Will Technology Save or Destroy our Friends?

The solution is probably not only having AI friends

The Loneliness Epidemic

Ezra Klein thinks there’s a quiet catastrophe brewing in our social lives.

Between 1990 and 2021, there was a decrease of 25 percentage points in the number of Americans who say they have five or more close friends……

Loneliness is the result of a structure. It is imposed, in some ways, by culture. We make choices as a society about what we value. We chase our jobs. We live far from our families. Who move away from our friends. We spread out into suburbs and into single-family homes set back behind fences and lawns. We sprawl out with automobiles. We design for atomization and isolation. And so no wonder we get lonely. [1]

Here’s a key definition for atomization: the act or process of splitting into smaller parts, sections, groups, etc. And yet when we have 100% atomized our lives, this means working is meant for your career and income, the gym is for only for building muscles and burning fat, and that going on vacation is only for taking Instagram pics and making your friends not on vacation jealous.

And as Nat Eliason adds, it’s no wonder why we’re feeling so lonely as a society. Even when we specifically atomized our social time:

Then when we feel lonely, painfully isolated by our atomized life, we schedule some atomized social time like going to a bar or coffee to see friends in between our lonely work and lonely dinner because we’ve removed most of the natural socializing elements from all of the other parts of life. Atomization turns an integrated day of socializing, eating, exercising, and working into discrete hurried chunks of trying to move from one thing to another, wondering why we never seem to have time for everything.[1]

The solution it would seem is to be more open to the world around us and blend much of our lives together. Just like in high school when you would “study with your friends” (and effect not study at all), we should be finding ways to make sure our busy lives are maybe not so busy but full of areas that they can be blended together.

So why aren’t we de-atomizing our lives?

Is Technology the Problem

One interesting concept on the state of our society was from Strange Loop Canon’s piece about Seeing Like a Network. He agues in the beginning that we are all living on nodes in a network.

We’re living through a phase change that is at the root of a lot of our societal problems. It’s the fact that our information networks have become much more dense. You exist as a node in a network. Other people are other nodes. They send you information, the edges. You process it, you create your own. Information flows in all directions. There are all sorts of networks. If you imagine all of us as nodes and the information we receive from each other via edges, then the shape of the network defines the type of information that spreads. When the type of information is extremely tantalising, one that spreads fast, then the whole network gets taken over with that information. And there’s even a tipping point at which the information breaks containment and spreads through the whole network.

…

Ok, this is basics about what a network is. But what happens when the entire world gets interconnected such that we’re all connected to each other much more densely? What would have changed in ourselves? Our culture?

If we all exist on these nodes - our relationships overall get blended across these nodes.

For example, I have a friend in Seattle that I see once or twice a year when I manage to visit or when they come down to SF. Whenever we see each other, we catch up for a few hours over coffee or dinner. And yet I probably chat with them a few times a year over text, see their stories / posts on Instagram or Twitter, and maybe even get an update about their life from one or two other mutual friends.

And so while most of our interactions are in-person, the rest of it is actually a one to many or many to many information node relationship. The information nodes that make up our relationship and perceptive reality are a blend of personalized conversation and mass produced content of what they wants to share with their friends / world.

This happens for all of us. The version of Jay that you get can be very convoluted depending on what contextual cues we're in. If you know me through my Youtube videos on data science, it's really not the same as the perception my nieces get from their uncle “Bo Bo Jay”, which is then different from the thoughts that you might read when you receive my Substack emails.

This effect can be magnified with celebrities. Many famous podcasters mention how people will come up to just talk to them like they’re best friends, because they’ve listened to more hours of them talking than from their actual friends. And yet many of these podcasters while seeming totally approachable are extremely introverted and potentially terrified of strangers who are oblivious that this they could enjoy anything but talk about [INSERT SUBJECT] all the time.

How much of our friendships are then truly baked in reality? We’ve know that Instagram shows the best but false side of anyones life. But I’d argue that the inputs received from all of my friends is now “reality” whether I’m aware it’s a tailored perspective or not, I just synthesize it together with every other interaction to understand our relationship.

But how these information silos actually affect us in the long term is likely negatively impacting the feeling that our friendships are strong. If I’ve caught up with my friends through their Instagram Stories every day, is it making me feel like I don’t need to catch up with them or that they’ve left me behind? Or am I contextualizing some facet of their life without them actually explaining their reasoning behind their latest politically charged online post?

Can Technology Save Us?

One interesting passage from Warren Buffett’s biography The Snowball describes the unselfishness of his late wife, Susan Buffett. While Warren’s purpose was making money, Susan’s purpose was giving it away (and maybe slowly teaching her husband to do the same).

Susie counseled a long list of “clients,” traveled to see her grandchildren and hosted them in Emerald Bay, and accompanied son-in-law Allen Greenberg on Buffett Foundation business, which took them to places as far away as Vietnam….wrote hundreds of cards and sent hundreds of gifts, and traveled frequently with friends. And she willingly dropped everything to be at the bedside of anyone who was dying or ill. Peter (her son) took to calling his mother the Dalai Mama.

…

Her attention was increasingly taken up by an expanding number of “vagrants,” as one friend put it, who wandered through the house, sought her help, and occupied her time. Since she almost always accepted people unconditionally, some of those “clients” had pasts as felons, con men, addicts, or, in one case, as the purported proprietor of a bawdy house. From time to time these people conned her out of money. She didn’t really mind. Buffett was infuriated at the thought of being cheated himself, but he had eventually come to think of this as part of Susie’s giveaway budget and even accepted it as part of her charm.

Susan really exemplified the role of being a “connector” or someone who put in the effort to reach out and re-buy into the friendships as well as understand the fragility of human connection. And while Susan Buffett is an extreme example of this, in many ways she was a person who consistently sacrificed her life for others to the point of exhaustion for her well-being.

I think most of us cannot fathom keeping up with even ten friends at a time. And there’s some consensus by the Silicon Valley elite that the solution is an all-knowing AI like in the movie “Her”, where each person has an OS that caters to their every need, including companionship.

Apart from (spoiler) all the AI’s leaving Earth for bigger and better things, the world looked very distant and apart for a good reason. The sensitive and soulful main character started out lonely from their divorce, managed to fall in love with his OS, and then became lonely again from the end.

And while this might be fiction, it’s definitely did not paint the most optimal picture of what might happen if this occurs. “AI can solve this” is one way to raise $10 million for your next startup. Yet it’s a harder sell if by the time AGI occurs, we’re all destitutely living in VR headsets alone.

The Soft Landing

At the beginning of every month, I have an urgent small item on my to-do list that must be completed that day. On iPhone, I use Shortcuts to send an automated “Happy Birthday” text to all of my friends for the upcoming month that I track an Airtable spreadsheet. For some reason Shortcuts won’t allow me to schedule it out the whole year, which explains why I have to make sure on the 1st of the month to make sure someone doesn’t get a random happy birthday text.

This is a asynchronously high ROI interaction for me. By batching it, I now have 12 birthdays to remember instead of 100+. And I think everyone is better off for it given how meaningful it can be to get a text now that only your grandma and your old 1st grade teacher use Facebook.

I sometimes wonder if we’re just jumping the gun on the all knowing AI. While the first few months after ChatGPT’s release had everyone in craze that AGI was here, all content writers and software engineers were extinct, and we’d be enslaved to our AI gods in a few short years, we’ve come to the soft realization that we are instead better suited for building “human in the loop” systems that make our work more productive.

And yet making our lives outside of work better has been slightly more of an afterthought. If I think about the many ways that I myself can be extended to a Susan Buffett type figure, I think selfishly that I don’t really want to gift them an AI to them to keep them company, but rather I want an AI that can help me become better friends with my friends.

I think most friendship decay is on the high effort of just managing the workload. Our lives get busy with work, travel, chores, family, etc.. and yet we think little about how friends matter until we suddenly realize we have less friends than ever before. Maybe we think loneliness is an effect of growing up as we all change and become more introverted and have kids and careers. But I would push back and say that technology has existed to solve ALL of our problems, not just our work ones. And if that means building a cold email AI to generate more leads, or a Therapy AI to talk some sense into our lives, why not direct some of those resources to something that will instead enrich our own relationships as well.

In my mind, the future of AI is finding time for friends between our busy calendars without a constant back and forth, or a therapeutic AI that reaches out about our friends latest job interview or break up, or one that even guides me through a hard text message that I wouldn’t be able to send without a little push.

It’s personally compelling for me because I feel like establishing community has never been harder after moving back to SF. For the first five years I lived in San Francisco, I never found much issue making friends with a house full of rotating roommates and two different close knit startups. But now that I’ve atomized my life to living with my girlfriend in a house by ourselves and running a remote startup, it’s suddenly something that hits me after a week of realizing, if I don’t actually do anything, I could not see anyone for weeks.

Technology may come to save us soon enough, though we just have to hope we have enough friends around when it does.

Things to Share

• This piece on CSAAS prompted me down a rabbit-hole of how Metromile failed even though it was projected to be an amazing growth company during the SPAC / ZIRP period of 2021 (they ended up being acquired by Lemonade for nothing). Something I’ve been thinking about is the nature of fraud in tech / finance that on the fact don’t really seem like fraud. You could argue that Chamath really did think Metromile was bullish based on his points of argument or maybe did not do enough diligence to realize it was (as we know now) a pretty bad company. But at the end of the day - it destroyed shareholder value, wiped out any equity from investors that read the Tweet, and ultimately was not a great outcome for anyone involved.