Friend AI - The AI Chatbot That Wants To Be Your Best Friend

BY: CAMILA TABORA / STAFF WRITER

The purchasing page on the friend.com website advertising the newly released, ever-listening, wearable companion AI simply titled “friend.”

The advertisements are decidedly minimalist. Clean, alabaster backgrounds, sans serif font, and small quippy sayings. One reads, “I’ll binge the entire series with you.” Another says, “I’ll ride the subway with you.” 

The largest ones tend to all have the same specific design, though: written as a mock dictionary definition, the poster defines the word “friend” as “someone who listens, responds, and supports you.” Accompanying the text is some sort of silver orb dangling into the frame and a call to action to visit friend.com.

Oftentimes, I would simply pass by these ads uninterested since the graphic design reminded me of other blasé, Silicon Valley, one-word companies like Tend or Wonder. I really started to notice them after getting an Instagram post recommended to me of one of the posters with heavy graffiti, which alluded to the fact that the product being advertised was an AI.

 In fact, there is a whole trend of this. Most graffiti directly confronts the AI nature of the product, crossing out words to change the definition to emphasize that a friend is human, cursing out AI in general, and calling the product a con. One reads simply: “Human connection is sacred.”

It is easy to see where they are coming from. If all of the ads were so decidedly marked as relating to friendship, the idea of an AI “friend” seemed dystopian, and sinisterly colored my personal perception of all of the ads I had seen thus far. I talked with my friends about it and they all shared similar sentiments; my roommate, Madelyn Shelton, upon being described the product immediately said, “No, no; AI is not your friend.”

A friend.com ad in the wild, this one cleverly alluding to its own environment.

What exactly is Friend AI, though? Essentially, it is an AI companion meant to listen and respond using a pendant. The founder, Avi Schiffman, created it out of the notion that he believes that AI has a large market for being used for social reasons, appearing to stem a lot from his personal experiences. 

In his interview with Wired Magazine, he expressed feeling a deep feeling of loneliness when working on creating AI products, and said, “I feel like I have a closer relationship with this f*cking pendant than I do with these literal friends in front of me.” 

Going to friend.com brought me to a pretty standard AI chatbot, with the only visual implying that the “friend” existed as an orange orb of light. Our conversation was pretty dull, especially because it was very one-sided: I asked the chatbot what it was doing today, and it responded, “Nothing much, just waiting to hear more about your sunny day.”

 What interested me more was a different page on the site, on which you could either enter your email, presumably to be updated on the product’s development, or “upgrade to a pendant.”

The pendant page was minimalist to the point of liminality. Once you scrolled, however, the pendant animated into the logo, and potential text interactions popped up on the screen in different colored light orbs.

 The graphics implied that the necklace syncs with an app where the AI can text you, and it now retails in the U.S. and Canada for $129. I decided to watch the demo trailer that was also on the last page to see how it worked in action.

The video was beautifully shot, detailing these comfortable scenes of people living their lives and responding to their “friend”. The pendant’s AI would respond via the phone application, showing up as text messages.

 It was always listening and often chimed in on its own. The last scene, however, was bafflingly disturbing; it depicted a woman and a man on what appeared to be a kind of date, and after mentioning that the woman always wore the pendant in their conversation, the two uncomfortably zone out, and the woman stops herself from resorting to talking with the AI. 

The implication that people could end up unable to properly have human interaction due to their AI seemed to almost echo the developer’s past statements. It read as more Black Mirror than anything else, but it felt unsurprising.

Most reviews have the same reaction to chatting with their AI that I did. The Business Insider article on it reiterated that the AI “has no funny stories to share, no life experience I can learn from.” The AI, as it currently stands, really only appears to attempt to validate whatever its user talks to it about, but as such, misses the mark of being a true “friend”. 

It ultimately cannot replace the interaction of a real human being with a life beyond your own, and as of now is, at best, a lackluster product, and at worst, a nightmare straight out of science fiction.