Key Takeaways
- Dating apps have been experimenting with AI this year, including recommending matches based on values, as a way to boost conversation, or even match folks based on their long-term goals.
- AI can act like an honest best friend, helping daters understand what they want and where they go wrong as a true co-matchmaker.
- AI needs human oversight. Without it, AI could nudge users toward empty “good on-paper” matches or amplifies bias.
Remember the good old days when Tinder felt revolutionary, when downloading an app instead of surfing the web on your desktop for a date became the norm and using dating apps felt like something out of “The Jetsons”? Well, after the initial novelty of dating apps wore off — hello, swipe fatigue — what’s next? Naturally, using artificial intelligence on dating apps.
Just like I used to feel about the hype over Bumble or the freedom of matching with someone 0.2 miles away, AI is bringing a new height to the dating game that folks haven’t seen before. And you may be thinking, “But Hayley, isn’t AI just those silly chatbots that never work?” Or profile prompts that suck, or maybe a “smart match” here and there to mix things up?
But it’s becoming so much more. Dating apps have met their matchmaker, and in case it’s not obvious, AI is acting as a real matchmaker, now. Yes, really. But is that a good thing? And who cares? Could it make dating better, forever?
With all the crazy AI hype, I believe we (yes, I’m looking at you and your team in the industry) are facing a turning point: If you run a dating platform, you need to decide whether AI is central to your matchmaking philosophy or meet your match, trying.
When It Actually Started Feeling Real
It’s no secret that apps have been experimenting with AI this year, including recommending matches based on user values, using AI as a way to boost conversation or even match folks based on their long-term goals, not just mutual swipes or bio keywords.
Take Lior Baruch — the co-founder of AlgoAI Tech — for instance. He describes in a recent article how they’ve used their AI-matchmaking to match folks based on their ambitions in life.
“We took about 1,500 different research criteria this year to understand what’s really working, what makes a good relationship, and what we can do to help people really find a match.”
And what they found in their research is that AI can help. AlgoAI recently launched a-love — a matching platform that offers organizations, communities, and brands a customizable, white-label solution to build meaningful connections through data and AI — to meet the demand we’re now seeing for AI matchmaking.
Why, though? Because the data talks: Users are tired of swiping through endless faces. According to a Forbes Health Survey, 78% of surveyed dating app users said they felt emotionally exhausted by it. Just like I am tired of swiping endlessly on non-monogamy apps through a bunch of people I’d never be interested in, other dates are looking for a real connection over volume.
Hence the new wave of AI as a matchmaker. It makes sense to me… but, as with all things, it has to be done right.
Whether you’re a fan of AI or not, one thing is true: It’s coming. So we may as well make it help provide real tools that can help people find matches, right? I, for one, think predictive compatibility is pretty cool, like matching based on lifestyle goals, or the elusive communication style. Another potential AI feature I’ve seen, like conversation analysis, can examine responses and a user’s tone to gauge how emotionally compatible someone is with their match. I’m not sure how I feel about that one, yet.
But I do love the idea of verification via AI, that can filter out bots or mismatches.
When AI Becomes a “Co-Matchmaker”
You know the best friend that tells you the honest truth about that terrible guy you dated in college? Well, what if your team could implement AI that acted as that? It could help folks really understand who they are, what they want and where they are going wrong. In that sense, AI becomes a co-matchmaker.
Plus, we all know someone or have been that person who doesn’t know or can’t articulate what they want in the dating game. So maybe the co-matchmaker could help a dater figure it out based on their app history, how they show up and what they say? It wouldn’t just match based on what users swipe or say, either. What if it could really help match folks based on their level of openness, whether they were queer-affirming or not, emotionally mature, etc.?
And some people really want this co-matchmaker in their dating profiles.
“If this AI bot has the same advice, or is programmed to have the same responses, as someone like me would tell them to say, then that’s a win,” says Lindsey Metselaar, a dating-podcast host on using AI for messaging.
Dating is hard. It’s always been hard. Sooo… maybe AI could make it that much easier.
Humans + AI = Matchmaking Evolved
But don’t get it twisted: AI only works well if humans are continually involved.
“We don’t want AI or chatbots to replace human connection or do the human connecting for us,” McLeod told Insider recently, “But it can facilitate getting us out on a date… and even help maybe coach us along the way.”
We still need and want human matchmakers. But would it surprise you to know that many of them are already leaning on AI data to match their clients? The difference, though, is that the AI can help reveal patterns while the matchmaker interprets context, values, ethics, and of course, the old gut feelings.
Take Sitch, a pay-per-setup AI matchmaking service that combines humans and machine learning to bring daters the romance they crave, founded by Nandini Mullaji. One user of the service said that for every first date they went on, there was a second date afterward. They attribute it to the human element and AI, combined.
And it could be that way, for many dating app experiences, too.
When AI Starts Suggesting Who You Should Date
Now, it’s not to say AI is all perfect as a matchmaker. I mean, c’mon, it’s a machine, after all. It’s all fun and games until AI starts recommending matches to daters based on what “makes sense” on paper and not what they “feel.”
A user may feel “nudged” toward certain types of people even if their heart wants something else. Or, even more importantly in my opinion, biases baked into AI could amplify. For example, underrepresented identities, like my polyamorous and bisexual self, could be filtered out. Plus, other marginalized groups, like minorities, may get filtered out if AI isn’t built consciously.
Obviously, things happen, but it’s a responsibility for app makers to own and take accountability. Otherwise, matchmaking could go terribly wrong and help none of your users.
Will AI Take Over or Evolve with Humans?
“Three to five years from now, it will feel relatively arcane to go through hundreds of profiles in order to get out on a date,” McLeod said about the future of swiping.
Oddly enough, I couldn’t agree with him more. I suspect we’ll only see more AI-first tools. But the trick? I don’t think that the ultimate dating apps will be pushing purely AI-first automation. They’ll be platforms that use AI and have human guidance, too.
Here’s my hot take: If you make dating apps, don’t just ride the AI trend. Do it in a meaningful way. Do it with the match in mind.
