Her Words Don't Show Attraction; Her Behavior Does

Can ChatGPT tell who is attracted to whom? Not really – but neither can people reading date transcripts. You must be able to OBSERVE BEHAVIOR to tell!

Here’s a paper published this year from the University of New Brunswick’s psychology department that shows that, once again, most of attraction is not coded into words but behavior. From the paper:

What makes people 'click' on a first date and become mutually attracted to one another? While understanding and predicting the dynamics of romantic interactions used to be exclusive to human judgment, we show that Large Language Models (LLMs) can detect romantic attraction during brief getting-to-know-you interactions. Examining data from 964 speed dates, we show that ChatGPT (and Claude 3) can predict both objective and subjective indicators of speed dating success (r=0.12-0.23). ChatGPT's predictions of actual matching (i.e., the exchange of contact information) were not only on par with those of human judges who had access to the same information but incremental to speed daters' own predictions. While some of the variance in ChatGPT's predictions can be explained by common content dimensions (such as the valence of the conversations) the fact that there remains a substantial proportion of unexplained variance suggests that ChatGPT also picks up on conversational dynamics. In addition, ChatGPT's judgments showed substantial overlap with those made by the human observers (mean r=0.29), highlighting similarities in their representation of romantic attraction that is, partially, independent of accuracy.

The paper aimed to see whether an LLM like ChatGPT was capable of predicting who’d go out with whom from a speed dating event. So naturally, they focus on what ChatGPT was able to do in the abstract.

However, ChatGPT wasn’t actually that good at it – its predictions only correlated with reality 12% of the time – but that’s not the interesting part. The interesting part is that other humans simply reading the transcripts of the speed date conversations had almost exactly the same low level of accuracy (13%). Meanwhile, humans able to watch videos of the speed dates were 2.5x as accurate (31%); the participants themselves were on-the-mark a full 50% of the time.

From the paper

What that means is that if you are trying to judge a woman’s intentions toward you, you need to be basing that off her body language, not her words.

I’ll explain.