Could AI Entice You Into An Unhealthy Relationship?

Are you lonely? An AI chatbot could help, according to some. Given the conversational nature of chatbots and the fact that the persona some of them have is ‘chatty', some are treating these virtual human personas almost like real people. For instance, there's a chance that if you ask some of the GenAI chatbots about the movies playing in cinemas, the bot may ask whether you want a movie recommendation and that could spark a conversation. That's the big difference between search engine results and the personal, human-like response of an AI chatbot.

Ultimately, every tech company wants to keep users hooked in a world where engagement is king, so you cannot blame the developers for building models that are designed to keep you engaged. But what that could mean is that while parents for instance are worried about their child's screen time on Instagram or YouTube, they should perhaps also start to worry about time with AI chatbots because impressionable minds could get emotionally hooked. And this is not a sci-fi movie scenario—it may be happening with many people. While it will not happen with ChatGPT or Gemini, such bonds are certainly being built with AI-driven chatbots and apps designed to function as friends.

We have a fascinating piece this week on this trend, and it is a must-read, because it tells you about this topic, what is happening with people using such apps, how developers of almost every AI chatbot are enabling some of this (inadvertently or otherwise) and what needs to be done to protect users. 

And while he certainly wasn't talking about emotional attachments, OpenAI's Sam Altman has said that recent updates to the GPT-4o model have resulted in a chatbot personality that is "too sycophant-y and annoying." Anything with personality is a human-like feature and when I read this, I was wondering about the chatbots that want to us keep us hooked and want us to build deeper relationships with virtual humans.

We should be all the more concerned about this trend in India, because, as a study by Google and Kantar revealed, 60% of Indians they surveyed weren't familiar with AI, and only 31% have tried any generative AI tool. There was a time when Indians weren't familiar with smartphones or hadn't tried one and look where we are today. While it's a matter of genuine pride that India has leapfrogged many developed nations in mobile data usage, it's possible that as the AI wave brings about as much disruption as the internet itself (if not more), Indians will take to AI chatbots in droves. The data explosion in India resulted in many wonderful things—like how India leapfrogged developed nations in payments, but will it be all sunshine and roses with AI chatbots too? The Google-Kantar report speaks of how Indians have an innate desire to excel and how GenAI can help boost productivity, enhance creativity and communicate more effectively. But knowing Indian ingenuity, it is possible we need to look at how such tools could be repurposed to form emotional bonds which may not be entirely healthy. The benefits of AI far outweigh the bad and Google is absolutely right about how AI can benefit India, but it wouldn't hurt to build some safeguards.

Meanwhile, here are some of the other interesting AI-related stories from the week gone by:

• Duolingo To Go 'AI First', To Replace Contract Work With Artificial Intelligence

• Gen AI To Drive 43% Of AI Spending In India By 2025, But Adoption Slow: Lenovo Study

• Microsoft And Amazon Capex In Focus Amid Potential AI Pullback

• Meta Launches Standalone AI App In Bid To Compete With ChatGPT

• Sarvam AI To Build India's First AI Platform In Six Months