Oh my god, did you see Brian’s hat? Because we are all Brian in the context of AI
[click to view transcript]
In the early days of social media, there was a lot of buzz about ‘context collapse’. This is the idea that people communicate with some basic expectations for what context their communication will live within. For example if you’re talking to your friend on the phone, you expect your communication to be to your friend and not to let’s say your boss at work. But when social media came along, all the sudden you might be sharing with your friend in mind, but your boss could somehow end up seeing it. This was termed context collapse.
Today I’m pulling from a skit from the show I Think You Should Leave and from a 2018 paper titled ‘Does context really collapse in social media interaction?’. In the paper, they put forward an explanation founded in interactionist theories that context actually expands rather than collapses.
The tl;dr for their argument is that online interaction both creates and lives within its own context, so if you look for example at comments made on a social media post, each piece of communication actually works to clarify, shift, or expand the context. Stated simplistically, online communication is action and action generates more context, meaning it doesn’t collapse. BUT here’s the thing…
The Brian’s Hat skit from I Think You Should Leave is a courtroom scene where a text message exchange gets read out loud. I will try to walk a fine line here between referencing the skit and spoiling it. Neither of the defendants expected their texts to each other to end up being read in front of a courtroom. Their experience is actually a decent example of both context collapse and context expansion, depending on how you look at it.
So then we come to Brian and his hat. Now Brian, who loves his new hat, did not choose to be discussed in this text message exchange. Yet he was. He didn’t know that his behavior would be described in detail, that his words to his coworkers would be transcribed, and he certainly didn’t expect any of that to be read out in a courtroom.
You could argue that Brian, by sitting in the audience in court in his hat, is taking action and shifting the context. But I don’t think that’s quite right. I actually think Brian is a great representation of us–all of us–as we live through this moment in AI development.
To co-opt the framework in this paper, AI has retroactively taken action on every online post or conversation that gets used as training data. And because of this new and extensive interaction from AI, the context of every post has radically expanded. My Brian’s Hat (just way less funny and thankfully not embarrassing) is, for example, the articles I posted as a mid-career designer. Today’s recruiters and sales people are using them, filtered through AI tools, to determine what my personality is. Just like Brian, rolling his hat down his arm in a meeting room, I don’t think any of us expected our words and actions to expand into such unintended contexts.
I would love to have a term to describe this AI-driven, Brian’s Hat moment of radical context expansion, so if you know of a term better than “Brian’s Hat” please let me know. Thanks for listening.
Sources: https://www.degruyter.com/document/doi/10.1515/applirev-2017-0119/html and ‘Brian’s Hat’ from I Think You Should Leave, Season 2 (currently available on Netflix and YouTube)