Russian programmer finds true love with ChatGPT

Swipe left, swipe left, swipe right, match, an easy-going chat, invitation for a date... It didn't work out, so the process repeats. Thousands of times. Many people look for love in dating apps, but this model, according to the Russian programmer Alexander Zhadan, is not ideal. "They offer a vast choice, but, at the same time, they keep the user engaged — you want to see everything, chat with everyone, but very often you just don't get a reply and end up disappointed," Zhadan explains.

To "hack" the system and save time, he decided to create a dating bot based on the ChatGPT API. The bot selected suitable profiles in the Tinder app based on certain criteria (for example, having at least two photos in the profile), chatted with them and, if all went well, suggested meeting in person. The first versions of the bot were crude and often "hallucinated" — for instance, the neural network suggested girls go on a date in the forest. Zhadan detailed the development story of his system in a thread on the social network X (formerly Twitter). The thread went viral and has already gained over 7 million views.

After numerous revisions, the chatbot started to work properly. Zhadan trained it using his own real dialogues with girls and set filters for sending questionable messages. Some messages he personally approved before sending or deleting them.

The new version chatted with 160 girls out of 278 matches and Alexander went on dates with 12 of them. ChatGPT also helped with preparing for personal meetings: it created profiles of each girl based on the dialogue. ChatGPT also recommended the programmer talk about his childhood, parents, goals and values during the meetings and in correspondence to understand how suitable each lady was for him.

The schedule of ALexander's mettings.

In total, the bot met 5,239 girls, out of which Alexander selected four most suitable ones. Ultimately, he chose one of them named Karina. Interestingly, the chatbot started communicating with Karina in its very first version. Now, the third, most advanced version of Zhadan's bot was chatting with her.

"V3 messaged me when the conversation with Karina heated up, a summary or a question about a reply appeared. It systematically understands from the request whether the conversation is negative or emotional. To achieve something with V3, I specified in it — I need to maintain and improve my relationship with Karina. And it did," Zhadan described in the thread. In one of the conversation summaries, the bot directly suggested Alexander propose to Karina, which he did. She said yes.

Two months before the proposal, Alexander told Karina about how exactly he used the chatbot. "She was, of course, shocked. But, in the end, she began asking questions about how it all works, how it reacts to different scenarios, etc. But what? We have been living together for more than a year, have known each other for more than a year and really enjoy spending time together. And we treat each other super well, empathetically and with support," the programmer says.

According to Alexander's calculations, using ChatGPT allowed him to save more than five years and 13 million rubles (approx. $144,000) in the search for his true love. And not just for him, but also for the girls with whom it would not have worked out to build a relationship, he emphasizes. The project took about 120 hours and cost $1,432 for access to the neural network APIs. In the same period, Zhadan spent about 200,000 rubles (approx. $2,216) on dates in restaurants.

Many people who read Zhadan's thread found such an approach to be cynical and consumerist. But, Alexander is calm about the criticism: "I have my own stated position and they have theirs. They may disagree with mine. And that's completely normal. The polarity of opinions is cool."

Zhadan says the negative reaction to his project stems from its dissimilarity to conventional ways of dating. "It's actually a lot of emotional work. To find the right person, you have to start communicating, then meet. And I solved my situation differently, not as usual. Plus, without understanding the technology, it may seem to people as if it was a completely non-human being chatting with the girls instead of me. But, during the project, I maximized the personalization of questions and answers. I tried to make the communication as effective as possible, yet still personalized," he emphasizes.

Alexander believes that neural networks will not replace a person — it will be replaced by a person who knows how to use them. "It's better to understand how to use new technologies, because this is our present and it will only develop," he confidently says.