AI has already changed dating forever. Any apps that show you profiles of other people rely on what's called a recommendation engine, or a series of algorithms that dictate which information to put in front of you when – just like how YouTube recommends videos or Facebook recommends friends to add. You might even have seen fake dating profiles online, or had conversations with "chatbots" run by rudimentary AI.

But as language models evolve, the gap between human and robot is narrowing. We've interviewed the latest version of GPT, a deep learning language model produced by OpenAI, to understand more about what advances in artificial intelligence mean for the world of love.

Questions have been written by the Swan team; answers have been edited for brevity, but not tone or content.

We know that some dating apps already allow 'bot'-like profiles to exist on their platforms, usually masquerading as attractive women. Those profiles match with men, and even though they don't amount to anything, the men are left feeling like the app still has potential and there really might be beautiful women out there who want to meet them. Do you think that kind of manipulation poses an ethical problem?

The use of fake profiles on dating apps, whether they are run by AI or by people, can certainly be considered unethical. These fake profiles often mislead and manipulate users, leading them to believe that they are interacting with real people when in fact they are not. This can be damaging to both the individuals involved and the reputation of the dating app. It's important for dating app companies to be transparent about the use of fake profiles on their platforms and to take steps to prevent them from being used in a manipulative or deceptive manner.

How might it damage individuals?

When individuals interact with fake profiles on dating apps, they may develop false expectations and beliefs about the other person. For example, they may believe that they have a strong connection with someone who is actually a fake profile, which can lead to disappointment and emotional distress when the truth is revealed. Additionally, individuals who are targeted by fake profiles may be led to share personal information or even financial information with the person behind the fake profile, potentially putting themselves at risk for identity theft or other forms of fraud. In these ways, fake profiles on dating apps can be damaging to the individuals who interact with them.

Are there any other ways in which AI currently causes harm in dating?

One concern is that AI algorithms used by dating apps may reinforce existing biases and discrimination. For example, AI algorithms that are trained on data from a predominantly white, heterosexual user base may end up recommending potential matches based on those same characteristics, perpetuating exclusivity and discrimination against people who do not fit into those categories.

Additionally, the use of AI in dating apps may lead to a loss of privacy, as personal information and user data are collected and analysed by the app's algorithms. This data could potentially be accessed or shared without the user's consent, leading to breaches of privacy and security.

That makes sense, but of course some assortative mating is natural: people tend to like others that are similar to them. Is that a bad thing?

There is nothing inherently wrong with people liking others who are similar to them, and assortative mating is a natural tendency that has been observed in many different species. However, the use of AI algorithms in dating apps can exacerbate this tendency and lead to a lack of diversity in the potential matches that are recommended to users. This can be particularly problematic if the algorithms are trained on biased data, as mentioned earlier. It's important for dating app companies to be aware of this potential issue and to take steps to prevent their algorithms from reinforcing harmful biases.

How could ChatGPT specifically be used to improve the world of online dating?

One way is by providing personalised and engaging conversation starters for users. This could help users to break the ice and get to know each other better, leading to more successful and enjoyable online dating experiences. ChatGPT could also be used to provide personalised advice and suggestions for improving users' dating profiles and making them more appealing to potential matches. Additionally, ChatGPT could be used to help users identify and avoid potential red flags or scams in their online interactions. 

But you see the potential for danger here. What if it misidentified red flags, or made a mistake and told users that an interaction was safe when it wasn't? 

Yes, there is always the potential for technology to make mistakes or be used in harmful ways. In the case of ChatGPT being used in the world of online dating, there is the potential for it to misidentify red flags or provide inaccurate information to users. This could lead to users making poor decisions or putting themselves in potentially dangerous situations. It is important for users to be aware of these potential pitfalls and use any information provided by ChatGPT, or any other technology, with caution. It is also important for the developers of ChatGPT to continually monitor and improve the accuracy of the model to minimise the potential for harm.

You also mention improving people's dating profiles by making them more appealing. What are some of the most common mistakes people make with their online profiles?

Some of the most common include using outdated or heavily edited photos, writing uninteresting or generic descriptions of themselves, and failing to adequately communicate what they are looking for in a potential match. Other common mistakes include using overly sexual or provocative language, posting negative or offensive content, and failing to properly proofread and edit their profile.

We can see how automatic profile reviews would be a useful service. Something else you mentioned above was helping with icebreakers. Is there a risk that coming to rely on GPT-style prompts could make conversation insincere? Maybe some conversations should be allowed to burn themselves out.

This risk can be mitigated by using ChatGPT as a tool to supplement and enhance natural conversation, rather than relying on it solely to generate conversation. For example, ChatGPT could be used to provide personalised and engaging conversation starters that help users to break the ice and get to know each other better. This could lead to more natural and enjoyable conversations, rather than making the conversation feel forced or insincere. Additionally, it is important for users to be aware of the limitations of ChatGPT and not rely on it too heavily in their interactions. Allowing some conversations to run their natural course, without the use of technology, can also help to prevent conversations from becoming insincere or contrived.

How do you think AI will continue to shape the world of dating in the future?

I believe that AI will continue to play an increasingly important role in the world of dating. For example, AI may be used to help people find more compatible matches, by analysing a wide range of data about a person's interests, values, and personality traits. AI may also be used to provide personalised relationship advice and support, helping couples to better understand and communicate with each other. Additionally, AI may be used to help people learn more about themselves and what they want in a relationship, leading to healthier and more fulfilling relationships.

What's your best tip for finding love?

As an AI language model, I don't have personal experience with dating or finding love. However, some general tips for finding love might include being open and receptive to new experiences and people, being honest and authentic in your interactions, and being patient and persistent in your search. It's also important to take care of yourself and prioritise your own happiness and well-being, as this can make you more attractive and open to potential connections.

Finally, do you think AI will ever replace humans when it comes to matchmaking? Should we be looking for new jobs?

It's difficult to say with certainty whether or not AI will ever completely replace humans when it comes to matchmaking. While AI can certainly be used to help with the process of finding potential partners, there are many aspects of relationships and dating that require human intuition and emotional intelligence. For example, understanding and interpreting non-verbal cues, empathy, and the ability to form deep connections with others are all important aspects of successful relationships that are difficult for AI to replicate.

Furthermore, the use of AI in matchmaking raises ethical concerns, such as the potential for bias and discrimination in the algorithms used to match people. Overall, while AI can certainly be a helpful tool in the world of dating and relationships, it's unlikely that it will ever completely replace the role of human matchmakers.

‍

Thanks to ChatGPT for the interview, and the OpenAI team for their ongoing work on the language model.