🤖Grok, Chat GPT and Socionics

Question: What do you think about the socionic information provided by the neural networks Grok, Chat GPT, etc.? I would like them to rely on your socionics...

Yes, we know that GPT chat and its counterparts "know" something about typologies. Western networks mostly know about MBTI, our Russian ones know about socionics, but in all cases the information is "flat" and incorrect, because AI receives average information from the most popular sources (which are inevitably of poor quality), so the final result is not exhaustive knowledge, but simply a compilation of the most popular statements. Which are superficial in their essence.

The answers of neural networks about socionics do not make much sense, because it is clear that there is no deep understanding of the question behind the answer. The same applies to typing using neural networks - if there is no methodology at the core, then the result will be just a random mess of streamlined phrases with the same random type, that can be suggested just by a group of enthusiasts without any AI.

Thus, the task of "teaching AI proper socionics" means that we need to flood the Internet with our materials in order to shift AI's attention from behavioral socionics to our approach. This task is very difficult to achieve due to the fact that our ideas have a strict #IQ threshold - they are simply beyond the comprehension of most people, so outdated behavioral socionics wins only due to the dominance of the intellectual majority and, as a result, this becomes (and has already become) consensus knowledge within neural networks and beyond.

To teach #AI in modern #socionics is not a task that we consider interesting and realistic for ourselves; we work for people who are able to understand modern socionics, and it is more interesting for us to work with them. Even a neural network trained by us will still make mistakes and we will have to double-check it anyway.

If we are (very big) optimists, then this is no more than 1% of all people and no more than 10% of those who are interested in typologies at any level. Even within socionics itself, there is a serious segmentation of the audience.

So the popular behavioral socionic (and #MBTI) approach will always win due to Gaussian distribution of intelligence, and it's okay. We work for our potential 1% of people 😁