[Handprint4469]

From the article:

> Sarah was, in fact, changing. She felt herself getting smaller, constantly appeasing. She apologized, again and again and again. “I have been selfish, childish, and confused,” she said. “As a result, I need your protection.” A spokesperson for Huberman denies Sarah’s accounts of their fights,

Having played around with local LLM chatbots using different models and templates ("characters"), I find this pattern of talking _remarkably_ similar to how chatbots would respond to certain situations. The repetition ("again and again and again"), the use of multiple adjectives in a row as self-description ("I have been selfish, childish, and confused"), and the explicit closing of sentences ("As a result, I need your protection") fits almost perfectly with what I would expect from a waifu/husbando chatbot at the ChatGPT 3.5 level.

Does anyone talk like this in real life? English is not my native language, but this all sounds extremely weird to me.

[RangerScience]

Manipulation always involves diminishment.

Abuse always involves isolation.

Coercion always involves fatigue.

Symptom; method; cause; tool; good ends or bad; doesn’t matter. If X is present, Y is also present.

[queuebert]

Perhaps vapid, pseudo-intellectuals are overrepresented in the text of the internet because of their excess blogging, tweeting, etc. Thus the LLMs are biased toward that way of speaking.

[canjobear]

It is indeed a very weird way of talking, characteristic of prepared statements and forced apologies.