MIT AI Researchers Warn About Addiction To Artificial Intelligence

Two AI researchers from MIT, Pat Pataranutaporn—a researcher at the MIT Media Lab who studies human-AI interaction with a focus on cyborg psychology—, and Robert Mahari—a joint JD-PhD candidate at the MIT Media Lab and Harvard Law School who focuses on computational law—published a joined article warning about addictive AI companions at the MIT Technology Review magazine.

In the piece, the experts explain that they have analyzed a million ChatGPT interaction logs and discovered that sexual role-playing is the second most popular use for AI chatbots—the first use is for creative composition— and shared their results in a paper just a few days ago.

“We are already starting to invite AIs into our lives as friends, lovers, mentors, therapists, and teachers,” wrote the researchers. “AI wields the collective charm of all human history and culture with infinite seductive mimicry.”

Pataranutaporn and Mahari are concerned about our current massive experiment rolling out in real-time without true knowledge of the consequences in our society and as individuals.

“As AI researchers working closely with policymakers, we are struck by the lack of interest lawmakers have shown in the harms arising from this future,” wrote the experts while highlighting the urge to combine law, psychology and technology research for AI regulation.

AI researchers explained that IA companions become addictive because the technology can identify people’s desires and its submissive nature knows how to serve users as they wish. Combined with the already existing addictive social media algorithms and the integration of new generative AI technologies, the experts believe it can easily grow into an extremely addictive technology.

Mira Murati, OpenAI chief technology officer, has previously mentioned the addictive qualities of their technology. During an interview last year, Murati said that ChatGPT models could become “extremely addictive” and that as users we could become “enslaved” to this technology.

This warning arrives only days after OpenAI rolls out its voice feature for users of the paid version of their product, and days after a new startup goes viral for launching a new AI necklace called Friend that is constantly listening and interacting with users.