Modern romance gets an AI twist

Online dating apps embrace AI-powered tools for higher safety, better matches; some are looking at AI-based chatbots that can improve flirting skills

Swipe right for yes. Left for no. About a decade ago, the founders of the dating app Tinder came up with the swipe feature that revolutionised the way people date in the digital age. Today, app-based matchmaking has become a ubiquitous part of the dating landscape in India, with over seven crore users – from young adults to older generations – embracing dating apps.

Delhi-based Narayani Gupta, a frequent user of the dating app Meet7, says as a busy professional, finding time for dating seemed impossible a few years ago. But today, she says, apps are easy to use and secure too.

“The smart algorithm makes it easy for me to find compatible matches in my area. I appreciate the app’s focus on meaningful connections, and I was pleasantly surprised by the quality of conversations I had with potential partners,” she adds.

Dating apps traditionally came with downsides: identity fraud, misconduct, ghosting. Besides, casting your net too wide meant being bombarded with hundreds of potential matches – some real, some scammers. For women particularly, the experience could often be extremely uncomfortable.

Now, AI and machine learning are coming in to address these huge issues, just as they have disrupted almost all other industries. Many big players in the industry are increasingly making huge investments in their tech platforms to improve usability, personalisation, and safety.

Meet7, for instance, has introduced a selfie verification process. Instead of relying solely on still images, the system helps thwart bots and affirm authenticity of users by asking them to perform simple actions like smiling or blinking. The process further entails cross-checking selfies with over 16 distinct markers. This includes age, facial similarity with uploaded profile photos, and user-provided details such as gender. Should discrepancies arise, the profile is flagged for manual inspection. Until cleared, such profiles remain inactive and unseen by other users.

Saurabh Awasthi, CEO of Meet7, says the app uses an AI-driven feature called KnowQ, to break free from the superficial ‘swipe-only’ culture.

For Mumbai-based Tanvi Haru, safety is always a top priority when it comes to online dating. Personalisation, she says, makes the match-making process much easier. “I’ve had engaging interactions with like-minded individuals who genuinely resonate with me. It’s refreshing to know that I’m starting conversations with people who have a higher likelihood of being compatible with me. This personalised approach has saved me time and effort, as I no longer have to endlessly swipe through profiles hoping for a connection,” she says.

App developer Shrenik Shah, who has worked on dating apps, says most users are looking for a customised experience. “We largely rely on AI. We use analytics to rank the users based on the time spent on the app. Our aim is always to help the user reach their ultimate goal of finding a match closer to their expectations, as soon as they log in,” he says.

At TrulyMadly, every profile that goes live undergoes a moderation process where they leverage their proprietary gate-keeper algorithm, aptly nicknamed ChowkAiDar1.0. Founder Snehil Khanor says ChowkAiDar1.0 checks if there’s a human face in the photo. “Then it verifies that the face is not of a celebrity, checks for any obscenity in the profile photo, and also verifies that the gender on the uploaded photo is the same as specified on the profile,” he says.

Bumble has an AI tool that detects lewd images like nudes on the platform.

The tool was open-sourced last year. The company is also looking at an AI-based chatbot that will help users improve their flirting skills. The idea is to improve the confidence levels of users before they talk to real people.

Match Group, which owns dating apps like Tinder and OkCupid and which has several teams working on AI-related initiatives, announced last month that it’s establishing a new team focused on generative AI and other emerging tech. The group said that over the next six months, the company plans to launch AI-based features that would help users select optimal photos and use AI to highlight why a profile may be a good match. The company also said there are bigger AI projects that are in the works that will “more holistically improve the end-to-end dating experience.”

CEO Bernard Kim said in a recent investor call that by the end of the year, “we expect to have launched a number of initiatives that will use generative AI to eliminate awkwardness, make dating more rewarding and surprise and delight users, all in a way that focuses on authenticity and maintaining the highest ethical and privacy standards.”


GLOSSARY OF IN-GROUP TERMS USED ON THIS PAGE
Neural network
A neural network is a computational model inspired by the structure and functioning of the human brain’s neural networks. It is a fundamental component of machine learning and deep learning, used for various tasks such as image recognition, natural language processing, and more. Neural networks consist of interconnected nodes (artificial neurons or units) organized into layers, each with specific functions. Neural network models vary in complexity, architecture, and depth, with deep neural networks (deep learning) having many hidden layers. Convolutional Neural Networks (CNNs) are commonly used for image-related tasks, while Recurrent Neural Networks (RNNs) are used for sequential data, such as natural language processing.

Computer vision
Computer vision is a field of artificial intelligence (AI) and computer science that focuses on enabling computers to interpret and understand visual information from the world, much like the human visual system. It involves the development of algorithms, models, and techniques that allow machines to analyse and make sense of images and videos, extract relevant information, and make decisions or perform actions based on the visual input.

Level 5 autonomy (for drones)
Level 5 autonomy for drones, similar to autonomy levels in the context of self-driving cars, refers to the highest level of autonomous operation. In the case of drones, Level 5 autonomy implies full automation and complete independence from human intervention in all aspects of operation.