Good piece. AI will help bring epistemic convergence and rigor after social media shattered it, which will be largely positive even as it creates new issues.
What makes you think or say this? Have you actually used AI at all? It's literally trained via reinforcement learning to provide answers that make people feel good, rather than what's truthful. It tells you what you want to hear, and if it doesn't, all it takes is the slightest nudge to get it to shift its position to yours. It's confirmation bias on steroids.
If we look at both AI and social media as tools (tech) which we use to shine light on reality/ourselves, then we will see things more as being "revealed by" than "caused by" them.
(I believe social media and AI can both go under the "largely positive even as it creates new issues" umbrella, even tho we tend to demonize one or the other — or indeed "all tech"!)
That said, I'm not sure what *exactly* the difference between causes and revelations is... 😋
"Instead, the action resides in the competing systems of interpretation through which people select, categorise, frame, connect, explain, and narrate those facts."
It's not different worlds, it's "Interpretive polarisation " - a new term to me.
i'd also argue, the desire for speed of reaction and outrage also short circuits our ability to think clearly and address our own biases of thought
As you said Dan - "Nevertheless, the core narrative is a gross simplification and distortion of a far more complex reality"
Great pièce. I read it telling myself that I still operate a pretty good pseudo reality in comparison with many people haha! But, seriously, I needed to read this as a counterpoint to all the overdetermined commentary from the right on the Novak murder.
You mentioned your students struggle with Lippman's criticism of "lived experiences". It might help them to consider a non-political example that they've probably encountered first hand: The dress.
I'm referring to the dress that some people interpreted to be blue and black, and others interpreted to be white and gold. Most people have "lived experiences" in interpreting the color of things, and so they felt very confident in their ability to interpret the color of this dress. And yet...
Great post, and I mostly agree. I have two aspects to nitpick, though. We are indeed cursed by the complexity of the human world in a way that doesn't allow itself to be reduced to the simple models that humans can handle - I wish we were a tad more like electrons!-. Still, I believe there exists enough truth of the matter even for social issues that could, ideally, be reached and agreed upon given a slightly idealized (but not impossible) scenario in which we had more general access to core facts and their more consistently fitting interpretations, along with a greater openness to consider different interpretations and not engage tribally-emotionally in them. This ain't utopia: to some degree, we already do it all the time (with subjects and topics we don't invest in).
My second nitpick is just my deep distrust of gatekeepers and technocratic solutions. I'd say my cynical support in liberal democracy mostly stems of a negative appreciation of humans: we are mostly self-serving, tribalist animals, but one advantage of liberal-democratic policies is we just put a partial leash on the ruling elites to restrain how much they just look after themselves and their individual and group interests and make it somewhat incentivized to try to maximize and generalize good outcomes for a wider circle. The problem with each and every technocratic class and solution is that they will inevitable bend things towards their self-interest, whether consciously or unconsciously, and will be likely intelligent enough to be able to rationalize their biases. We got this in the Great Awokening, and there's no indication it won't happen again. In theory, we *could* set up incentives for maximally truth-seeking and truth-generating institutions that had checks and balances to insulate themselves from self-interest and pandering to the public, but in practice this is extremely hard.
As happened before, you formulate half-baked, fuzzy suspicions and intuitions that I have often been holding for many years with a clarity that slaps me almost literally in the face. :-)
I think that one of humankind’s profoundest delusion is that our senses communicate the objective reality to our cognitive system—what we (mainly) hear and see is treated as unconditionally true. We are not aware of the massive amount of interpretation that goes on before we’ve even noticed a ‘fact’.
(And the same - let’s be generous, and call it fashioning rather than distorting - fashioning occurs when we communicate those facts: we believe not only that it is possible to accurately describe facts in words, but also that we effortless do so. And when the other party gets it wrong, it’s them that misinterpret it, not us being sloppy and vague.)
If one thing is remarkable, it’s that, despite all this mad interpreting, we still manage to cooperate and communicate with way more effectiveness than we can reasonably expect.
This is a great analysis and argument for an - in hindsight - obvious superior explanation of our polarized world. 👏
Good analysis. To me, it almost sounds like common sense that academics have forgotten about in the last 20 years.
I'm skeptical of your endorsement of DiResta's assertion that "scholars grounded in fact-based argumentation ... must grapple with this shift [to the social media age]." DiResta made the same argument in her book "Invisible Rulers," but I think that her response to the challenge of social media actually contributed to many of the errors made by anti-misinformation warriors that you have charted on your blog.
When I took a closer look at her virality project (https://www.viralityproject.org/), it was pretty clear to me that her effort to get public health experts to respond rapidly to "vaccine misinformation" on social media only served to further undermine public trust in those experts. The problem from Lippmann's perspective is that you can't respond rapidly without deploying your own stereotype to quickly tell you who is good and who is bad before you have had time to think through the evidence. Thus, the Virality Project encouraged public health experts to deploy a good pro-vaccine stereotype to counteract the bad stereotype deployed by the anti-vaxxers. Unfortunately, new facts kept coming in that exposed the oversimplifications of the pro-vax stereotype. And, when your authority is based on your supposedly objective expertise, having your stereotype undermined by new information is much more damaging to your credibility than it is for random social media influencers.
I propose the opposite approach. Rather than having social scientists jump into the social media fray as DiResta advocates, I think that they should retreat back into their detached ivory towers. Analyze social conflict from the Man from Mars perspective. That's basically what you do here in your Nowak/Floyd analysis. I can't tell if you agree with the right-wing, left-wing stereotype or neither. That allows you to dissect both of them clinically. It's unlikely that the Man from Mars analyses will ever be popular, but it may be helpful just having them out there.
Great article! I too am cautiously optimistic that *thoughtfully developed* AI-driven tech could improve some of these concerns, perhaps even (as you discussed in a previous article) the basic chatbot form factors of contemporary AI (though I think those come with the same splintering of trust and trustworthiness, the ideas/rationalization marketplace just hasn't caught up to that yet). See several of my articles for more, e.g. https://www.oliversourbut.net/p/citations-needed
Excellent piece - insightful and balanced. I fear that AI will lead us further away from shared reality, as the AI providers are in my opinion more likely to take advantage of opportunities to pander to biases rather than to act to counter them (just as has occurred with news media) but I will read your take on this with interest.
We may want institutions that serve as guardians of more objective truth, but the challenge is that there may be no such thing in the ultimate sense, in the sense that all of our interpretations of the world are squeezed into narratives, and narratives are always disputable.
The two things that neutral institutions should guard against, to mitigate the social and political polarization, with the resulting harm to the body politic.
One is exaggeration of how representative (common) an event is. In both the Floyd and the Novak cases, what makes the story compelling to a partisan is the idea that "this is one small example of what's happening everywhere, all the time". In the aftermath of Floyd, for example, white liberals (in surveys) dramatically overestimated the how many black men are killed by cops (real number in the teens annually, estimates by liberals went up into the thousands). Putting every story into numerical context might help.
The second is the magnitude of the role of grievance in the narrative. "THEY are attacking our people" is the unifying motivator for both the right and left wing polarization narratives. The more a story can be told in terms of impersonal causes with unfortunate effects, the less emotionally compelling it is, and the less it drives polarization.
Obama attacked the 1619 project for its historical inaccuracy and its ridiculous and rather forced attempt to discredit capitalism via slavery-as-an-explanation-for-all-bad-thngs.
But ideology trumped realism and the project is now a core subject in US public schools. Just wait for the views of THIS next generation…
As for tribalism, natives' tribalism fluctuates quite dramatically: when a dem president is in charge democrat voters have significant trust in democracy while for reps it’s the opposite. When a rep president takes office the trust in democracy immediately reverses along party preference lines.
In Europe something else is going on: while most immigrants to the US traditionally have been over-educated, those to Europe are almost always under-educated or illiterate (and moving to the most generous welfare states on earth).
Plus, where immigrants to the US typically come from low tribal regions, immigrants to Europe mostly come from North Africa, the Middle East, the Horn of Africa, Pakistan and Afghanistan. The most tribal regions on the planet:
Good piece. AI will help bring epistemic convergence and rigor after social media shattered it, which will be largely positive even as it creates new issues.
What makes you think or say this? Have you actually used AI at all? It's literally trained via reinforcement learning to provide answers that make people feel good, rather than what's truthful. It tells you what you want to hear, and if it doesn't, all it takes is the slightest nudge to get it to shift its position to yours. It's confirmation bias on steroids.
If we look at both AI and social media as tools (tech) which we use to shine light on reality/ourselves, then we will see things more as being "revealed by" than "caused by" them.
(I believe social media and AI can both go under the "largely positive even as it creates new issues" umbrella, even tho we tend to demonize one or the other — or indeed "all tech"!)
That said, I'm not sure what *exactly* the difference between causes and revelations is... 😋
This is the key part Dan
"Instead, the action resides in the competing systems of interpretation through which people select, categorise, frame, connect, explain, and narrate those facts."
It's not different worlds, it's "Interpretive polarisation " - a new term to me.
i'd also argue, the desire for speed of reaction and outrage also short circuits our ability to think clearly and address our own biases of thought
As you said Dan - "Nevertheless, the core narrative is a gross simplification and distortion of a far more complex reality"
Great pièce. I read it telling myself that I still operate a pretty good pseudo reality in comparison with many people haha! But, seriously, I needed to read this as a counterpoint to all the overdetermined commentary from the right on the Novak murder.
You mentioned your students struggle with Lippman's criticism of "lived experiences". It might help them to consider a non-political example that they've probably encountered first hand: The dress.
I'm referring to the dress that some people interpreted to be blue and black, and others interpreted to be white and gold. Most people have "lived experiences" in interpreting the color of things, and so they felt very confident in their ability to interpret the color of this dress. And yet...
Great post, and I mostly agree. I have two aspects to nitpick, though. We are indeed cursed by the complexity of the human world in a way that doesn't allow itself to be reduced to the simple models that humans can handle - I wish we were a tad more like electrons!-. Still, I believe there exists enough truth of the matter even for social issues that could, ideally, be reached and agreed upon given a slightly idealized (but not impossible) scenario in which we had more general access to core facts and their more consistently fitting interpretations, along with a greater openness to consider different interpretations and not engage tribally-emotionally in them. This ain't utopia: to some degree, we already do it all the time (with subjects and topics we don't invest in).
My second nitpick is just my deep distrust of gatekeepers and technocratic solutions. I'd say my cynical support in liberal democracy mostly stems of a negative appreciation of humans: we are mostly self-serving, tribalist animals, but one advantage of liberal-democratic policies is we just put a partial leash on the ruling elites to restrain how much they just look after themselves and their individual and group interests and make it somewhat incentivized to try to maximize and generalize good outcomes for a wider circle. The problem with each and every technocratic class and solution is that they will inevitable bend things towards their self-interest, whether consciously or unconsciously, and will be likely intelligent enough to be able to rationalize their biases. We got this in the Great Awokening, and there's no indication it won't happen again. In theory, we *could* set up incentives for maximally truth-seeking and truth-generating institutions that had checks and balances to insulate themselves from self-interest and pandering to the public, but in practice this is extremely hard.
Music to my ears, Dan!
As happened before, you formulate half-baked, fuzzy suspicions and intuitions that I have often been holding for many years with a clarity that slaps me almost literally in the face. :-)
I think that one of humankind’s profoundest delusion is that our senses communicate the objective reality to our cognitive system—what we (mainly) hear and see is treated as unconditionally true. We are not aware of the massive amount of interpretation that goes on before we’ve even noticed a ‘fact’.
(And the same - let’s be generous, and call it fashioning rather than distorting - fashioning occurs when we communicate those facts: we believe not only that it is possible to accurately describe facts in words, but also that we effortless do so. And when the other party gets it wrong, it’s them that misinterpret it, not us being sloppy and vague.)
If one thing is remarkable, it’s that, despite all this mad interpreting, we still manage to cooperate and communicate with way more effectiveness than we can reasonably expect.
This is a great analysis and argument for an - in hindsight - obvious superior explanation of our polarized world. 👏
Good analysis. To me, it almost sounds like common sense that academics have forgotten about in the last 20 years.
I'm skeptical of your endorsement of DiResta's assertion that "scholars grounded in fact-based argumentation ... must grapple with this shift [to the social media age]." DiResta made the same argument in her book "Invisible Rulers," but I think that her response to the challenge of social media actually contributed to many of the errors made by anti-misinformation warriors that you have charted on your blog.
When I took a closer look at her virality project (https://www.viralityproject.org/), it was pretty clear to me that her effort to get public health experts to respond rapidly to "vaccine misinformation" on social media only served to further undermine public trust in those experts. The problem from Lippmann's perspective is that you can't respond rapidly without deploying your own stereotype to quickly tell you who is good and who is bad before you have had time to think through the evidence. Thus, the Virality Project encouraged public health experts to deploy a good pro-vaccine stereotype to counteract the bad stereotype deployed by the anti-vaxxers. Unfortunately, new facts kept coming in that exposed the oversimplifications of the pro-vax stereotype. And, when your authority is based on your supposedly objective expertise, having your stereotype undermined by new information is much more damaging to your credibility than it is for random social media influencers.
I propose the opposite approach. Rather than having social scientists jump into the social media fray as DiResta advocates, I think that they should retreat back into their detached ivory towers. Analyze social conflict from the Man from Mars perspective. That's basically what you do here in your Nowak/Floyd analysis. I can't tell if you agree with the right-wing, left-wing stereotype or neither. That allows you to dissect both of them clinically. It's unlikely that the Man from Mars analyses will ever be popular, but it may be helpful just having them out there.
Thomas Kuhn — and politics!
Excellent. Thanks for this!
Superb contribution on an important topic. If you are not careful, you may find yourself becoming a moderate. :)
Great article! I too am cautiously optimistic that *thoughtfully developed* AI-driven tech could improve some of these concerns, perhaps even (as you discussed in a previous article) the basic chatbot form factors of contemporary AI (though I think those come with the same splintering of trust and trustworthiness, the ideas/rationalization marketplace just hasn't caught up to that yet). See several of my articles for more, e.g. https://www.oliversourbut.net/p/citations-needed
Excellent piece - insightful and balanced. I fear that AI will lead us further away from shared reality, as the AI providers are in my opinion more likely to take advantage of opportunities to pander to biases rather than to act to counter them (just as has occurred with news media) but I will read your take on this with interest.
This may be the most nuance I've ever seen in a substack piece. Food for thought.
A couple of thoughts.
We may want institutions that serve as guardians of more objective truth, but the challenge is that there may be no such thing in the ultimate sense, in the sense that all of our interpretations of the world are squeezed into narratives, and narratives are always disputable.
The two things that neutral institutions should guard against, to mitigate the social and political polarization, with the resulting harm to the body politic.
One is exaggeration of how representative (common) an event is. In both the Floyd and the Novak cases, what makes the story compelling to a partisan is the idea that "this is one small example of what's happening everywhere, all the time". In the aftermath of Floyd, for example, white liberals (in surveys) dramatically overestimated the how many black men are killed by cops (real number in the teens annually, estimates by liberals went up into the thousands). Putting every story into numerical context might help.
The second is the magnitude of the role of grievance in the narrative. "THEY are attacking our people" is the unifying motivator for both the right and left wing polarization narratives. The more a story can be told in terms of impersonal causes with unfortunate effects, the less emotionally compelling it is, and the less it drives polarization.
Obama attacked the 1619 project for its historical inaccuracy and its ridiculous and rather forced attempt to discredit capitalism via slavery-as-an-explanation-for-all-bad-thngs.
But ideology trumped realism and the project is now a core subject in US public schools. Just wait for the views of THIS next generation…
As for tribalism, natives' tribalism fluctuates quite dramatically: when a dem president is in charge democrat voters have significant trust in democracy while for reps it’s the opposite. When a rep president takes office the trust in democracy immediately reverses along party preference lines.
In Europe something else is going on: while most immigrants to the US traditionally have been over-educated, those to Europe are almost always under-educated or illiterate (and moving to the most generous welfare states on earth).
The Mirage of America's Special Sauce Theory
https://www.philippelemoine.com/p/the-mirage-of-americas-special-sauce
Plus, where immigrants to the US typically come from low tribal regions, immigrants to Europe mostly come from North Africa, the Middle East, the Horn of Africa, Pakistan and Afghanistan. The most tribal regions on the planet:
Individualism and cooperation: I
https://www.notonyourteam.co.uk/p/individualism-and-cooperation-i?hide_intro_popup=true#:~:text=of%20the%20clan-,Across%20the%20globe%2C,-most%20cultures%20are
the problem isn't lack of shared facts it's conflicting interpretations that people weaponize