So I have this framework for how I see all human relationships | Part 1/3: Statefulness
Celine Dion expressed as only vowels just the old MacDonalds song
2 upfront notes before we get into it:
Look I know I've been away for like four months - this isn't what I wanted for us either! But I'm on week 13 of 16 consecutive weeks of travels or having stay at home guests, which is I suppose what summertime is. Also, we bought a house (spectacular macroeconomic timing!) and moved during the fine shavings of time that were the interstitial spaces between trips so it's been a crazy few months.
But, imagining Sisyphus happy, as ever, I'm falling back to being a creature of habit and this means I'll actually finish the half dozen snippet columns I've drafted in the last few months.
I missed you. I still love you and we are so back, baby!Also, you know, with all this travel and as I'm still working on figuring out what elicits consciousness in me, I've been more intentionally listening to music (vs podcasts and audiobooks) and I've realized there's a specific micro genre(?) of music I love, which deals with the subject matter of wistfully joyous traveling. Wanted to share 2 songs that I love listening to when traveling:
good vibes!
Ok so listen, this column as I wrote it ended up being really LONG, because it's IMPORTANT and hopefully INTERESTING and perhaps TRANSFORMATIVE. I keep my columns under 2K words (because I earnestly try for under 1500 lol) but this one has three parts that I think fit really well together so I'm splitting it and sending it in 3 parts - expect parts 2 and 3 in the coming weeks. Here we go!
Prelude
One of the great things about spending my twenties in a very dense city like New York was that my rate of collision with other people was just very high. For better or worse, I met a lot of new people, and I met new people often. In some ways, it was exhilarating to experience that kind of novelty, but at times i also found it really paralyzing.
I've written before about how the notion of making friends in New York was in fact really difficult, and that the persistent, high throughput nozzle of humanity aimed at me every week had a bit of a chilling effect (sonder, from the ever excellent Dictionary of Obscure Sorrows, if you recall!). I struggled then to square why two things that felt intuitive:
high frequency of meeting new people is great - I get to better and better approximate the exact kind of friends i'm trying to make, and test hypotheses often!
low stakes encounters are great - barrier to entry and exit are low so everyone can be more open and experimental!
were still not serving me well and I was left feeling unfulfilled and disconnected.
Eventually, as I was struggling with this, I ran into a fairly transformational framework that has defined how I see all all relationships - an intersection of two ideas: a) statefulness and b) the concept of finite and infinite games.
Today I'll cover statefulness.
The big idea - the metaphor of state or statefulness borrowed from computer science is a defining aspect of how we experience humanity with other people.
Statefulness, explained
Statefulness is an idea I encountered initially when reading The Most Human Human by one of my all time favorite authors Brian Christian. The book from 2012 is an wildly fun read, exploring the incredibly prescient idea of what it takes for computers to converse with people convincingly. The basic idea is that the Turing Test (when machines can convince humans they are human via conversation) was actually run as a competition called the Loebner Prize and was run roughly as follows:
every round a panel of human judges would converse briefly with a robot and a person at the same time
judges would then rate the degree to which they believed their conversant was a human
the computer that earned the highest score of appearing human won the prize
but because of the scoring there was also a human who could earn the prize of appearing the most human! - this is what the author tries to win in the book
One of the core ideas Christian explores in the book is the idea of state, and how having stateful conversations (that is, conversations that don't begin from zero and instead have 3 core characteristics: memory, context and adaptability) is a critical part of how we impute humanity. Chatbots that could successfully channel context and seemingly retrieve past information were judged to be more human. Conversationalists that best utilized those memories and context and adapted the conversational tone and content to them were often winners. Interestingly, one subversive tactic to trick judges was to design chatbots that would steer conversations to places that were inherently stateless, like arguments! Imagine things like "shut up, you don't know anything!" or "ugh you would say something stupid" "i don't care what you have to say, all i know is..." which provoke high emotionality but actually don't follow from any specific context - this was a sweet spot for tricking humans!
Thinking about statefulness as the core marker of humanity in my interactions really blew my mind.
The absolute state of my interactions, I tell you
This was the key concept that helped me figure out how to reconcile my social interactions in the city. Even though I was getting a lot of shots at low stakes interactions, none of them were particularly stateful! Everything was the same set of cursory questions about what we did for work, where we lived, what the weather was, etc. There was no build, no growth, no experienced story of a human being. I started thinking about people I met in two different groups:
Group A - low state - friends of friends at parties, the server at my pizza place, people you meet at conferences, fellow subway riders, etc.
Group B - high state - coworkers, roommates, family, friends, etc.
Seeing that, there is maybe a temptation to believe that interactions with Group B are inherently better than those with group A, but in fact I'd argue they are simply different and both can be really great.
While Group A has low statefulness and you don't often get past a surface level of small talk (that can last hours) you can also engage in fun exploratory behavior without a lot of consequence.
Just got back from study abroad and wanna try pronouncing it Barthelona? This is the group for it.
Think you can be a better con artist than Anna Sorokin? Here's where you try!
Heck, wanna just pretend you're having a much better day/week/year/life than you actually are? Bring it on!
There was once a reddit comment I saw that said something like "Most people unwittingly bring a lot of baggage into conversations they don't have to, not realizing they can just leave a lot of their personal background at the door." and I think about that very regularly. Most of the time, most people don't know most things about you - that's radically freeing! This is group A and these are the joys of low state!
Group B on the other hand are people, for better or worse, that we are stuck to and stuck with. We will see them again and again so we have both the opportunity to see change but that means the things we do with them also have consequences! You can't just be weird one week and hope your roommate forgets the week after. All sides have some investment in these relationships staying predictable, so you are inherently constrained in what you can do. But on the other hand, it's a group that imparts a richness of story with every interaction. We are allowed to reminisce, to conspire, to build, and to never let our siblings forget their many petty mistakes. It's the best!
Once that clicked, I had a better sense of the share of all my interactions that I needed to be stateful (spoiler: a lot) and I could figure out ways to spend my time on constructing more repeat interactions vs novel ones. I've since been big on having regular scheduled touchpoints with my closest friends. I either set up straight up recurring calendar invites, or have ritualistic triggers (when I'm coming home from x, I call y.)
At work, when I meet or onboard someone new, I default to 3-5 intro calls every 2 weeks to a few months apart. This lets me both repeat exposure to the person (spaced repetition is one of the best ways to learn!) and see how they're adapting and growing into their role and whatever else is going on in their life (how was that vacation you went on, home improvement project you started, etc.)
Cal Newport has this idea in Digital Minimalism that it's hard to stop a habit (in his context, doomscrolling) by itself, but the more successful path is instead to crowd your time with things you love and force a tradeoff that squeezes out the unfulfilling stuff. This is what I ended up doing. Because I was setting up so many repeated interactions with people, I had less time for bottomless brunches and house parties, which ended up serving me well. The latter didn't go to zero, of course, but sort of right-sized as a share of how I interacted with people, and how much novelty and freewheeling statelessness I desired.
What does this mean in the age of AI?
The Loebner Prize no longer exists, and its discontinuation is no surprise given where we are in 2024 with multiple foundational large language models that mimic conversation very well. After all, the idea of state has long been in computer science, and has informed areas like protocol development, natural language processing (NLP), human-computer interactions (HCI) as researchers think about iterative ways to better approximate human interaction.
Amara's Law states that we tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. My guess is that applies here in a weird way too - the ease of manufacturing state hasn't caught up to our felt sense of what's real. For example, when I was a senior in college desperately looking for a job, I used to literally sneak into research buildings with my laptop. I would walk down a hallway and then quickly google the professors names I had seen and double back to ones that looked interesting and pretend I had opportunistically stumbled across their office as a long time fan ("OMG Professor X? Hey, I loved your paper from last year where you talked about...") Sometimes it went great, but sometimes it got real awkward as the conversation developed and I flew right off my knowledge cliff. Today, finding information and collating it quickly to provide context and more-than-cursory depth is a LOT easier and faster, and I could probably get a lot father with that approach now. At least, until we all catch up with our expectations.
So, as statefulness becomes more commodified (literally more tokenized!), it's interesting to think about how our standards will change for what "feels" human. Knowing things about someone, even if contextually specific and reliant on memory, will become easier over time - imagine a Google Glass style AR app that scans someone's socials and gives a summary brief on publicly available information from their professional field, location, hobbies, etc. With that squeeze underway, what's left for us to really differentiate authentic caring and connection?
What a time to be alive!
As always here is a thematic QR code for this post if you want to share it, or just enjoy the fun future we live in :)
