Can AI Replace Human Friendship?

Startup founder Matt Oriente's take on the AI friend industry

I recently viewed a new video advertisement that flat-out disturbed me. No, this wasn’t a preview for the latest A24 horror film. It was a commercial for the wearable A.I. startup Friend, founded by Avi Schiffmann.

The ad displays numerous situations of individuals doing pretty common human activities (hiking outside, playing video games with friends, eating a gyro, etc.), but in each of these circumstances, individuals are checking their phones as messages sent from “Friend” are flooding in, offering playful and catchy replies to the verbal communication or outspoken events going on in their real lives. Well, “Friend” is not a human, but a small wearable A.I. necklace. The way that “Friend” can offer unsolicited banter and quick one-liners is through its small microphone that records live audio, analyzes what was said and by who, and then sends a text message reply to your phone. I suggest you watch the full advertisement as well, found here. 

Putting these unorthodox logistics aside (normalizing conversing aloud to a plastic rock and wearing something that is undeniably listening to your every second), my response was to try and understand the problem that this company is trying to solve. Based on their website and my initial intuition, I would imagine that Schiffmann would reply back to me with a single term: “loneliness.” And this is a serious problem today, especially in the US. As a country, about 60% of Americans classify as lonely. This was only exacerbated following the COVID-19 pandemic, as loneliness jumped 7% just between 2018 - 2020. We know that these feelings can also cause extensive health complications including increased risk of heart disease, stroke, and developing dementia. 

But interestingly, young adults are twice as likely to be lonely than seniors. Almost four out of every five (79%!) adults aged between 18 to 24 report feeling lonely, compared to 41% of seniors aged 66 and older. Furthermore, more than two in five adults (42%) aged 18 to 34 report “always” feeling “left out,” compared to just 16% of people aged 55 or older. This is unreal! Almost four in five young adults feel lonely in our country. So why is this? Well, it's too hard to pinpoint singular direct causes to explain this current situation, so I looked at historical data. 

Between the late 1970’s to the 2000’s, average loneliness rates for US high school students (typically 15-18 years of age) were actually decreasing steadily. However, we start to see some substantial increases between the timeframe of 2007 and 2012, which was the last year of data recorded in the survey. Another study shows similar results. Between 2000 and 2003, school loneliness was still decreasing. However, it continually spikes every ensuing reporting year, as seen in the figure below. 

Fig. 1. School Loneliness Means, Worldwide, by Sex.

Why has loneliness become so rampant? This cannot be attributed to a single cause, but the usual suspects of rising screen time, dopamine loss, and social media addiction are key players. Looking at the stretch between the mid 2000’s until now, a similar trend line appears: social media and screen time substantially increase. 

To an extent, we’re all aware of the increased trends of internet use over the following decades post-dotcom era. So what effects does this movement have today, especially on the adults of tomorrow? Well, 95% of US teens have access to a smartphone. That is not necessarily a bad thing, but what are they using it for and how often are they on it? 

Nearly half of all US citizens aged 13-18 admit to using the internet “almost constantly.”  A similar study found that 41% of these teens have a screen time of more than 8 hours a day. Additionally, average entertainment screen time among children has amassed to over 5 and half hours daily. The reason for these statistics? A growing dopamine culture fixated on social media dependence by scrolling apps like Tik Tok and Instagram. This trend is not going away. More than half of US teens now say it would be difficult for them to give up social media. 

Relying on technologies to avoid the social world and cope with social anxiety can heighten feelings of loneliness. Numerous studies have demonstrated this for years. In 2000, an experiment assessing just under 300 undergrad university students found that pathological internet use significantly increased their score on the Loneliness Scale, but nobody seemed to listen. Between 2000-2015, the number of teens who got together with their friends everyday dropped by more than 40%. So there is absolutely a relationship between screen time and loneliness. This is actually alarming. We’re a social species that craves connection. But online connectivity is clearly not working for everyone.

So now a tech startup is trying to solve loneliness by mass-producing a physical LLM that is going to constantly send the user quick texts. Thus forcing them to spend even more time interacting with their phones, which is one of the core drivers associated with loneliness. This almost qualifies for satire. Is our best solution for human loneliness really laying in a plastic necklace that floods my phone with dry, surface level reactions to my personal lived experiences? I personally don’t think so, but apparently others do. 

Let's look at the list of angels and VCs that invested in Friend. They include Perplexity CEO Aravind Srinivas, Z Fellows founder Cory Levy, Solana founders Anatoly Yakovenko and Raj Gokal, Caffeinated Capital's Raymond Tonsing, and Google senior product manager Logan Kilpatrick. Well this makes some sense, but maybe in a cynical way. Each of these media and tech moguls has either directly contributed to the rise of screen time or at least benefited from it, right? So the same people that aim to make money on this “wearable loneliness solution” are also the ones that in one way or another exacerbated this problem in the first place? 

The great Sun Tzu quotes “Victory comes from finding opportunities in problems”, but why is the solution arising from the same industry that intensified the problem? We are trying to create human-type artificial acquaintances to solve loneliness, when really the answer is just organic human connectivity amongst ourselves. We’ve known this for centuries, but the thing that Silicon Valley realized is that they can’t generate an ARR business model on creating real human connection. Instead, they have tried to sell technology like Friend’s as the novel panacea.

As a 24-year old urban single who works as a co-founder for an environmental startup, I am probably their exact target customer! Yet the concept of Friend’s product seems ridiculous to me. Sure, there are wild-eyed startups that come and go, promising grand ideas trying to forever change the world, but this is not the only company trying to mass produce your A.I. best friend. Just check out Replika or Rabbit R1 for a few more examples. 

I believe social media algorithms and addictive scrolling culture have captured our vulnerability and dependance by tapping into our evolved primal instincts. Our access to social media is the easiest it has ever been in history, yet our connectivity needs are not being fulfilled. I am worried that this model of monetizing humanity’s necessity for some sort of human connection is now being pursued to create other types of products and services that cater to the rest of our assured instinctual needs, things we can’t feel fulfilled without.

Out of the basic human instincts of connectivity, creativity, reflection, activity, sexuality and hunger (factors that helped our societies prosper during our history’s ups and downs), an argument can be made that modern tech startups are trying to sell products that either streamline an instinctual need or use it to further market other products. Is that really progressing the species? It’s easier to formulate solutions first when you initially form the problem. 

Time will tell if Friend really takes off. Who knows? Maybe I’ll rock a Friend in the next decade, but after doing the research, I’ve donned a tinfoil hat. Creating a scalable business model that seeks to profit by replacing the human connection that would truly meet our needs seems, well, non-human to me. It's not hard to think that social media has now poisoned the majority of us to believe we require some witness in our lives. This belief is fueled by the constant desire to glorify ourselves to our friends, family, and acquaintances. We’re also now comfortable with that witness being run on Python, and that's a whole other issue in and of itself. It’s as if we’ve traded genuine human connection for a digital viewer, and we’ve become strangely content if not more comfortable with it than actual humans. I just hope that next time I’m in a socially awkward position with another human, I can make a joke of it with my interlocutor, instead of resorting to checking my phone. And who knows, that situation may be the beginning of a long-term friendship.