WEBVTT 00:00:00.000 --> 00:00:03.960 Welcome back to the Deep Dive. Today we are really 00:00:03.960 --> 00:00:05.919 opening up the hood on something fundamental, 00:00:06.660 --> 00:00:08.560 learning. Right. It's something we all do every 00:00:08.560 --> 00:00:11.119 day. Exactly. But everyone thinks they sort of 00:00:11.119 --> 00:00:13.460 understand it, you know, but when you actually 00:00:13.460 --> 00:00:15.580 look closely at the research, the physical stuff, 00:00:15.699 --> 00:00:17.839 the brain science, even the philosophy behind 00:00:17.839 --> 00:00:21.039 it, well, the picture gets way bigger. It really 00:00:21.039 --> 00:00:22.780 does. It's not just sitting quietly with a book. 00:00:22.820 --> 00:00:25.359 No. Not at all. We often think of learning as 00:00:25.359 --> 00:00:28.339 this purely internal thing, happens in your head, 00:00:28.440 --> 00:00:31.120 end of story. But our deep dive today, looking 00:00:31.120 --> 00:00:34.500 at these sources, it forces us to stretch those 00:00:34.500 --> 00:00:37.719 boundaries, like a lot. So our mission basically 00:00:37.719 --> 00:00:42.380 is to unpack four key ideas. First, where does 00:00:42.380 --> 00:00:45.000 learning actually happen? And, spoiler, it's 00:00:45.000 --> 00:00:46.659 inside and outside the brain. That's a big one. 00:00:46.820 --> 00:00:49.340 Huge. Second, what is learning, really? It's 00:00:49.340 --> 00:00:52.520 not just memorizing facts. It's multi -dimensional. 00:00:52.640 --> 00:00:54.799 We'll get into that. Third, how do we get better 00:00:54.799 --> 00:00:57.159 at it? How do we improve our minds? It involves 00:00:57.159 --> 00:00:59.479 effort, but maybe not just individual effort. 00:00:59.700 --> 00:01:01.960 The social angle. Very important. And finally, 00:01:02.140 --> 00:01:04.659 a bit of a reality check. How much can we actually 00:01:04.659 --> 00:01:07.400 trust our own memories, our internal hard drives? 00:01:07.780 --> 00:01:11.939 Ah, yes. The faulty hardware issue. Exactly. 00:01:12.420 --> 00:01:14.299 OK, let's unpack this. Let's start right there 00:01:14.299 --> 00:01:17.040 with the location of learning. You said it's 00:01:17.040 --> 00:01:20.180 inside and outside. The inside part feels intuitive, 00:01:20.379 --> 00:01:23.099 right? The brain's the engine. That's the classic 00:01:23.099 --> 00:01:25.799 starting point, yes. And it's well documented. 00:01:26.340 --> 00:01:28.739 Scientists can literally see changes. Like physical 00:01:28.739 --> 00:01:31.519 changes. Physical changes. They track how blood 00:01:31.519 --> 00:01:33.359 flows in the brain when someone learns something 00:01:33.359 --> 00:01:37.500 new. And what they see suggests that... neural 00:01:37.500 --> 00:01:39.560 pathways that weren't very active before suddenly 00:01:39.560 --> 00:01:42.900 light up, they get used. And they even get, well, 00:01:43.340 --> 00:01:46.540 insulated myelinated to work faster. Then, crucially, 00:01:46.700 --> 00:01:48.939 when you use that knowledge later, those exact 00:01:48.939 --> 00:01:51.620 same physical pathways activate again. Wow. So 00:01:51.620 --> 00:01:54.170 learning is literally like... building new roads 00:01:54.170 --> 00:01:56.590 in your brain. You could put it that way. Constructing 00:01:56.590 --> 00:01:59.109 a new neural landscape. Yeah. Inside your skull. 00:01:59.209 --> 00:02:01.349 OK, that's the inside story. Makes sense. But 00:02:01.349 --> 00:02:03.069 then the sources kind of shift gears, don't they? 00:02:03.250 --> 00:02:05.090 They argue humans are special because we move 00:02:05.090 --> 00:02:07.290 cognition out of the brain. That's right. This 00:02:07.290 --> 00:02:10.550 idea of transferring cognition or maybe outsourcing 00:02:10.550 --> 00:02:13.229 it to the environment, it's gained a lot of traction. 00:02:13.530 --> 00:02:16.150 Outsourcing cognition. How does that work? Well, 00:02:16.449 --> 00:02:18.289 think about writing. It's a technology, right? 00:02:18.830 --> 00:02:21.490 Once we learn to write, suddenly a huge chunk 00:02:21.490 --> 00:02:23.889 of our memory doesn't need to be stored biologically. 00:02:23.889 --> 00:02:27.090 Ah, I see. We can just download it onto paper 00:02:27.090 --> 00:02:29.889 or, you know, into digital files, bytes of information. 00:02:30.229 --> 00:02:32.930 So when I jot down notes or use a calculator, 00:02:33.789 --> 00:02:36.629 that's external cognition. That's a perfect example. 00:02:36.949 --> 00:02:38.889 The calculation isn't only happening in your 00:02:38.889 --> 00:02:41.129 head. The paper, the calculator, they become 00:02:41.129 --> 00:02:43.590 part of the cognitive loop. The tool is part 00:02:43.590 --> 00:02:45.710 of the thinking. Precisely. It connects to these 00:02:45.710 --> 00:02:48.509 bigger ideas like the extended mind theory or 00:02:48.509 --> 00:02:52.169 what some call 4E cognition. Basically, yourself 00:02:52.169 --> 00:02:54.509 isn't just confined to your skin and skull. Okay, 00:02:54.530 --> 00:02:56.449 that's kind of mind -bending. What's fascinating 00:02:56.449 --> 00:03:00.530 here is the memory, say, of that complex formula 00:03:00.530 --> 00:03:03.729 you wrote down, it now exists on the page. not 00:03:03.729 --> 00:03:06.110 just in a sun app somewhere. And this changes 00:03:06.110 --> 00:03:08.870 everything. It means human learning isn't purely 00:03:08.870 --> 00:03:12.229 biological. It's also fundamentally a socio -environmental 00:03:12.229 --> 00:03:14.430 thing. You learn to write and boom, you've just 00:03:14.430 --> 00:03:16.409 massively upgraded your working memory capacity. 00:03:16.509 --> 00:03:19.409 That makes sense. So learning isn't just passively 00:03:19.409 --> 00:03:22.090 absorbing stuff into those new brain roads. No. 00:03:22.330 --> 00:03:26.069 It has to be about engaging with Well with both 00:03:26.069 --> 00:03:28.370 the internal brain stuff and these external tools 00:03:28.370 --> 00:03:31.129 we create exactly right which leads nicely into 00:03:31.129 --> 00:03:33.389 the next point Yeah, let's talk about what learning 00:03:33.389 --> 00:03:36.310 actually is the sources say it's multi -dimensional 00:03:36.310 --> 00:03:39.370 So it's more than just remembering things much 00:03:39.370 --> 00:03:41.629 more if learning was just recollection. We'd 00:03:41.629 --> 00:03:43.610 be missing the whole point Okay, the sources 00:03:43.610 --> 00:03:47.669 list activities like Analyzing information creating 00:03:47.669 --> 00:03:50.449 new things based on it critiquing ideas. It's 00:03:50.449 --> 00:03:53.430 active. It's manipulation perception recollection. 00:03:53.449 --> 00:03:55.830 Yeah but also analyzing, creating, engaging. 00:03:55.870 --> 00:03:59.569 Right. It moves learning from just having information 00:03:59.569 --> 00:04:02.590 to using information. So how do you know? For 00:04:02.590 --> 00:04:04.310 the listener, how can you tell if you've really 00:04:04.310 --> 00:04:07.330 learned something versus just memorized it? That's 00:04:07.330 --> 00:04:08.830 the key question. And the sources are pretty 00:04:08.830 --> 00:04:11.990 clear. Someone who has truly learned can do way 00:04:11.990 --> 00:04:14.430 more than just repeat the information back. OK. 00:04:14.689 --> 00:04:17.560 The real test is application. Can you take that 00:04:17.560 --> 00:04:20.060 knowledge and use it in a new situation? Can 00:04:20.060 --> 00:04:22.220 you analyze something unexpected with it? Can 00:04:22.220 --> 00:04:25.199 you create something original from it? Give me 00:04:25.199 --> 00:04:29.360 an example. OK. Think about physics. A student 00:04:29.360 --> 00:04:32.100 might perfectly recite Newton's laws. That's 00:04:32.100 --> 00:04:34.920 recollection. Internal pathways firing. Got it. 00:04:35.220 --> 00:04:37.860 But an engineer who uses those same laws to figure 00:04:37.860 --> 00:04:41.319 out why a bridge collapsed. Or even better, to 00:04:41.319 --> 00:04:45.110 design a completely new stable structure. That's 00:04:45.110 --> 00:04:47.490 learned knowledge. They're analyzing, creating, 00:04:48.410 --> 00:04:50.829 using that multi -dimensional skill set. Ah, 00:04:51.029 --> 00:04:53.350 I see. So moving from being a parrot to being 00:04:53.350 --> 00:04:55.189 an architect. That's a great way to put it, yeah. 00:04:55.310 --> 00:04:57.290 From repetition to creation. That's the sign 00:04:57.290 --> 00:04:59.170 you've truly acquired the knowledge. Okay, that's 00:04:59.170 --> 00:05:00.810 a really important distinction. But being an 00:05:00.810 --> 00:05:03.550 architect takes work, right? Effort. Definitely. 00:05:03.850 --> 00:05:05.990 Which brings us neatly to the next bit. The effort 00:05:05.990 --> 00:05:09.430 economy. How do we actually get smarter? How 00:05:09.430 --> 00:05:12.050 do we build those skills? Well, the basic idea 00:05:12.050 --> 00:05:14.730 is probably familiar. The brain is a muscle analogy. 00:05:14.889 --> 00:05:17.389 Pretty much. Our cognitive abilities seem to 00:05:17.389 --> 00:05:20.889 improve with practice, with effort. If you want 00:05:20.889 --> 00:05:23.009 those neural pathways, those roads we talked 00:05:23.009 --> 00:05:25.930 about, to become stronger and faster, you got 00:05:25.930 --> 00:05:27.930 to use them. You got to use them repeatedly. 00:05:28.430 --> 00:05:31.290 You need to stress them a bit, push them so they 00:05:31.290 --> 00:05:33.990 adapt and get more efficient. That's where the 00:05:33.990 --> 00:05:36.110 myelination of the insulation comes in. Right. 00:05:36.410 --> 00:05:38.970 So practice makes permanent, neurologically speaking. 00:05:39.050 --> 00:05:41.329 That's the idea. Active engagement. Purposeful 00:05:41.329 --> 00:05:44.490 practice. That's how you get those effort dividends 00:05:44.490 --> 00:05:47.269 that make knowledge stick. Use it once. The pathway 00:05:47.269 --> 00:05:51.029 is weak. Use it often. It deepens, becomes easier 00:05:51.029 --> 00:05:53.810 to access. Makes sense. But here's where it gets 00:05:53.810 --> 00:05:56.550 really interesting. The sources throw a curveball. 00:05:57.029 --> 00:05:59.529 They say this practice, this effort, it's not 00:05:59.529 --> 00:06:01.769 just a solo activity. Absolutely not. And this 00:06:01.769 --> 00:06:03.870 connects back to the external cognition idea. 00:06:04.189 --> 00:06:06.029 If learning extends into the environment. Then 00:06:06.029 --> 00:06:08.209 maybe the effort needed does too. It gets shared. 00:06:08.439 --> 00:06:11.139 Exactly. This is where the community aspect is 00:06:11.139 --> 00:06:14.300 just crucial. Individual practice lays the groundwork, 00:06:14.319 --> 00:06:17.639 builds that initial path. But the strongest, 00:06:17.759 --> 00:06:20.540 most flexible, most deeply understood knowledge, 00:06:21.060 --> 00:06:24.279 that gets forged socially. How so? Well, think 00:06:24.279 --> 00:06:26.800 about it. When you have to interact with other 00:06:26.800 --> 00:06:30.439 people, maybe a mentor, maybe peers, maybe even 00:06:30.439 --> 00:06:33.240 someone you disagree with, you're forced to use 00:06:33.240 --> 00:06:35.439 your knowledge differently. You have to explain 00:06:35.439 --> 00:06:38.509 it. Defend it. Maybe modify it based on their 00:06:38.509 --> 00:06:40.509 input. You have to connect it to what they know. 00:06:40.889 --> 00:06:43.870 That process It activates completely different 00:06:43.870 --> 00:06:45.709 parts of your understanding than just reviewing 00:06:45.709 --> 00:06:48.129 notes alone. It's like stress testing the knowledge. 00:06:48.410 --> 00:06:50.990 Perfectly put, that social negotiation, the back 00:06:50.990 --> 00:06:53.490 and forth, the synthesizing, it's like concentrated 00:06:53.490 --> 00:06:55.970 practice. It builds much tougher memories. And 00:06:55.970 --> 00:06:57.769 really importantly, it helps you see connections, 00:06:57.850 --> 00:07:01.029 how what you know links up with the bigger network 00:07:01.029 --> 00:07:03.290 of what others know. So it's the difference between 00:07:03.290 --> 00:07:05.790 practicing basketball drills alone versus playing 00:07:05.790 --> 00:07:07.870 an actual game. That's a great analogy, yeah. 00:07:08.129 --> 00:07:10.389 You might learn the move solo, but the real skill, 00:07:10.509 --> 00:07:12.870 the adaptive intelligence, comes from constantly 00:07:12.870 --> 00:07:15.930 adjusting to other players. So what you know 00:07:15.930 --> 00:07:18.790 and how you know it is fundamentally shaped by 00:07:18.790 --> 00:07:21.009 the people around you, by your social environment. 00:07:21.490 --> 00:07:23.990 You simply get a better return on your cognitive 00:07:23.990 --> 00:07:26.250 effort when you make it collaborative. OK, this 00:07:26.250 --> 00:07:28.930 taints a really empowering picture. We can improve. 00:07:29.009 --> 00:07:31.430 We can learn deeply. We use tools. We leverage 00:07:31.430 --> 00:07:36.230 community. But now, the final piece. The reality 00:07:36.230 --> 00:07:39.350 check. Ah, yes. The counterpoint. The sources 00:07:39.350 --> 00:07:42.230 acknowledge that learning creates these somewhat 00:07:42.230 --> 00:07:44.990 permanent brain changes. But then they hit us 00:07:44.990 --> 00:07:47.170 with the fact that our cognition isn't static. 00:07:48.370 --> 00:07:52.000 And worse. Our memories are basically unreliable. 00:07:52.459 --> 00:07:54.319 Yes. Unreliable is putting it mildly. The sources 00:07:54.319 --> 00:07:56.779 are quite blunt. Our memories are very faulty. 00:07:56.920 --> 00:07:59.500 Very faulty, okay. We aren't like video recorders 00:07:59.500 --> 00:08:02.060 capturing pristine data. We're more like storytellers 00:08:02.060 --> 00:08:04.639 constantly editing. Well, wait, how does that 00:08:04.639 --> 00:08:06.480 square with the permanent neural pathways? If 00:08:06.480 --> 00:08:08.759 I built the road, shouldn't the memory be solid? 00:08:09.000 --> 00:08:12.209 That's a really sharp question. The way to reconcile 00:08:12.209 --> 00:08:15.189 it seems to be differentiating between skills 00:08:15.189 --> 00:08:18.689 and specific content. The ability to analyze 00:08:18.689 --> 00:08:21.550 the skill of using language, those underlying 00:08:21.550 --> 00:08:23.790 neural structures, the how -to pathways, they 00:08:23.790 --> 00:08:25.750 can be very resilient, built through practice. 00:08:26.430 --> 00:08:29.290 But the content, the specific details of what 00:08:29.290 --> 00:08:31.610 happened, the sequence of events, who said what, 00:08:32.210 --> 00:08:35.210 that kind of episodic memory is incredibly susceptible 00:08:35.210 --> 00:08:37.929 to error. So I might remember how to ride a bike 00:08:37.929 --> 00:08:40.700 forever. But I might misremember the color of 00:08:40.700 --> 00:08:43.580 my first bike. Exactly. Or, much more significantly, 00:08:43.879 --> 00:08:46.059 misremember important details of past events 00:08:46.059 --> 00:08:48.639 or things you supposedly learned. Think about 00:08:48.639 --> 00:08:50.620 how much stuff you learned in school that's just 00:08:50.620 --> 00:08:53.679 gone. True. Lots. And even the stuff you think 00:08:53.679 --> 00:08:56.299 you remember clearly. Chances are you're misremembering 00:08:56.299 --> 00:08:58.559 parts of it. Our brains are constantly filling 00:08:58.559 --> 00:09:01.480 in gaps, smoothing out inconsistencies, rewriting 00:09:01.480 --> 00:09:03.639 the past to fit our present beliefs or needs. 00:09:04.059 --> 00:09:07.220 It's called confabulation. Confabulation. Every 00:09:07.220 --> 00:09:09.330 time you recall a memory you're not accessing 00:09:09.330 --> 00:09:12.730 a perfect file. You're basically rebuilding it. 00:09:13.029 --> 00:09:15.730 It's an active construction process, and errors 00:09:15.730 --> 00:09:18.450 creep in constantly. So my internal memory isn't 00:09:18.450 --> 00:09:21.330 just fading, it's actively changing the story 00:09:21.330 --> 00:09:24.549 behind my back. Pretty much. Which makes... Relying 00:09:24.549 --> 00:09:27.490 only on your internal memory, a very risky strategy, 00:09:27.590 --> 00:09:30.210 doesn't it? It really does. And this inherent 00:09:30.210 --> 00:09:32.690 faultiness is probably the strongest argument 00:09:32.690 --> 00:09:35.370 for why we evolved this capacity for external 00:09:35.370 --> 00:09:38.250 cognition in the first place. We need the notes, 00:09:38.429 --> 00:09:40.789 the books, the recordings, the shared knowledge 00:09:40.789 --> 00:09:43.669 of the group because our individual recall is 00:09:43.669 --> 00:09:46.830 just too biased and leaky. Okay, so what does 00:09:46.830 --> 00:09:48.610 this all mean? Let's try and tie these threads 00:09:48.610 --> 00:09:52.230 together. We learn both inside our brains and 00:09:52.230 --> 00:09:54.990 using external tools. Right, our cognition extends 00:09:54.990 --> 00:09:57.070 outward. We need to do more than just recall. 00:09:57.210 --> 00:09:59.909 We need to analyze and create. The multi -dimensional 00:09:59.909 --> 00:10:02.269 aspect. Our effort is best when it's social, 00:10:02.590 --> 00:10:04.789 practice with others. Because interaction strengthens 00:10:04.789 --> 00:10:07.149 understanding. And underpinning it all is the 00:10:07.149 --> 00:10:10.110 fact that our own internal memory is fundamentally 00:10:10.110 --> 00:10:12.889 unreliable. Very faulty. Yes. What does that 00:10:12.889 --> 00:10:15.789 leave us? Well, if we connect this to the bigger 00:10:15.789 --> 00:10:19.259 picture... Learning emerges as this kind of constant 00:10:19.259 --> 00:10:21.779 negotiation, doesn't it? A balancing act. Between 00:10:21.779 --> 00:10:25.860 what? Between our own constantly shifting error 00:10:25.860 --> 00:10:30.419 -prone neural pathways inside and the more stable 00:10:30.419 --> 00:10:33.980 external downloads, the notes, the files, the 00:10:33.980 --> 00:10:35.820 collective knowledge we create outside. The high 00:10:35.820 --> 00:10:37.539 wire act you called earlier. Yeah, something 00:10:37.539 --> 00:10:39.519 like that. And it means our effort has to be 00:10:39.519 --> 00:10:42.620 ongoing. It needs that social dimension precisely 00:10:42.620 --> 00:10:44.799 because we can't fully trust ourselves alone. 00:10:45.049 --> 00:10:47.730 We need those external tools and communal checks 00:10:47.730 --> 00:10:50.610 and balances as a direct response to how leaky 00:10:50.610 --> 00:10:53.169 our own memory is. We outsource knowledge not 00:10:53.169 --> 00:10:55.870 just for convenience, but for accuracy. For accuracy, 00:10:56.129 --> 00:10:58.629 for reliability, maybe even for survival. Which 00:10:58.629 --> 00:11:00.409 brings us right back to that initial startling 00:11:00.409 --> 00:11:03.870 idea. If writing things down or storing data 00:11:03.870 --> 00:11:06.870 digitally is actually part of learning happening 00:11:06.870 --> 00:11:10.090 outside the brain, and if we lean on this external 00:11:10.090 --> 00:11:13.379 network because our internal one is flawed, How 00:11:13.379 --> 00:11:16.519 does that change what we mean by individual intelligence? 00:11:16.759 --> 00:11:18.440 That's the million -dollar question, isn't it? 00:11:18.620 --> 00:11:21.440 Is being smart just about how many facts you 00:11:21.440 --> 00:11:23.759 can cram into your head anymore? Or is it more 00:11:23.759 --> 00:11:26.759 about how skillfully you can navigate and use 00:11:26.759 --> 00:11:29.139 this vast external knowledge system we've all 00:11:29.139 --> 00:11:31.399 built together? A question worth mulling over. 00:11:31.620 --> 00:11:32.940 Definitely something to think about until our 00:11:32.940 --> 00:11:33.580 next deep dive.