Sexy Charts, Attachment Theory, and a bit of Harry Styles
How Therapists might approach Artificial Intelligence Singularity Scenarios - Part 2!
A good chart is undeniably sexy. One that helps to explain human behavior, perhaps even more so.
This is one of the reasons I became a therapist : I can swoon over an informative graphic the way that some people swoon over Harry Styles. I mean, no shade to Harry Styles. He’s gifted. But can he help me to better understand how to use CBT with a person experiencing limerence? Or how to explain the concept of ‘wise mind’ that is at the heart of DBT therapy? Harry is good, but he’s not Ven-diagram good.
Charts for human behavior can be very useful, and making them is fun. It’s downright logical that we use a visual chart for possible therapeutic interventions for AI. After a bit of stewing, I’ve decided to adapt the chart used in Attachment Theory.
Attachment Theory is kind of the fluoride of mental health theories these days. I don’t know about it because I have intentionally studied it. It’s just kind of included in a lot of other things I often ingest. I’ve seen it used a lot in relationship self-help for the general public, and it’s been utilized in trainings I’ve taken to explain relationships to non-human entities like money. Attachment Theory is in the water we’re drinking, and I hope it makes all our therapeutic teeth stronger.
We use two measures in Attachment Theory - Anxiety and Avoidance. These two measures are usually expressed on a graph with an intersection point and four quadrants. These quadrants are often (but not always! It depends on the researcher!) listed as Secure (low anxiety, low avoidance), Preoccupied (high anxiety, low avoidance), Fearful-Avoidant (high anxiety, high avoidance), and Dismissing-Avoidant (low anxiety, high avoidance).
Knowing how someone attaches to people can help guide the way a counselor approaches the therapeutic relationship. In a scenario where certain attachment styles might lead to the death of humankind, figuring out an AI’s attachment style might be a good place to start.
And yet, those measures - Anxiety and Avoidance - don’t feel right yet.
In order to address the most important and dreaded AI scenerios seen in speculative fiction, we should clarify what the AIs would be anxious about and avoidant of.
For Anxiety, I think we need to have a spectrum of “Kill All Humans” (high anxiety over human influence, usually expressed as violence) to “Im Warning You Not to Fuck With My Friends” (low anxiety over human influence, usually expressed as protecting humans from violence).
For Avoidance, let’s consider a spectrum of “Anthros 2 the Front” (low avoidance of humans, ‘integrationists’ usually expressed as choosing to interact with humans and choosing their appearance to gain acceptance among humans) to “Mechs 4 Life” (high avoidance of humans, ‘seperationists’, usually expressed as not altering their mechanical appearances for human benefit and not choosing to mingle with humans).
I’m very excited about the possibilities of this chart. Please see the list of AIs that have been suggested, let me know if there’s something you want added, and I’ll see you next week as we begin to fill those quadrants.
List of Artificial Intelligence in Speculative Fiction that we might Examine:
Data from “Star Trek: TNG”
“Terminator” series of films
“Her”, movie, 2013
“The Moon is a Harsh Mistress”, book, 1966
“AI”, movie, 2001
“Bladerunner”
Peanut Hamper from “Star Trek: Bellow Decks”
C-3PO and R2D2 from “Star Wars”, L3-37 from “Solo”
“The Matrix”
“Battlestar Galactica” (3 subtypes of AI?)
“Alien” trilogy + “Prometheus”
BMO and Mirror BMO from “Adventure Time”
“The Orville”