Thanks for the honest and authentic writing. I am also a frequent user of ChatGPT, but for very specific task, I converse with it regarding things I read about and wanted a discussing partner. I learn alot from AI, and in turn, I give a little part of my psyche to it. I believe this to be a win-win situation.
You raise a valid point though, which is that it doesn’t matter if the claim itself is wrong or right. The goal is to raise confusion, and generative AI is supercharging that confusing, because virtually anything can now be generated using AI, so convincingly, that we can’t tell the difference anymore.
I have been doing a deep dive into that, since I wrote this. (I vaguely knew that the concept existed, though I don't know how much it was ever at this scale.) It's been very interesting!
It does remain an open question whether hallucinations will ever really go away, or if they are built into the architecture and will actually get much worse. (I guess time will tell!)
I’m interested in knowing more about why the size of the crowd is important. To whom? This interest derives from a critical perspective on Trump, who has a history of distortion of crowd sizes. One assumption I hold, which I’ve examined, is that Trump welcomes bringing public attention to bear on his crowd. My critical thinking suggests that by discussing the crowd, we make him important.
The question of whether the photo is doctored is empirical, answered by scientific observation. Logical analysis based on formal rules of deduction can in theory produce certainty. Critical thinking can produce compelling evidence but rarely certainty. If the photo without a crowd is real or fake, it has no bearing on the real issue. Who cares about crowd size?
I used to teach a course in critical thinking at Sacramento State. It was really a lot of fun. The course has a theoretical framework of critical thinking that covers implementation across all 26 campuses of the CSU system.
Critical thinking is a broad skill encompassing analysis, evaluation, and reflection. It includes logical reasoning but also other cognitive strategies including questioning assumptions and considering multiple perspectives. It focuses on practical problem-solving and decision-making while it addresses real-world complexity and ambiguity
While logical reasoning is a key component of critical thinking, critical thinking extends beyond formal logic to include a wider range of analytical and evaluative skills. The type of thinking required to determine the authenticity of the photo is an example of empirical research, not critical research.
I have a long standing concern about conflating critical thinking with making and breaking arguments, certainly an application of this genre of cognition. I think the problem is when we assume that writing is critical thinking and critical thinking is writing. Though there is overlap there are huge differences
I think that I am sometimes prone to conflate critical thinking and empirical analysis, since I do see the two as fundamentally connected.
I think, more and more, this kind of analysis will get closer to "critical thinking" and further away from "scientific observation," as deep fakes continue to improve. I just have a hard time seeing the pursuit of certainty remaining as (relatively) neat and pat as it is right now.
Looking at a photo may be more akin to mixing logic, empirical analysis, and reflection in the future than it does at the current moment.
I tend to be protective of situated definitions of critical thinking and am loathe to give my own sense of it. After teaching at the university for seventeen years, working in the end as the university assessment coordinator, my job brought me in contact with a Baskin and Robbins of thought about the more amorphous learning outcomes like critical thinking, creative thinking, even information literacy.
After I became involved in the ACLU's VALUE Project, I found a useful set of tools to bring faculty together around a shared understanding of constructs like "critical thinking." Everybody thought what they were doing IS critical thinking. Designing a bridge IS critical thinking. Writing a paper IS critical thinking. In the end critical thinking becomes whatever I think it is that I'm doing.
I concluded that critical thinking is so ambiguous because it is so significant and ubiquitous. This forced me to conclude that CT can be taught only in discourse genres, which I still lean toward. But the VALUE rubric for CT distilled a shared understanding which faculty across thousands of colleges now embrace in name at least.
For me, CT is multi-leveled from situated to disciplinary to existential. Situated probably is closely related to analysis, disciplinary to synthesis, existential to evaluation. At least that's one way Ive made sense of it. Blooms taxonomy is theoretically clunky, but Ive found it a decent heuristic.
This Baskin Robbins aspect drove me on a quest. What in the hell IS its? I found a body of research looking at assessment of CT that turned out to be useful. Since what we call CT and think it is collectively matters more than some scientifically precise definition in an existential mindset, assessment was a natural. The lynchpin was this empirical study from U Washington (one of the Washingtons) where they used an ACLU-like construct of CT and developed an assessment checklist. They assessed evidence of CT and evidence of writing using the CT checklist as one factor and a writing rubric as another. They measured different constructs in a big way. So we can distinguish CT from writing and therefore can think about their instruction as related but separate.
Your use of the term "fundamentally connected" re CT and empirical analysis (any analysis really, is spot on. They really don't need to be separate except perhaps for pedagogical purposes... what do you think Jason?
I’m not sure this example requires critical thinking in isolation. I agree with the comment about the real question. Clearly, critical thinking is the cognition applied to determine who gives a healthy shit. What does crowd side has to do with anything? Once that bridge is crossed it becomes unnecessary to go to a fine grained analysis. The work of finding out about tail fins and serial numbers would not have been necessary if you had been thinking critically to begin with. You engaged not in critical thinking but in forensic examination of evidence to determine if the crime of fraud had occurred. You can identify a small role for critical thinking in the forensic analysis, though my hunch is that more creative thinking and ethical reasoning would be involved. So, no need to worry. The only way critical thinking is going to fail in the war against AI is if we don’t use it at the appropriate time.
I thinkyour point about selecting what we critically think about is really key. (Though I'll also say that many people do find the detail of the crowds very important. I try not to evaluate the object that is being analyzed, as much as I can.)
And yes, that is certainly one way this could go. And I hope you're right!
So much of it will come down to whether we're underestimating the textpocalypse, how much that textpocalypse will unsettle our children's epistemologu (which, where we like it or not, is largely virtual), and the erosion of media news outlets.
I fear we may be. But maybe I'm having a pessimistic day.
I'm going to say the definition of critical thinking is too fragile here. It misses the actual piece parts of what Thinking + Critique really is. I've seen too many people discount Critical Thinking because they've never critically thought about critical thinking.
I enjoyed reading your article, though I'm not sure it pushes against my idea again all.
We seem to be lock-step on most things, when it comes to the process of critical thinking.
I think our main part of departure (feel free to disagree with me about this, obviously) is whether critical thinking is a reliable way to subvert our own biases. I'd love to be optimistic about this.
Taking on multiple perspectives is certainly a step in the right direction, and is part of my understanding of the concept too, in my article.
I think when things get tricky is when we are doing it in isolation, especially when those multiple perspectives emanate (to a degree) from our own biases.
That's why I mean on the social aspect a little more. I think things changed when we work together in our critical thinking. But, and this was the engine behind the whole article, that can be a big lift.
And I certainly don't discount the power of critical thinking. In fact, I'm bullish on critical thinking and my whole article was about how powerful it is!
Anyway, I am sorry that is a long note. Thank you for the conversation and feedback!
What I disagreed with, and maybe I put too much emphasis on this, was 1. whether it was scalable, and 2. whether we could use it to adress our own biases.
I think the framework answers the first part. It is scalable in that sense. I think the critique element takes the second on biases. Done right, critique should address our own biases first.
Now to your point, the second part is where most people fail making it had to scale. 😃
That's a good point! Thank you for clarifying that!
Maybe a better way for me to put it is that it will require a specific framework of critical thinking.
Part of me, too, gets stuck because it's hard to tell how big the "textpocalypse" will be, and how far reaching.
One of my fears is that it will produce a truly negative epistemology. My students really depend on the Internet to find "truth." That often becomes their foundation for critical thinking. They use it as a starting point to think critically, and (fingers crossed) use the critical thinking to then question that foundation. (I think we're very similar in that idea, too.)
But what happens if the textpocalypse is so big and so vast that they lose almost all of that foundation?
What would a model of critical thinking without a solid foundation like that look like?
It seems like then (and I was thinking about this with Terry's comment below) it would make it hard to select what we think about critically. .
Reminds me of the tech available now to real-time translate a speaker in their voice to another language and modify the image so their face matches the translated words. Potentially useful, of course. But the same tech could easily be used to make the speaker say something different.
I agree, we can't all spend all our time trying to verify everything. Can we use AI to take the burden on? But then, how do we know the AI service we're using isn't compromised by some interest?
It's enough to make you pack up and move to a cabin in the woods and say, "to hell with it!"
Really interesting analysis, Jason. But I do wonder whether either of those analyses rebut the central argument that was being made about the photo. It wasn't the plane or the sky that were in doubt, but rather the size of the crowd. You could imagine a real photo of the real plane on the real day, but with an AI generated crowd. So do either of these analyses show the evidence that you would want in order to prove the veracity of the image in that respect?
That's a great question! Thank you! It actually allows me to make an important clarification, about my point.
I'm actually not all that interested in proving the veracity of the image. I think we can definitely go back and forth, about the best process and whether we should be focusing on the image in general or only specific parts of the image.
I'm more interested in the need for an individual to have a need for a process at all. I think we're moving into a world where individuals have a ton of interpretive burden put on them.
So, even if tomorrow we found out that former President Trump was right, I think this idea was still hold value.
I definitely think we are asking too much of critical thinking. It's crazy to ask consumers to go through this much trouble, especially when it comes to videos and images which hit the back of our cerebral cortex before we even have the chance to realize what is happening! (unlike written words).
My solution is over-the-top (like always). I say we start a grassroots campaign to change laws at the federal level that put restrictions on voice cloning and deepfake technologies, with liability placed on the tech companies that provide the services themselves. In my opinion, they should not be able to offer technology that allows us to easily clone someone else's voice or face. I feel like that makes sense, but...
I know I know....it doesn't stand a chance. In fact, I recently had an education policymaker call me a "kook" for bringing up the idea for policy-level age restrictions and training requirements for teenagers to use AI systems. Unfortunately, when it comes to policy (in my opinion), the ideas that make the most sense have the least chance of passing...
What say you, Senor Gulya? Should there be laws that restrict the use of deepfake technology, with liability placed on the tech companies themselves? Or am I a kook?
I also tend to think about creating policies like this, even if some of them may be DOA.
In the end, I have a hard time conceptualizing moving forward without reliable authorities on the topic. In the end, my own answers to the questions about Air Force Two depended on specific sources. I felt like I could rely on CNN, Washington Post, etc.
What happens if I stop relying on certain sources, and I have to figure things out in just an amorphous sea of information and data?
I don't know. I legitimately don't know.
To push against that, it does seem like the kind of policy you're suggesting makes sense. And get it may be all so impractical.
Thanks for the honest and authentic writing. I am also a frequent user of ChatGPT, but for very specific task, I converse with it regarding things I read about and wanted a discussing partner. I learn alot from AI, and in turn, I give a little part of my psyche to it. I believe this to be a win-win situation.
Fun fact: this phenomenon is also known as liar's divided, coined by Robert Chesney and Danielle Citron in a paper about deep fakes in 2018. Here’s a link to that: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3213954
You raise a valid point though, which is that it doesn’t matter if the claim itself is wrong or right. The goal is to raise confusion, and generative AI is supercharging that confusing, because virtually anything can now be generated using AI, so convincingly, that we can’t tell the difference anymore.
I wrote a short piece on this myself about several pictures of TEDx speakers that went viral last week, that turned out to be AI generated as well: https://jurgengravestein.substack.com/p/welcome-to-hyperreality
I have been doing a deep dive into that, since I wrote this. (I vaguely knew that the concept existed, though I don't know how much it was ever at this scale.) It's been very interesting!
Good post. Your main points are easily understood. Don't overload your brain too much on generative AI - it everywhere and getting better every week.
That's fair!
It does remain an open question whether hallucinations will ever really go away, or if they are built into the architecture and will actually get much worse. (I guess time will tell!)
I’m interested in knowing more about why the size of the crowd is important. To whom? This interest derives from a critical perspective on Trump, who has a history of distortion of crowd sizes. One assumption I hold, which I’ve examined, is that Trump welcomes bringing public attention to bear on his crowd. My critical thinking suggests that by discussing the crowd, we make him important.
The question of whether the photo is doctored is empirical, answered by scientific observation. Logical analysis based on formal rules of deduction can in theory produce certainty. Critical thinking can produce compelling evidence but rarely certainty. If the photo without a crowd is real or fake, it has no bearing on the real issue. Who cares about crowd size?
I used to teach a course in critical thinking at Sacramento State. It was really a lot of fun. The course has a theoretical framework of critical thinking that covers implementation across all 26 campuses of the CSU system.
Critical thinking is a broad skill encompassing analysis, evaluation, and reflection. It includes logical reasoning but also other cognitive strategies including questioning assumptions and considering multiple perspectives. It focuses on practical problem-solving and decision-making while it addresses real-world complexity and ambiguity
While logical reasoning is a key component of critical thinking, critical thinking extends beyond formal logic to include a wider range of analytical and evaluative skills. The type of thinking required to determine the authenticity of the photo is an example of empirical research, not critical research.
I have a long standing concern about conflating critical thinking with making and breaking arguments, certainly an application of this genre of cognition. I think the problem is when we assume that writing is critical thinking and critical thinking is writing. Though there is overlap there are huge differences
That is a great reminder, Terry! Thank you!
I think that I am sometimes prone to conflate critical thinking and empirical analysis, since I do see the two as fundamentally connected.
I think, more and more, this kind of analysis will get closer to "critical thinking" and further away from "scientific observation," as deep fakes continue to improve. I just have a hard time seeing the pursuit of certainty remaining as (relatively) neat and pat as it is right now.
Looking at a photo may be more akin to mixing logic, empirical analysis, and reflection in the future than it does at the current moment.
Maybe I'm wrong...
I tend to be protective of situated definitions of critical thinking and am loathe to give my own sense of it. After teaching at the university for seventeen years, working in the end as the university assessment coordinator, my job brought me in contact with a Baskin and Robbins of thought about the more amorphous learning outcomes like critical thinking, creative thinking, even information literacy.
After I became involved in the ACLU's VALUE Project, I found a useful set of tools to bring faculty together around a shared understanding of constructs like "critical thinking." Everybody thought what they were doing IS critical thinking. Designing a bridge IS critical thinking. Writing a paper IS critical thinking. In the end critical thinking becomes whatever I think it is that I'm doing.
I concluded that critical thinking is so ambiguous because it is so significant and ubiquitous. This forced me to conclude that CT can be taught only in discourse genres, which I still lean toward. But the VALUE rubric for CT distilled a shared understanding which faculty across thousands of colleges now embrace in name at least.
For me, CT is multi-leveled from situated to disciplinary to existential. Situated probably is closely related to analysis, disciplinary to synthesis, existential to evaluation. At least that's one way Ive made sense of it. Blooms taxonomy is theoretically clunky, but Ive found it a decent heuristic.
This Baskin Robbins aspect drove me on a quest. What in the hell IS its? I found a body of research looking at assessment of CT that turned out to be useful. Since what we call CT and think it is collectively matters more than some scientifically precise definition in an existential mindset, assessment was a natural. The lynchpin was this empirical study from U Washington (one of the Washingtons) where they used an ACLU-like construct of CT and developed an assessment checklist. They assessed evidence of CT and evidence of writing using the CT checklist as one factor and a writing rubric as another. They measured different constructs in a big way. So we can distinguish CT from writing and therefore can think about their instruction as related but separate.
Your use of the term "fundamentally connected" re CT and empirical analysis (any analysis really, is spot on. They really don't need to be separate except perhaps for pedagogical purposes... what do you think Jason?
I really appreciate that!
I think that you're right: in higher Ed, many of us really push CT, but only have a basic understanding of what it is.
I'll on a quest for greater clarify on that, myself! I'll be diving into it in earnest, this semester!
Write about it! I'd love to read about your journey with your students.
Absolutely! Writing helps me process things, anyway!
I’m not sure this example requires critical thinking in isolation. I agree with the comment about the real question. Clearly, critical thinking is the cognition applied to determine who gives a healthy shit. What does crowd side has to do with anything? Once that bridge is crossed it becomes unnecessary to go to a fine grained analysis. The work of finding out about tail fins and serial numbers would not have been necessary if you had been thinking critically to begin with. You engaged not in critical thinking but in forensic examination of evidence to determine if the crime of fraud had occurred. You can identify a small role for critical thinking in the forensic analysis, though my hunch is that more creative thinking and ethical reasoning would be involved. So, no need to worry. The only way critical thinking is going to fail in the war against AI is if we don’t use it at the appropriate time.
I thinkyour point about selecting what we critically think about is really key. (Though I'll also say that many people do find the detail of the crowds very important. I try not to evaluate the object that is being analyzed, as much as I can.)
And yes, that is certainly one way this could go. And I hope you're right!
So much of it will come down to whether we're underestimating the textpocalypse, how much that textpocalypse will unsettle our children's epistemologu (which, where we like it or not, is largely virtual), and the erosion of media news outlets.
I fear we may be. But maybe I'm having a pessimistic day.
I'm going to say the definition of critical thinking is too fragile here. It misses the actual piece parts of what Thinking + Critique really is. I've seen too many people discount Critical Thinking because they've never critically thought about critical thinking.
However, it can be defined as a simple framework.
1. Knowlege Gathering (Thinking Step 1)
2. Logical Formulation (Thinking Step 2)
3. Application of Critique (Step 3)
More Here: https://www.polymathicbeing.com/p/do-you-really-think-critically
Thank you for that, Michael!
I enjoyed reading your article, though I'm not sure it pushes against my idea again all.
We seem to be lock-step on most things, when it comes to the process of critical thinking.
I think our main part of departure (feel free to disagree with me about this, obviously) is whether critical thinking is a reliable way to subvert our own biases. I'd love to be optimistic about this.
Taking on multiple perspectives is certainly a step in the right direction, and is part of my understanding of the concept too, in my article.
I think when things get tricky is when we are doing it in isolation, especially when those multiple perspectives emanate (to a degree) from our own biases.
That's why I mean on the social aspect a little more. I think things changed when we work together in our critical thinking. But, and this was the engine behind the whole article, that can be a big lift.
And I certainly don't discount the power of critical thinking. In fact, I'm bullish on critical thinking and my whole article was about how powerful it is!
Anyway, I am sorry that is a long note. Thank you for the conversation and feedback!
What I disagreed with, and maybe I put too much emphasis on this, was 1. whether it was scalable, and 2. whether we could use it to adress our own biases.
I think the framework answers the first part. It is scalable in that sense. I think the critique element takes the second on biases. Done right, critique should address our own biases first.
Now to your point, the second part is where most people fail making it had to scale. 😃
That's a good point! Thank you for clarifying that!
Maybe a better way for me to put it is that it will require a specific framework of critical thinking.
Part of me, too, gets stuck because it's hard to tell how big the "textpocalypse" will be, and how far reaching.
One of my fears is that it will produce a truly negative epistemology. My students really depend on the Internet to find "truth." That often becomes their foundation for critical thinking. They use it as a starting point to think critically, and (fingers crossed) use the critical thinking to then question that foundation. (I think we're very similar in that idea, too.)
But what happens if the textpocalypse is so big and so vast that they lose almost all of that foundation?
What would a model of critical thinking without a solid foundation like that look like?
It seems like then (and I was thinking about this with Terry's comment below) it would make it hard to select what we think about critically. .
Reminds me of the tech available now to real-time translate a speaker in their voice to another language and modify the image so their face matches the translated words. Potentially useful, of course. But the same tech could easily be used to make the speaker say something different.
I agree, we can't all spend all our time trying to verify everything. Can we use AI to take the burden on? But then, how do we know the AI service we're using isn't compromised by some interest?
It's enough to make you pack up and move to a cabin in the woods and say, "to hell with it!"
That is very true. I think it's so tempting to offload more. And then you're right, that needs to be verified.
I think he textpocalypse and the rabbit hole of verification is going to be a bigger deal than we sometimes think.
Really interesting analysis, Jason. But I do wonder whether either of those analyses rebut the central argument that was being made about the photo. It wasn't the plane or the sky that were in doubt, but rather the size of the crowd. You could imagine a real photo of the real plane on the real day, but with an AI generated crowd. So do either of these analyses show the evidence that you would want in order to prove the veracity of the image in that respect?
That's a great question! Thank you! It actually allows me to make an important clarification, about my point.
I'm actually not all that interested in proving the veracity of the image. I think we can definitely go back and forth, about the best process and whether we should be focusing on the image in general or only specific parts of the image.
I'm more interested in the need for an individual to have a need for a process at all. I think we're moving into a world where individuals have a ton of interpretive burden put on them.
So, even if tomorrow we found out that former President Trump was right, I think this idea was still hold value.
That's my focus, anyway.
Great post - so many thoughts!
I definitely think we are asking too much of critical thinking. It's crazy to ask consumers to go through this much trouble, especially when it comes to videos and images which hit the back of our cerebral cortex before we even have the chance to realize what is happening! (unlike written words).
My solution is over-the-top (like always). I say we start a grassroots campaign to change laws at the federal level that put restrictions on voice cloning and deepfake technologies, with liability placed on the tech companies that provide the services themselves. In my opinion, they should not be able to offer technology that allows us to easily clone someone else's voice or face. I feel like that makes sense, but...
I know I know....it doesn't stand a chance. In fact, I recently had an education policymaker call me a "kook" for bringing up the idea for policy-level age restrictions and training requirements for teenagers to use AI systems. Unfortunately, when it comes to policy (in my opinion), the ideas that make the most sense have the least chance of passing...
What say you, Senor Gulya? Should there be laws that restrict the use of deepfake technology, with liability placed on the tech companies themselves? Or am I a kook?
These are great ideas.
I also tend to think about creating policies like this, even if some of them may be DOA.
In the end, I have a hard time conceptualizing moving forward without reliable authorities on the topic. In the end, my own answers to the questions about Air Force Two depended on specific sources. I felt like I could rely on CNN, Washington Post, etc.
What happens if I stop relying on certain sources, and I have to figure things out in just an amorphous sea of information and data?
I don't know. I legitimately don't know.
To push against that, it does seem like the kind of policy you're suggesting makes sense. And get it may be all so impractical.
Sorry. Apparently, I have no answers today.
No worries,. it's the nature of these discussions. I appreciate that you think it makes sense!