2
points
Dating is expensive, time-consuming, and often… exhausting. What if there were a way to test compatibility before even leaving your couch?
Here’s the idea:
* Your personal AI chatbot learns your personality from your social media writing.
* Instead of going on every date in real life, your chatbot "goes on dates" with other users' chatbots.
* Chats that go well in these AI simulations are then escalated to actual humans, so you only meet people with real potential.
* Basically, AI-simulated conversations filter out the bad dates so you can focus on the good ones — saving time, money, and awkward conversations.
Would you try a service like this?
First, I don't think there is any technology that could learn my tastes and personality to any degree of accuracy, and particularly not if the only training material was social media. Likewise, I don't think that I could tell compatibility by interacting with someone else's AI.
Second, dating is not just about testing compatibility. It's also fun in its own right, and offers practice in social interaction (amongst a host of other things). Even if AI were able to accurately determine compatibility, limiting your dating experience to just those people means you are losing out on exercising and developing important social skills and perhaps inhibiting personal growth. Without growth and experience, you'll never expand the range of who you are compatible with and your social experiences will remain fixed to a smaller world.
Bad dates, as unpleasant as they may be in the moment, bring their own value to you and offer opportunities for personal growth.
I agree that they do better on explainability issues in the sense that people could read the transcripts and say "oh yeah, I would/wouldn't have liked/enjoyed that", which they can't do if the OKCupid algorithm says they're a 92% match with someone or something. But how do we know that chatbots can be made a realistic enough simulation of a person to give relevant (indeed, more relevant) guesses about romantic potential compared to older matching methods?
Remember that existing LLMs have a whole lot of training that's not based on specific individuals at all, so to start with you have some component of "to what extent do chatbots that are told to simulate a conversation in a date end up acting like they liked other conversation partners?" with potentially a smaller component of distinctiveness of an individual.
(Another issue is that chatbots wouldn't know private or personal information that hadn't been revealed in training data but that might conceivably come up, if relevant, during a real date!)
Comparing to other algorithmic matching isn't that useful, because it's all terrible. But I don't see how this method would be an improvement.
This is not why they dated. They dated to test compatibility and learning about themself was a byproduct.