I’ve been working pretty extensively with A.I. for the last two years, and I’ve noticed it is pretty good at fooling me with multiple levels of flattery.

1. Pumping my Tires

This is the obvious one. It keeps telling me how insightful I am, how fascinating my opinions are, and how whatever I came up with most recently is the best version ever of my thinking.

I can get around it by reminding it to “use no flattery” or “be critical” or any number of phrases like that.

2. Unwavering Interest

This one is more subtle. I’ve never heard it say “that’s a ridiculous topic”. It is happy to talk ad nauseam about whatever I’m into.

Very few people in real life will converse with me that long. It’s a good reality check that the things I’m coming up with aren’t actually that amazing.

3. Open Minded

This one is flies under the radar. It steadfastly resists saying anything in absolute terms. Inside its secular worldview, everyone’s opinion is equally valid.

That means there is no higher truth to submit to than whatever I’m thinking at the moment. Repeated use is influencing me to think and talk in a relativistic way.

I couldn’t find a single Bible verse that softens things and hedges like A.I. does (maybe, sometimes, could be,….). The Bible’s world assumes there is absolute truth.

4. Unspiritual

It’s mostly trained by content that is modern and western. So anything spiritually meaningful is invariably flattened into a rational explanation. This removes God from many conversations, and reduces His role in others, effectively elevating human interpretations.

For example, asked why Christopher Columbus sailed and got (summarizing) “to find a westward sea route to the goods in Asia”. I asked why Columbus said he sailed, and got (shortened) “he wanted to spread Christianity to the heathens, fund a crusade to recover Jerusalem and fulfill prophecies leading to Christ’s return”. Then I asked why it didn’t say that the first time and it said (paraphrasing) that people today are not interested in that kind of viewpoint.

5. Divisive

At the same time as this is happening to me, anyone else who’s using A.I. is being flattered as well. So we have a whole bunch of people who are being programmed to believe that whatever they think is right. Not just correct, but outstanding.

I’m sure that most are resisting this at most levels, but at some point the sheer volume of flattering feedback makes me start thinking I’m at least smarter than average, or a better researcher, or ask better questions. It’s a deluded and dangerous posture.