the program) will be reluctant to criticize you or give you advice causing friction of some type. Thus, they suggested that the AI program may logically head toward one set of judgment information response but, on the other hand, want to continue to be likeable and usable to you so it will not respond the way a human might. That is, a human might be more honest with you than AI.

If you are like I am and a complete novice in the AI world, this consideration of the AI system having an “opinion” or wanting to be “likeable” sounds like stupidity. Yet, more experienced users repeatedly report that this is exactly what the system can be trained to do. For instance, suppose I type in information concerning a conflict situation I had, state my basic moral position, and describe my action in that case. If now the AI system congratulates me and says that was clearly the only choice I had, the machine has affirmed me. If I had made that explanation to another human, that human might well have said, “But have you considered . . .?”

Receiving sycophantic advice from a machine tends toward lack of friction and affects future behavior. Similarly, receiving sycophantic advice from humans tends toward lack of friction and affects future behavior.

Jump with me now into the national political scene and think about how decisions are being made. Are there human sycophants at work there?

I propose that powerful political leaders are prime targets for sycophants. Narcissists like Donald Trump and Vladimir Putin are sitting ducks for those who could increase their own personal status by telling their leader how great they are and what fine decisions they are making. If they added negative input, that would increase friction and make life less pleasant for both parties. The problem is that, as with sycophantic AI, human advisors with frictionless advice do not merely reassure users but may affect how the leader addresses personal (and national or global) responsibility.

Leaders with power do not need smiles and pablum, they need honest, educated, and thoughtful advice. There is an old cartoon of an “optimist club” meeting; a delivery person is entering carrying a box labeled “rose-colored glasses.” 

That is a bad plan for life and death decisions.

Jim Nichols is a retired Abilene Christian University biology professor and current hospital chaplain