WEBVTT 00:00:00.000 --> 00:00:02.839 Imagine for a second that you have spent your 00:00:02.839 --> 00:00:06.080 entire professional life begging for one specific 00:00:06.080 --> 00:00:08.539 thing. Like your whole career. Right. I mean, 00:00:08.759 --> 00:00:12.000 for decades, you have been pounding the table, 00:00:12.240 --> 00:00:15.859 you know, altering your entire workflow, restructuring 00:00:15.859 --> 00:00:18.760 a whole industry, all just to get this one highly 00:00:18.760 --> 00:00:20.780 specific outcome. Yeah, you're just totally obsessed 00:00:20.780 --> 00:00:24.100 with it. Exactly. And then one day someone invents 00:00:24.100 --> 00:00:27.039 a machine that just, it just gives it to you 00:00:27.039 --> 00:00:30.469 instantly. perfectly in endless quantities. You'd 00:00:30.469 --> 00:00:32.549 think you'd be throwing a parade. Right. It sounds 00:00:32.549 --> 00:00:35.109 like an absolute dream come true. Or a classic 00:00:35.109 --> 00:00:37.729 monkey's paw situation. Oh, yeah. Because, you 00:00:37.729 --> 00:00:40.649 know, getting exactly what you asked for is quite 00:00:40.649 --> 00:00:43.409 often the quickest way to realize you were asking 00:00:43.409 --> 00:00:45.509 for the completely wrong thing all along. And 00:00:45.509 --> 00:00:48.189 that is exactly the brutal realization that's 00:00:48.189 --> 00:00:51.380 currently just... sweeping through our education 00:00:51.380 --> 00:00:53.560 system, which is what we are getting into today. 00:00:53.659 --> 00:00:56.219 It's a huge shift. It really is. Welcome to the 00:00:56.219 --> 00:00:58.740 Deep Dive, everyone. We are jumping right into 00:00:58.740 --> 00:01:03.000 a fascinating piece from April 2026 by Gary Ackerman. 00:01:03.159 --> 00:01:06.340 Over at hackscience .education. Yeah, yes, exactly. 00:01:06.840 --> 00:01:10.079 The post is titled, Generative AI. We are getting 00:01:10.079 --> 00:01:13.400 what we asked for. And it outlines this massive. 00:01:14.340 --> 00:01:16.480 unintentional irony playing out in classrooms 00:01:16.480 --> 00:01:18.780 right now. Such a great framing of the issue. 00:01:18.980 --> 00:01:21.680 It is because teachers and administrators, they've 00:01:21.680 --> 00:01:24.200 spent decades demanding data. You know, give 00:01:24.200 --> 00:01:26.379 us the measurable data. Show us the metrics. 00:01:26.540 --> 00:01:29.200 Right. But now that generative AI is basically 00:01:29.200 --> 00:01:32.400 handing them pristine, flawless data on a silver 00:01:32.400 --> 00:01:35.079 platter, they are fundamentally dissatisfied. 00:01:35.340 --> 00:01:38.069 They hate it. They really do. Okay, let's unpack 00:01:38.069 --> 00:01:40.829 this. Because to understand why educators are 00:01:40.829 --> 00:01:42.909 suddenly dreading the very thing they demanded, 00:01:43.450 --> 00:01:46.549 we have to, um... we have to look at how our 00:01:46.549 --> 00:01:49.010 brains actually work. That is the core of Ackerman's 00:01:49.010 --> 00:01:51.650 argument right there. He grounds this whole crisis 00:01:51.650 --> 00:01:54.189 in the actual mechanics of human thought. So 00:01:54.189 --> 00:01:56.950 psychology, basically. Right. Specifically, he 00:01:56.950 --> 00:01:59.310 references Daniel Kahneman's framework from the 00:01:59.310 --> 00:02:02.170 2011 book, Thinking Fast and Slow. Yeah, a classic. 00:02:02.230 --> 00:02:04.810 It really is. And to understand the trap we've 00:02:04.810 --> 00:02:08.150 fallen into with AI, we have to understand Kahneman's 00:02:08.150 --> 00:02:10.569 two systems. Right. So for those who might not 00:02:10.569 --> 00:02:13.150 have read it, Kahneman argued that our brains 00:02:13.150 --> 00:02:16.150 basically operate using in two distinct systems 00:02:16.150 --> 00:02:18.550 of thought. System one and system two. Exactly, 00:02:18.710 --> 00:02:22.229 so system one is your fast, intuitive, automatic 00:02:22.229 --> 00:02:26.569 thinking. It's the trust your gut reflects. The 00:02:26.569 --> 00:02:28.669 instinct. Yeah, like if you're walking down the 00:02:28.669 --> 00:02:31.069 sidewalk and a bicycle suddenly swerves toward 00:02:31.069 --> 00:02:34.210 you, you don't stop to do a bunch of like physics 00:02:34.210 --> 00:02:36.389 calculations about mass and velocity. Or you'd 00:02:36.389 --> 00:02:38.610 get hit. Exactly, you just jump out of the way. 00:02:38.889 --> 00:02:42.550 That is system one keeping you alive, it's instantaneous, 00:02:42.550 --> 00:02:45.650 and it feels practically effortless. Effortless 00:02:45.650 --> 00:02:47.969 is the key word there. It burns almost no cognitive 00:02:47.969 --> 00:02:50.349 calories. Right. But then we have system two, 00:02:50.449 --> 00:02:53.189 which Ackerman characterizes as the mandate to 00:02:53.189 --> 00:02:56.889 trust your brain. OK, so gut versus brain. Precisely. 00:02:57.069 --> 00:02:59.830 System two is deliberative cognition. It is incredibly 00:02:59.830 --> 00:03:02.830 slow. It requires intense focus. And crucially, 00:03:03.349 --> 00:03:05.909 it is biologically exhausting. Oh, man. Yeah. 00:03:06.009 --> 00:03:07.449 It's that feeling you get when you're, I don't 00:03:07.449 --> 00:03:09.469 know, sitting down to calculate a complicated 00:03:09.469 --> 00:03:13.400 monthly budget. Or doing your taxes. Yes. or 00:03:13.400 --> 00:03:15.659 trying to conjugate verbs in a new language you're 00:03:15.659 --> 00:03:18.800 learning. You can literally feel your gears grinding. 00:03:19.159 --> 00:03:22.580 You can. It is actual physical work for your 00:03:22.580 --> 00:03:25.240 body. Because you're trying so hard. Right. Because 00:03:25.240 --> 00:03:28.120 system two creates friction. When you are grappling 00:03:28.120 --> 00:03:30.740 with a complex concept, your brain is actively 00:03:30.740 --> 00:03:34.039 forging new neural pathways. Wow. OK. Yeah. That 00:03:34.039 --> 00:03:37.020 friction, that slowness, is the literal mechanism 00:03:37.020 --> 00:03:39.659 of learning. You are forcing your brain to reorganize 00:03:39.659 --> 00:03:42.080 itself. OK. So we have system one trusting the 00:03:42.080 --> 00:03:45.000 gut and system two trusting the deliberative 00:03:45.000 --> 00:03:48.639 brain. Yep. But Ackerman's post introduces a 00:03:48.639 --> 00:03:50.860 completely new wrinkle into this framework. Right. 00:03:50.919 --> 00:03:53.219 And this is where he brings in that 2026 pre 00:03:53.219 --> 00:03:55.240 -print paper, right? Yeah. By Shaw and Nave. 00:03:55.360 --> 00:03:57.180 Shaw and Nave. And the title of their paper really 00:03:57.180 --> 00:03:58.439 tells you where this whole thing is heading. 00:03:58.560 --> 00:04:01.599 Let me find it. Here it is. Thinking fast, slow, 00:04:01.680 --> 00:04:05.110 and artificial. How AI is reshaping human reasoning 00:04:05.110 --> 00:04:07.830 and the rise of cognitive surrender. Cognitive 00:04:07.830 --> 00:04:10.729 surrender. That phrase just hits hard. It really 00:04:10.729 --> 00:04:12.990 does. So Sean and Ava are proposing that because 00:04:12.990 --> 00:04:15.289 of modern technology, we now have to account 00:04:15.289 --> 00:04:18.129 for a system three. An artificial system three, 00:04:19.790 --> 00:04:21.589 which they summarize perfectly with the phrase 00:04:21.589 --> 00:04:24.529 trust the data. Trust the data. Right. If system 00:04:24.529 --> 00:04:27.949 one is trusting your gut reflex and system two 00:04:27.949 --> 00:04:31.180 is trusting your slow, deliberative brain, System 00:04:31.180 --> 00:04:34.000 3 is trusting an external artificial system. 00:04:34.000 --> 00:04:36.199 Okay, I see. It's looking at the data the machine 00:04:36.199 --> 00:04:39.420 provides and just accepting it, completely bypassing 00:04:39.420 --> 00:04:41.839 your own internal cognitive processes. That makes 00:04:41.839 --> 00:04:46.449 me think of navigation. like using a GPS. Oh, 00:04:46.550 --> 00:04:48.709 that's a great example. Right. So system one 00:04:48.709 --> 00:04:50.829 is me jumping out of the way of that moving bike. 00:04:51.050 --> 00:04:53.110 Yep. System two is me sitting down with a big 00:04:53.110 --> 00:04:55.910 paper map, analyzing the topography, and like 00:04:55.910 --> 00:04:58.370 carefully mapping out a route to avoid busy roads. 00:04:58.709 --> 00:05:01.089 Lots of cognitive friction there. Exactly. System 00:05:01.089 --> 00:05:03.250 three then would be me just blindly following 00:05:03.250 --> 00:05:06.029 the blue line on my smartphone's GPS, even if 00:05:06.029 --> 00:05:08.089 it is actively routing me right into a lake. 00:05:08.360 --> 00:05:11.000 Which happens. Right. Because I'm not navigating 00:05:11.000 --> 00:05:13.100 anymore. I'm just trusting the data in the machine 00:05:13.100 --> 00:05:15.699 handed me. That is a perfect visualization of 00:05:15.699 --> 00:05:18.300 the mechanics at play here. You've surrendered 00:05:18.300 --> 00:05:20.899 the rote finding entirely. I've cognitively surrendered 00:05:20.899 --> 00:05:23.639 to the phone. Exactly. The artificial system 00:05:23.639 --> 00:05:26.910 did the thinking, and you just... consume the 00:05:26.910 --> 00:05:28.949 output. But wait, hold on though. Hasn't trusting 00:05:28.949 --> 00:05:31.490 the data always been considered like the ultimate 00:05:31.490 --> 00:05:33.529 objective good in our society? Oh, absolutely. 00:05:33.689 --> 00:05:36.269 I mean, we are constantly told in business and 00:05:36.269 --> 00:05:38.949 science, just in everyday life, to put aside 00:05:38.949 --> 00:05:41.689 our biases, stop relying on flawed gut feelings, 00:05:41.810 --> 00:05:44.029 and look at the hard data. Numbers don't lie. 00:05:44.290 --> 00:05:46.670 Right. It sounds like a system three approach 00:05:46.670 --> 00:05:49.329 should be the absolute pinnacle of logic and 00:05:49.329 --> 00:05:52.970 reason. Why is Shaw suddenly framing trust the 00:05:52.970 --> 00:05:55.959 data as a negative thing? What's fascinating 00:05:55.959 --> 00:05:59.180 here is how Shaw and Nave separate the concept 00:05:59.180 --> 00:06:02.420 of data from the process of cognition. The mandate 00:06:02.420 --> 00:06:05.160 to trust the data does sound entirely objective 00:06:05.160 --> 00:06:07.519 and scientifically rigorous. Yeah, it does. But 00:06:07.519 --> 00:06:10.480 in practice, just as the urge to trust your gut 00:06:10.480 --> 00:06:13.660 sways us toward lazy system one intuition, the 00:06:13.660 --> 00:06:16.420 modern mandate to trust the data sways us toward 00:06:16.420 --> 00:06:19.680 system three outsourcing. And here is the crucial 00:06:19.680 --> 00:06:22.480 mechanism that makes it dangerous. It completely 00:06:22.480 --> 00:06:25.579 bypasses human deliberative cognition. It entirely 00:06:25.579 --> 00:06:27.600 bypasses system two. Oh, I see. So you get the 00:06:27.600 --> 00:06:30.319 result without doing the math. Precisely. When 00:06:30.319 --> 00:06:33.560 you just accept the artificial output, the data, 00:06:34.639 --> 00:06:38.220 you aren't actually doing that slow, effortful, 00:06:38.620 --> 00:06:40.879 friction -filled thinking yourself. You're just 00:06:40.879 --> 00:06:43.560 taking the answer. Right. You are outsourcing 00:06:43.560 --> 00:06:45.240 the cognitive load. You aren't building those 00:06:45.240 --> 00:06:48.779 neural pathways. Wow. And this shift toward system 00:06:48.779 --> 00:06:52.139 three isn't just some abstract psychological 00:06:52.139 --> 00:06:54.740 theory happening in a vacuum. It has collided 00:06:54.740 --> 00:06:57.579 head on with a decades long obsession in our 00:06:57.579 --> 00:07:00.680 schools. Which brings us to Ackerman's slightly 00:07:00.680 --> 00:07:03.259 snarky premise in his post. He's got a point, 00:07:03.379 --> 00:07:05.759 though. He really does. He points out that for 00:07:05.759 --> 00:07:08.759 decades in education, the dominant unquestionable 00:07:08.759 --> 00:07:11.980 mantra from the top down has been trust the data. 00:07:12.079 --> 00:07:14.100 If we connect this to the bigger picture, the 00:07:14.100 --> 00:07:16.339 modern education system has been operating under 00:07:16.339 --> 00:07:19.660 an intense systemic pressure to quantify absolutely 00:07:19.660 --> 00:07:21.680 everything. Like an assembly line. It really 00:07:21.680 --> 00:07:23.899 is a highly industrialized approach to human 00:07:23.899 --> 00:07:25.560 development. Let's look at the reality of that 00:07:25.560 --> 00:07:27.360 for a second. Think about it from the perspective 00:07:27.360 --> 00:07:30.319 of a school administrator, or a district superintendent, 00:07:30.560 --> 00:07:32.620 or even just a classroom teacher over the last 00:07:32.620 --> 00:07:34.980 20 or 30 years. They're under so much pressure. 00:07:35.279 --> 00:07:37.040 Insane pressure. You can't just look at a school 00:07:37.040 --> 00:07:38.879 board and say, trust me, the kids are learning. 00:07:39.000 --> 00:07:40.459 No, they'd laugh you out of the room. Right. 00:07:40.459 --> 00:07:43.759 You have to prove it. Every single decision made 00:07:43.759 --> 00:07:46.459 in a classroom, every piece of evidence used 00:07:46.459 --> 00:07:48.819 to draw a conclusion about a school's funding, 00:07:49.319 --> 00:07:52.139 every solitary item recorded in a teacher's grade 00:07:52.139 --> 00:07:56.160 book, it all had to be grounded in data. Measurable, 00:07:56.519 --> 00:07:59.180 chartable, graphable data. Exactly. And this 00:07:59.180 --> 00:08:01.379 is where the conflict between data obsession 00:08:01.379 --> 00:08:04.740 and genuine learning really takes root. How so? 00:08:04.839 --> 00:08:08.220 Because true learning, that system to neuroplasticity, 00:08:08.399 --> 00:08:11.449 that internal cognitive friction... is practically 00:08:11.449 --> 00:08:14.370 invisible. Oh, wow. Yeah. You cannot easily log 00:08:14.370 --> 00:08:16.709 a student's internal struggle into a district 00:08:16.709 --> 00:08:18.810 -wide spreadsheet. You can't graph confusion 00:08:18.810 --> 00:08:21.209 turning into understanding. Exactly. So over 00:08:21.209 --> 00:08:23.829 time, the system became completely fixated on 00:08:23.829 --> 00:08:26.069 the measurable end product, because that was 00:08:26.069 --> 00:08:28.189 the only thing you could see. The proxies for 00:08:28.189 --> 00:08:31.250 learning became the goal itself. Yes. The worksheet, 00:08:31.689 --> 00:08:34.409 the five paragraph essay, the multiple choice 00:08:34.409 --> 00:08:37.519 score. Yes, those artifacts generate the hard 00:08:37.519 --> 00:08:40.259 data the system requires to prove that it's functioning. 00:08:41.159 --> 00:08:44.220 But in demanding those specific artifacts, the 00:08:44.220 --> 00:08:48.519 actual messy human process of deliberative learning 00:08:48.519 --> 00:08:50.980 was completely overshadowed. They just wanted 00:08:50.980 --> 00:08:53.820 the paper at the end. Right. The system optimized 00:08:53.820 --> 00:08:57.379 for the output, not the internal process. Let's 00:08:57.379 --> 00:08:59.720 build an analogy here to really solidify the 00:08:59.720 --> 00:09:00.940 mechanics of this, because I think it's something 00:09:00.940 --> 00:09:03.860 everyone experiences. Sure. Imagine you decide 00:09:03.860 --> 00:09:06.360 you want to get really fit. You want to build 00:09:06.360 --> 00:09:09.100 muscle mass. OK, good call. So you go out and 00:09:09.100 --> 00:09:11.379 buy a really expensive state -of -the -art smart 00:09:11.379 --> 00:09:13.740 scale. It measures your weight, your bone density, 00:09:13.980 --> 00:09:15.759 your body fat percentage, everything. It gives 00:09:15.759 --> 00:09:18.129 you all the data. all the data, and you become 00:09:18.129 --> 00:09:20.529 absolutely obsessed with the data from that scale. 00:09:20.850 --> 00:09:22.830 You track the numbers every single morning, you 00:09:22.830 --> 00:09:25.169 make intricate spreadsheets, you draw trend lines 00:09:25.169 --> 00:09:27.509 to project your future fitness. Sounds exhausting. 00:09:27.730 --> 00:09:30.149 It is. But you never actually go to the gym to 00:09:30.149 --> 00:09:33.269 lift heavy things. Oh no. You are entirely obsessed 00:09:33.269 --> 00:09:36.129 with the data while completely ignoring the actual 00:09:36.129 --> 00:09:38.730 physical process required to change your body. 00:09:39.429 --> 00:09:41.190 That is a brilliant way to look at it, really. 00:09:41.690 --> 00:09:44.190 Because the mechanics of muscle growth perfectly 00:09:44.190 --> 00:09:46.210 mirror the mechanics of system two thinking. 00:09:46.379 --> 00:09:49.320 How do you mean? Well, a muscle only grows when 00:09:49.320 --> 00:09:53.320 you subject it to intense stress, causing micro 00:09:53.320 --> 00:09:56.659 tears in the fibers. The lifting is what matters. 00:09:56.980 --> 00:10:00.019 The struggle. Yes. The weight on the scale is 00:10:00.019 --> 00:10:02.120 simply a byproduct. It's a lagging indicator 00:10:02.120 --> 00:10:05.279 of that physical exertion. You've confused the 00:10:05.279 --> 00:10:07.299 measurement of the thing for the thing itself. 00:10:07.460 --> 00:10:11.039 And in education, the data. the essay, the test 00:10:11.039 --> 00:10:13.340 score, is supposed to be a byproduct of a student's 00:10:13.340 --> 00:10:15.460 cognitive exertion. Their mental micro -tears. 00:10:15.559 --> 00:10:17.740 Right. Their system two deliberative thought. 00:10:18.360 --> 00:10:20.500 But the educational system gradually stopped 00:10:20.500 --> 00:10:23.039 caring about the invisible heavy lifting and 00:10:23.039 --> 00:10:25.080 started demanding only the visible numbers on 00:10:25.080 --> 00:10:28.039 the scale. So if you have a massive societal 00:10:28.039 --> 00:10:31.039 system that demands nothing but visible formatted 00:10:31.039 --> 00:10:33.970 data for decades. Decades of it. Right. It shouldn't 00:10:33.970 --> 00:10:35.850 be a surprise when someone eventually invents 00:10:35.850 --> 00:10:38.070 a machine perfectly designed to deliver exactly 00:10:38.070 --> 00:10:40.330 that. And here's where it gets really interesting. 00:10:40.750 --> 00:10:44.009 This is the core irony of Ackerman's post. The 00:10:44.009 --> 00:10:47.330 punchline. Yes. Generative AI enters the chat. 00:10:47.529 --> 00:10:50.590 And students, being incredibly adaptable, are 00:10:50.590 --> 00:10:54.350 now using these AI tools to give educators exactly, 00:10:54.669 --> 00:10:56.769 precisely what they have been demanding for 30 00:10:56.769 --> 00:10:58.870 years. They are delivering the pristine data. 00:10:59.070 --> 00:11:01.629 Instantly. By using AI to generate the answers, 00:11:01.850 --> 00:11:04.090 students are, as Eckerman puts it, cutting to 00:11:04.090 --> 00:11:06.789 the chase. They really are. Let's say a teacher's 00:11:06.789 --> 00:11:09.490 highly structured rubric demands a 500 -word 00:11:09.490 --> 00:11:12.730 essay with a clear thesis statement, three supporting 00:11:12.730 --> 00:11:15.929 arguments with historical citations, and a summarizing 00:11:15.929 --> 00:11:18.490 conclusion. A very standard design. Very standard. 00:11:18.769 --> 00:11:20.929 That is the required data for the gradebook. 00:11:21.309 --> 00:11:23.669 So a student feeds that prompt into a generative 00:11:23.669 --> 00:11:25.970 AI, and the machine generates it in about four 00:11:25.970 --> 00:11:28.990 seconds. No. Gone. Student hands it in. It checks 00:11:28.990 --> 00:11:31.730 every single box on the rubric. It's flawless 00:11:31.730 --> 00:11:34.070 data. But there's absolutely nothing else attached 00:11:34.070 --> 00:11:36.990 to it. Nothing. No struggle. No grappling with 00:11:36.990 --> 00:11:39.909 complex concepts. No synthesizing of disparate 00:11:39.909 --> 00:11:42.470 ideas. No late night stressing over a blank page. 00:11:42.830 --> 00:11:45.669 Exactly. No cognitive friction whatsoever. Zero 00:11:45.669 --> 00:11:49.220 system to deliberative thought. It is the equivalent 00:11:49.220 --> 00:11:52.139 of having a robot lift the heavy weights at the 00:11:52.139 --> 00:11:54.820 gym for you and then stepping on the scale and 00:11:54.820 --> 00:11:56.779 wondering why your muscles aren't growing. And 00:11:56.779 --> 00:11:59.220 this raises an important question, one that gets 00:11:59.220 --> 00:12:02.279 right to the heart of Shaw and Nave's subtitle, 00:12:02.519 --> 00:12:04.980 The Rise of Cognitive Surrender. Right, that 00:12:04.980 --> 00:12:07.419 phrase again. When we talk about cognitive surrender, 00:12:07.620 --> 00:12:09.759 we really have to look at the biology of why 00:12:09.759 --> 00:12:12.899 it happens so easily. It's biological. Oh, absolutely. 00:12:13.120 --> 00:12:16.039 The human brain is naturally efficient. Its primary 00:12:16.039 --> 00:12:19.639 evolutionary goal is to conserve energy. So we're 00:12:19.639 --> 00:12:22.700 inherently lazy. Well, evolutionarily thrifty, 00:12:22.740 --> 00:12:25.779 let's say. As we establish, system two thinking 00:12:25.779 --> 00:12:28.019 burns a massive amount of cognitive energy. It 00:12:28.019 --> 00:12:30.299 is genuinely tiring. Yeah, nobody likes doing 00:12:30.299 --> 00:12:33.000 taxes. Right. So when an external tool system 00:12:33.000 --> 00:12:35.860 three offers a way to produce the required data 00:12:35.860 --> 00:12:38.419 without burning any of that energy, The brain 00:12:38.419 --> 00:12:41.259 enthusiastically surrenders the task. Oh, wow. 00:12:41.360 --> 00:12:43.019 Think about that in your own life for a second. 00:12:43.360 --> 00:12:45.980 Think about the last time you had a massive complex 00:12:45.980 --> 00:12:48.120 email thread at work. Oh, I know where this is 00:12:48.120 --> 00:12:50.759 going. And instead of reading it, you use an 00:12:50.759 --> 00:12:54.100 AI tool to summarize it for you. Guilty. We all 00:12:54.100 --> 00:12:55.740 do it. You got the bullet points. You got the 00:12:55.740 --> 00:12:58.559 data. But did you actually retain any of the 00:12:58.559 --> 00:13:01.220 nuances of that conversation? Probably not. Did 00:13:01.220 --> 00:13:04.240 you grasp the subtle frustrations of your coworkers? 00:13:05.000 --> 00:13:07.940 No, because you bypassed the System 2 process 00:13:07.940 --> 00:13:10.480 of reading and synthesizing. You conserved your 00:13:10.480 --> 00:13:12.820 energy, but you surrendered the deep understanding. 00:13:12.980 --> 00:13:15.639 And students are doing exactly that on a systemic 00:13:15.639 --> 00:13:18.840 level. By handing over the pure data via AI, 00:13:19.419 --> 00:13:22.100 students are successfully fulfilling the system's 00:13:22.100 --> 00:13:25.320 explicit demands. But they are doing it by relying 00:13:25.320 --> 00:13:28.139 entirely on System 3. They're outsourcing the 00:13:28.139 --> 00:13:30.360 brain work. They have achieved maximum efficiency 00:13:30.360 --> 00:13:33.200 in data production with zero cognitive development. 00:13:33.279 --> 00:13:35.220 OK, I have to stop and ask a practical question 00:13:35.220 --> 00:13:37.419 here. Go for it. Let's put ourselves in the shoes 00:13:37.419 --> 00:13:40.159 of the system itself. You know, you are a school 00:13:40.159 --> 00:13:42.419 principal or a state education board member. 00:13:42.580 --> 00:13:45.659 OK, I'm the boss. You spend years setting up 00:13:45.659 --> 00:13:49.519 a rigorous, standardized system asking for pure, 00:13:49.940 --> 00:13:52.580 quantifiable data to ensure students are meeting 00:13:52.580 --> 00:13:55.620 specific benchmarks. Right. The student utilizing 00:13:55.620 --> 00:13:59.440 new technology provides that pure data with incredible 00:13:59.440 --> 00:14:02.899 efficiency. On a purely mechanical administrative 00:14:02.899 --> 00:14:06.159 level, shouldn't the system be thrilled? I mean... 00:14:06.159 --> 00:14:08.840 The demand was met. The spreadsheet is full of 00:14:08.840 --> 00:14:11.639 perfect scores. Mechanically speaking, yes. The 00:14:11.639 --> 00:14:14.059 system is operating at peak efficiency. Right. 00:14:14.600 --> 00:14:17.220 But the humans inside the system... the teachers 00:14:17.220 --> 00:14:20.679 and educators, are not thrilled at all. Not even 00:14:20.679 --> 00:14:23.779 a little bit. In fact, they are feeling an overwhelming 00:14:23.779 --> 00:14:27.539 sense of dread, frustration, and honestly an 00:14:27.539 --> 00:14:31.159 existential crisis. Which leads directly to Ackerman's 00:14:31.159 --> 00:14:34.039 brutal foundational conclusion. Yes. So what 00:14:34.039 --> 00:14:36.580 does this all mean? Why the dread if the data 00:14:36.580 --> 00:14:39.120 is perfect? Ackerman delivers a very striking, 00:14:39.299 --> 00:14:42.299 almost melancholic conclusion in his post. He 00:14:42.299 --> 00:14:44.500 says educators are getting exactly what they 00:14:44.500 --> 00:14:47.379 asked for, but they find it deeply dissatisfying 00:14:47.379 --> 00:14:49.299 because we were asking for something that does 00:14:49.299 --> 00:14:51.960 not matter. Wow. Let that sink in for a second. 00:14:52.019 --> 00:14:54.000 We were asking for something that does not matter. 00:14:54.159 --> 00:14:57.000 It is a devastating realization for an entire 00:14:57.000 --> 00:14:59.379 industry. I can't imagine realizing that after 00:14:59.379 --> 00:15:03.759 a 30 -year career. The pure data, the flawlessly 00:15:03.759 --> 00:15:07.379 generated AISA, the perfectly completed worksheet 00:15:07.379 --> 00:15:12.179 is fundamentally worthless. if it isn't a byproduct 00:15:12.179 --> 00:15:16.039 of the student's internal system to deliberative 00:15:16.039 --> 00:15:18.299 cognition. The essay itself was never the point. 00:15:18.740 --> 00:15:21.240 Exactly. The actual text on the paper was never 00:15:21.240 --> 00:15:24.259 the point. The point was the structural neurological 00:15:24.259 --> 00:15:26.440 changes that happen inside a student's brain 00:15:26.440 --> 00:15:28.679 while they struggle in the dark to write the 00:15:28.679 --> 00:15:31.919 essay. That is such a crucial, clarifying distinction. 00:15:32.379 --> 00:15:34.600 The output is meaningless without the process. 00:15:34.799 --> 00:15:36.519 Completely meaningless. It's the difference between 00:15:36.519 --> 00:15:38.960 buying a souvenir and actually going on the vacation. 00:15:39.039 --> 00:15:40.639 Oh, that's a good one. Let's say the goal of 00:15:40.639 --> 00:15:42.620 a trip to Paris is to broaden your worldview, 00:15:42.940 --> 00:15:44.899 right? To struggle with the language, to get 00:15:44.899 --> 00:15:47.059 lost in the streets, and to experience a new 00:15:47.059 --> 00:15:49.919 culture. The friction of travel. Exactly. An 00:15:49.919 --> 00:15:52.679 AI can instantly print out a beautiful, high 00:15:52.679 --> 00:15:54.940 -resolution photo of you standing at the Eiffel 00:15:54.940 --> 00:15:57.919 Tower. It looks perfect. Flawless. That's the 00:15:57.919 --> 00:16:00.120 souvenir that... That's the data, but it means 00:16:00.120 --> 00:16:02.279 absolutely nothing to your personal development 00:16:02.279 --> 00:16:04.279 because you never actually traveled there. You 00:16:04.279 --> 00:16:06.299 just have the picture. Right. You never tasted 00:16:06.299 --> 00:16:08.220 the food. You never walked the streets. You never 00:16:08.220 --> 00:16:10.820 experienced the friction. The education system 00:16:10.820 --> 00:16:13.840 spent 30 years demanding souvenirs, assuming 00:16:13.840 --> 00:16:16.100 that handing in a pristine souvenir meant the 00:16:16.100 --> 00:16:18.639 student had taken the grueling trip. And generative 00:16:18.639 --> 00:16:21.320 AI completely broke that core assumption. It 00:16:21.320 --> 00:16:24.120 shattered it. AI can manufacture an infinite 00:16:24.120 --> 00:16:27.620 number of perfect souvenirs at zero cost. It 00:16:27.620 --> 00:16:30.080 entirely severed the link between the product 00:16:30.080 --> 00:16:32.919 and the process, exposing the massive flaw in 00:16:32.919 --> 00:16:35.340 the metric itself. This synthesizes the entire 00:16:35.340 --> 00:16:38.480 journey for us so clearly. The conflict is fully, 00:16:38.759 --> 00:16:41.600 uncomfortably exposed now. It really is. This 00:16:41.600 --> 00:16:44.679 decades -long obsession with data metrics, with 00:16:44.679 --> 00:16:47.399 what Shaw and Knave now classify as System 3 00:16:47.399 --> 00:16:49.620 artificial thinking, accidentally highlighted 00:16:49.620 --> 00:16:52.240 a very hard truth. Which is? In genuine learning, 00:16:52.419 --> 00:16:54.700 that System 2 deliberative thought is inherently 00:16:54.700 --> 00:16:57.639 messy. It is invisible. It is really hard to 00:16:57.639 --> 00:17:00.080 measure on a spreadsheet or grade with a standardized 00:17:00.080 --> 00:17:03.220 rubric. It involves failure, hesitation, slow 00:17:03.220 --> 00:17:06.000 progress, and immense cognitive friction. Everything 00:17:06.000 --> 00:17:08.420 the system hates. Right. It is the exact opposite 00:17:08.420 --> 00:17:11.619 of a clean, instantly generated data point. And 00:17:11.619 --> 00:17:14.680 crucially, it is impossible to fake with an AI. 00:17:15.150 --> 00:17:18.329 By getting exactly what they asked for, instant, 00:17:18.529 --> 00:17:20.789 perfect data educators realized they had been 00:17:20.789 --> 00:17:23.789 asking the wrong question all along. They asked 00:17:23.789 --> 00:17:26.250 for the final product when the only thing that 00:17:26.250 --> 00:17:28.630 actually mattered was the internal process. The 00:17:28.630 --> 00:17:31.210 heavy lifting at the gym. The messy vacation. 00:17:31.529 --> 00:17:33.829 And the introduction of System 3 into everyday 00:17:33.829 --> 00:17:36.509 life means this isn't just an education problem 00:17:36.509 --> 00:17:39.349 anymore? Oh no. It's everywhere. This dynamic 00:17:39.349 --> 00:17:41.930 is everywhere. The temptation to bypass the messy 00:17:41.930 --> 00:17:45.069 process and just grab the clean data is practically 00:17:45.069 --> 00:17:47.849 universal now. So true. Let's quickly recap how 00:17:47.849 --> 00:17:49.990 we got here before we wrap up. Sure. We started 00:17:49.990 --> 00:17:51.970 with Daniel Kahneman's foundational framework 00:17:51.970 --> 00:17:55.089 of human thought. System one, the fast, intuitive, 00:17:55.210 --> 00:17:58.269 low -energy gut reaction. The bike swerve. Right. 00:17:58.549 --> 00:18:00.829 And system two, the slow, deliberative, heavy 00:18:00.829 --> 00:18:03.450 lifting brain power that actually builds understanding. 00:18:03.589 --> 00:18:06.180 The friction. Exactly. Then we looked at how 00:18:06.180 --> 00:18:09.059 Shaw and Nave's 2026 paper introduced the artificial 00:18:09.059 --> 00:18:11.910 system 3. trusting the data and outsourcing the 00:18:11.910 --> 00:18:13.930 thought process entirely to conserve energy, 00:18:14.309 --> 00:18:17.089 which leads to cognitive surrender. A great term. 00:18:17.289 --> 00:18:20.289 Really is. Finally, we arrived at Gary Ackerman's 00:18:20.289 --> 00:18:23.490 incredibly ironic realization. The education 00:18:23.490 --> 00:18:26.470 system's relentless decades -long demand for 00:18:26.470 --> 00:18:30.490 pure, measurable data led to a completely hollow 00:18:30.490 --> 00:18:33.569 victory when generative AI gave them exactly 00:18:33.569 --> 00:18:36.029 what they asked for, completely devoid of the 00:18:36.029 --> 00:18:38.130 actual human learning process. Which leaves us 00:18:38.130 --> 00:18:41.170 with a necessary and slightly uncomfortable application 00:18:41.170 --> 00:18:44.509 of this idea. Let's hear it. If a massive well 00:18:44.509 --> 00:18:47.390 -intentioned system like education fell so deeply 00:18:47.390 --> 00:18:50.309 into the trap of obsessing over data that ultimately 00:18:50.309 --> 00:18:52.630 didn't matter It is worth looking at your own 00:18:52.630 --> 00:18:55.230 life. Oh, wow. Yeah. Think about your daily routines, 00:18:55.569 --> 00:18:57.670 your workplace habits, or your long -term goals. 00:18:58.269 --> 00:19:00.789 Where are you blindly demanding system three 00:19:00.789 --> 00:19:03.069 data? Where are you taking the shortcut? Exactly. 00:19:03.369 --> 00:19:04.990 Where are you taking the hyper -efficient shortcut 00:19:04.990 --> 00:19:08.089 just to get the result, and in doing so, completely 00:19:08.089 --> 00:19:11.809 bypassing the messy human system two deliberative 00:19:11.809 --> 00:19:14.250 process that actually gives that result its value? 00:19:14.430 --> 00:19:16.150 Are you just stepping on the smart scale every 00:19:16.150 --> 00:19:18.230 day and ignoring the heavy lifting? That's the 00:19:18.230 --> 00:19:21.019 question. It is a phenomenal question to chew 00:19:21.019 --> 00:19:23.480 on as these tools become more and more integrated 00:19:23.480 --> 00:19:26.339 into our lives. Thank you so much for joining 00:19:26.339 --> 00:19:29.480 us on this deep dive into the mechanics of cognition 00:19:29.480 --> 00:19:32.099 and the irony of our data obsession. He's a fun 00:19:32.099 --> 00:19:34.920 one. Keep questioning the metrics. Keep leaning 00:19:34.920 --> 00:19:37.880 into that slow, friction -filled, deliberative 00:19:37.880 --> 00:19:40.680 thinking. And remember to always stay critically 00:19:40.680 --> 00:19:43.140 engaged in a world that is practically begging 00:19:43.140 --> 00:19:45.880 you to surrender to the easy answer. Don't just 00:19:45.880 --> 00:19:48.240 take the souvenir. Exactly. Because as we learned 00:19:48.240 --> 00:19:50.339 today, getting exactly what you asked for might 00:19:50.339 --> 00:19:52.000 just mean you are asking for the wrong thing 00:19:52.000 --> 00:19:53.380 all along. Catch you next time.