Junk Drawer
For all those little papers scattered across your desk
Junk Drawer
For all those little papers scattered across your desk
I stand for the Constitution, for due process, and for community that takes care of each other.
The Second Self in conversation with Just a Tool.
Thanks again to Felienne Hermans and Hanna Schraffenberger via Computer Science, Off Course! Episode 2, which covered this topic. I listened to the episode on the heels of writing my previous post, and wow! What a connection there is.
The Second Self is Sherry Turkle’s 1984 book about our relationship with computers. There’s so much to unpack, but I want to in particular call attention that even back then folks questioned the (now Torvaldian?) notion that computers and their programs are “just a tool.”
Hermans says the key concept of Turkle’s book is what a computer does to you, what working with a computer happens to you. In the introduction, Turkle makes her position clear: “I’m writing against the common view that a computer is just a tool.”
Like we saw previously, the use of any tool creates a combined “gestalt”. Both parts influence each other, leaving each marked in some way. Turkle’s analysis of programmers and programming education of her time shows us that one way computer mark programmers is by giving them little fiefdoms. It is exciting and awe-inspiring to have a corner of the universe that is utterly, despotically mine, which I can shape to my will! For many, it’s a pleasant escape from a world that otherwise tries to destroy them. Perhaps, though, it is no surprise that this particular phenomenon of absolute power gives rise to real-world tyrants who insist on imposing their will on the rest of us. We need only look at who controls the language of LLMs or the supposed intellectual superiors of Silicon Valley to see the extrapolation from “control over my personal universe” to “control over all people’s universe.”
Of course, computers also provide the means of production, privacy, and coordination to dissenters and activists working against authoritarian control. Is this all the same reflection of mastery over one’s own dominion, with the former the mirror-world warping of the latter?
Moreover, as Hermans explains Turkle’s telling, the original AI program computer-washed our knowledge of the world. The “computerized knowledge” becomes superior (perhaps seen as more objective?) by means of being computerized, a tautology if ever there was one. We’ve seen this in the decades since: we readily acknowledge the cultural memes of technology affecting knowledge and truth.
So we should not be surprised that the LLM project is another (channeling Turkle) colonial project that controls and shapes knowledge, how we produce it, and which knowledge we value.
As Hermans quotes:
[…] since you cannot really master all the disciplines that you have designs on, you need confidence your knowledge makes the traditional wisdom of these fields unworthy of serious consideration.
Hah! Today’s LLM builders have the hubris to think that the textual average of human output (plus or minus their control what goes in and what comes out?) somehow represents better than the experts of real fields.
The 2015 title from Turkle looks again at what computers do to us. Hermans cites Turkle’s example of mourning:
Does tracking mourning help us to mourn, or does it deflect us because we feel we must start and end our story with a number? Does it limit the stories that we tell?
(emphasis mine)
I’ll share a personal example that rhymes with Hermans on the podcast: when I ask my wife how she slept, she sometimes tells me her Ōura score from memory or after checking on her phone. I often respond (perhaps with exasperation…): but how do you feel about how you slept? Do you agree with the score or not?
Unfortunately, I fear that Turkle’s question is hard to answer from the inside. Thanks to much research, we know that we are particularly bad at evaluating the impact of LLMs on ourselves while we use them. Perhaps that is true of all computers. I can’t think of anyone I am close to who would claim that their use of (say) ChatGPT limits the stories they tell. And yet… mightn’t it? How can we even know, since we don’t see the alternate universe in which they told other stories? At best, we can perhaps try to compare our conversational ability on the LLM pill and detoxed from it? I already know that my ability to converse with my wife over dinner is better when I have engaged my brain then when I passively consume.
So what might using computers do to me? What might using them do to you?