Revisiting this because my gf and I watched Social Dilemma last night and the conversation of social media as neutral came up.
Now that this piece is almost 2 years old, I’d love to see you revisit your assessment of AI. Do you feel you understand it enough now to make a value judgement?
But do I have to make a judgment!? If I was rewriting this article today (and maybe I will rewrite it someday soon), I'd perhaps retitle it:
Why AI is neither good nor bad nor neutral
or perhaps
Why AI is both neutral and not neutral
Part of the issue, I think, stems from the word "neutral." Neutral comes from the Latin "neutrum," which actually means "neither/nor." I think what peeves me most about people calling technology "neutral" is that it presupposes that you have more control than you really do with the technology you use.
People will be like: "It's neutral. I can use it for good or for ill."
And I'm like: "Try to kill me with a wet noodle. Good luck."
It's going to be much harder to end me with some cooked capellini than with a 12 gauge.
Calling a tool neutral (like calling a word neutral) appears to give you some weird power over that thing or word. Or, at least it gives you the semblance of power (but not the actual thing).
In reality, that tool comes with a host of assumptions, biases, and bents that dictate, in a very real way, how you use it.
What do you think?
BTW, if your GF was saying that technology is neutral, I give you full permission to politely tell her that she is mistaken.
Haha no she was in agreement that technology has many inherent bents and biases.
As for the word neutral, I understand your hesitancy. People think of technology like a multipurpose tool that can be applied however the user sees fit, but the tool dictates the use many times.
The way I’ve been thinking of social media recently, and why I’ve been distancing, is because of who you must become to engage with it. Maybe that’s not a great way of putting it, but it’s like how everything on TV becomes entertainment, even news, and so everyone who engages with it (whether producing or consuming) is either entertainer or entertained. Everyone on social media, whether a musician, congressman, or priest, has to become an influencer (and often more precisely a comedian) to use the “tool”.
People can argue that it’s neutral and that you can create an account and use it however you want, but if they’re being honest, they know that content that performs well, that suits the medium, has specific, unneutral qualities.
I’ve been wondering what percentage of content on Instagram is humorous. My guess is the vast majority. Priests that have accounts are doing dances and quirky interviews. Why? Because how else do we expect 30 seconds to be used? Does that seem like a medium for profundity? Even if a physicist is saying something profound in a clip or video, my bet is most of the comments will be comedic. Which is a massive part of the medium.
Anyway, this is rambly. I’d very much look forward to a revisit of this. I think it’ll only continue to need revisiting. Thanks for replying!
I've also had similar thoughts recently. The comic/humorous tends to bears repetition on the Internet better than most other genres/styles.
For example, I'm not sure how many times I've watched or shared Denver the Dog (link below), but it is definitely more than once. Unlike most "content" or "information" that immediately goes stale, people tend to rewatch (or share) that which makes for a laugh.
Perhaps the rhetoric best adapted to the Internet has something to do with humor. But that doesn't mean that everyone or everything should be funny.
The problem with the joker is that he easily slides into being a buffoon. And when you're a buffoon, nobody can take you seriously.
Wow! To make a value judgement, this piece takes the cake 🎂! I feel like I'm reading McLuhan unleashed. From what? I'm not sure but he's definitely off the chain or hook if you prefer water. Much wisdom here. I'll be returning for another dip, no question!
Thanks for this thoughtful piece. And thank you for pointing out the problem with calling technology neutral; all that matters is its use. This is not to say there isn't a moral alignment. Regarding AI, I've been thinking a lot about Charlie Munger's quip: “Show me the incentive, and I'll show you the outcome," coupled with Merleau-Ponty's "Matter is pregnant with its form." A moral assignment to a technology can't be separated from its use and the intent motivating it. A knife used for chopping vegetables is a cooking knife; a knife used to stab someone is a murder knife. What's easy to miss is that this is not the same as a technology being neutral. It's perhaps a bit like a moral "superposition," like Schrödinger's cat. Technology embodies its myriad uses; it is never neutral. Thanks again for a thought-provoking piece!
Thanks for reading, Jim! Good to hear from you. Indeed, every technology has a certain set of potentialities wrapped up in it. I can make a podium out of a block of wood, but I can't use water to do the same thing.
Would you be willing to expand more on how Munger's quip relates to AI? I'm interested to hear more.
Charlie Munger thought that once you understood the incentive of an actor, you could determine the outcome of the actions. Of course, his context was companies and businesses. In the case of AI, if one first looks at the incentives of the largest interests pursuing its development, one sees profit as the motive (or power, but those usually go hand-in-hand). Certainly, there is nothing wrong with profit, but in a business milieu, profit comes from only one place: the gap between price and cost. The best way to impact costs apart from material is to increase operational efficiency, lower labor costs, or both. So, a first-order outcome (among others) is reducing the headcount or the amount of time the headcount must work to achieve an outcome, thereby increasing capacity, reducing compensation, or both. A second-order outcome is that profit as a first-order outcome renders one blind to second-order outcomes! And what are second-order outcomes but a manifestation of matter pregnant with its form? The potentialities of AI as a technology beyond just the use it is put to. Heidegger might rephrase Merleau-Ponty as “matter is pregnant with phenomena.”
Concerning AI itself, there is no incentive because AI, so far as we can tell, has no will! Otherwise, AI appears to fulfill the role of any technology: following a command for efficiency. The material from which the technology is constructed may lead to it being more or less efficient (your podium made of water certainly won’t help a speaker be heard “over the crowd,” though as a block of ice, it would serve better, and yet still not as well as a block of wood). Still, the more the technology achieves its intention as an efficiency-manifesting device, the more it becomes ready-to-hand.
There are two incentives: one belonging to the maker and the other belonging to the user, and both are interchangeable between agents.
Revisiting this because my gf and I watched Social Dilemma last night and the conversation of social media as neutral came up.
Now that this piece is almost 2 years old, I’d love to see you revisit your assessment of AI. Do you feel you understand it enough now to make a value judgement?
But do I have to make a judgment!? If I was rewriting this article today (and maybe I will rewrite it someday soon), I'd perhaps retitle it:
Why AI is neither good nor bad nor neutral
or perhaps
Why AI is both neutral and not neutral
Part of the issue, I think, stems from the word "neutral." Neutral comes from the Latin "neutrum," which actually means "neither/nor." I think what peeves me most about people calling technology "neutral" is that it presupposes that you have more control than you really do with the technology you use.
People will be like: "It's neutral. I can use it for good or for ill."
And I'm like: "Try to kill me with a wet noodle. Good luck."
It's going to be much harder to end me with some cooked capellini than with a 12 gauge.
Calling a tool neutral (like calling a word neutral) appears to give you some weird power over that thing or word. Or, at least it gives you the semblance of power (but not the actual thing).
In reality, that tool comes with a host of assumptions, biases, and bents that dictate, in a very real way, how you use it.
What do you think?
BTW, if your GF was saying that technology is neutral, I give you full permission to politely tell her that she is mistaken.
Haha no she was in agreement that technology has many inherent bents and biases.
As for the word neutral, I understand your hesitancy. People think of technology like a multipurpose tool that can be applied however the user sees fit, but the tool dictates the use many times.
The way I’ve been thinking of social media recently, and why I’ve been distancing, is because of who you must become to engage with it. Maybe that’s not a great way of putting it, but it’s like how everything on TV becomes entertainment, even news, and so everyone who engages with it (whether producing or consuming) is either entertainer or entertained. Everyone on social media, whether a musician, congressman, or priest, has to become an influencer (and often more precisely a comedian) to use the “tool”.
People can argue that it’s neutral and that you can create an account and use it however you want, but if they’re being honest, they know that content that performs well, that suits the medium, has specific, unneutral qualities.
I’ve been wondering what percentage of content on Instagram is humorous. My guess is the vast majority. Priests that have accounts are doing dances and quirky interviews. Why? Because how else do we expect 30 seconds to be used? Does that seem like a medium for profundity? Even if a physicist is saying something profound in a clip or video, my bet is most of the comments will be comedic. Which is a massive part of the medium.
Anyway, this is rambly. I’d very much look forward to a revisit of this. I think it’ll only continue to need revisiting. Thanks for replying!
I've also had similar thoughts recently. The comic/humorous tends to bears repetition on the Internet better than most other genres/styles.
For example, I'm not sure how many times I've watched or shared Denver the Dog (link below), but it is definitely more than once. Unlike most "content" or "information" that immediately goes stale, people tend to rewatch (or share) that which makes for a laugh.
Perhaps the rhetoric best adapted to the Internet has something to do with humor. But that doesn't mean that everyone or everything should be funny.
The problem with the joker is that he easily slides into being a buffoon. And when you're a buffoon, nobody can take you seriously.
Good thoughts here, George!
https://www.youtube.com/watch?v=B8ISzf2pryI
Well said! It'd been a while since I saw that video and even knowing the plot it still got me. Thanks for sharing!
It is one of my favorites.
Wow! To make a value judgement, this piece takes the cake 🎂! I feel like I'm reading McLuhan unleashed. From what? I'm not sure but he's definitely off the chain or hook if you prefer water. Much wisdom here. I'll be returning for another dip, no question!
Thanks, AB. You da man.
Let's catch up soon.
Thanks for this thoughtful piece. And thank you for pointing out the problem with calling technology neutral; all that matters is its use. This is not to say there isn't a moral alignment. Regarding AI, I've been thinking a lot about Charlie Munger's quip: “Show me the incentive, and I'll show you the outcome," coupled with Merleau-Ponty's "Matter is pregnant with its form." A moral assignment to a technology can't be separated from its use and the intent motivating it. A knife used for chopping vegetables is a cooking knife; a knife used to stab someone is a murder knife. What's easy to miss is that this is not the same as a technology being neutral. It's perhaps a bit like a moral "superposition," like Schrödinger's cat. Technology embodies its myriad uses; it is never neutral. Thanks again for a thought-provoking piece!
Thanks for reading, Jim! Good to hear from you. Indeed, every technology has a certain set of potentialities wrapped up in it. I can make a podium out of a block of wood, but I can't use water to do the same thing.
Would you be willing to expand more on how Munger's quip relates to AI? I'm interested to hear more.
Hi Justin,
Charlie Munger thought that once you understood the incentive of an actor, you could determine the outcome of the actions. Of course, his context was companies and businesses. In the case of AI, if one first looks at the incentives of the largest interests pursuing its development, one sees profit as the motive (or power, but those usually go hand-in-hand). Certainly, there is nothing wrong with profit, but in a business milieu, profit comes from only one place: the gap between price and cost. The best way to impact costs apart from material is to increase operational efficiency, lower labor costs, or both. So, a first-order outcome (among others) is reducing the headcount or the amount of time the headcount must work to achieve an outcome, thereby increasing capacity, reducing compensation, or both. A second-order outcome is that profit as a first-order outcome renders one blind to second-order outcomes! And what are second-order outcomes but a manifestation of matter pregnant with its form? The potentialities of AI as a technology beyond just the use it is put to. Heidegger might rephrase Merleau-Ponty as “matter is pregnant with phenomena.”
Concerning AI itself, there is no incentive because AI, so far as we can tell, has no will! Otherwise, AI appears to fulfill the role of any technology: following a command for efficiency. The material from which the technology is constructed may lead to it being more or less efficient (your podium made of water certainly won’t help a speaker be heard “over the crowd,” though as a block of ice, it would serve better, and yet still not as well as a block of wood). Still, the more the technology achieves its intention as an efficiency-manifesting device, the more it becomes ready-to-hand.
There are two incentives: one belonging to the maker and the other belonging to the user, and both are interchangeable between agents.