WEBVTT 00:00:00.000 --> 00:00:02.140 You know, usually when we welcome you to today's 00:00:02.140 --> 00:00:04.679 deep dive and we're looking at some massive, 00:00:05.259 --> 00:00:06.960 intimidating concept. Right, especially stuff 00:00:06.960 --> 00:00:09.839 sitting at the intersection of advanced math 00:00:09.839 --> 00:00:12.000 and computer science. Yeah, exactly. You kind 00:00:12.000 --> 00:00:14.039 of brace yourself for this absolute onslaught 00:00:14.039 --> 00:00:19.160 of information. You expect a sprawling architectural 00:00:19.160 --> 00:00:22.440 blueprint of knowledge, basically. Dense histories, 00:00:22.760 --> 00:00:25.960 competing theories, endless citations. Right. 00:00:26.079 --> 00:00:28.480 You get ready to wade through all of that. But 00:00:28.480 --> 00:00:31.410 today, Our source material is the shortest we've 00:00:31.410 --> 00:00:34.289 ever had, literally. It is remarkably short. 00:00:34.490 --> 00:00:36.969 Because you search for the term theorem prover, 00:00:37.390 --> 00:00:39.630 and you don't find a blueprint. You just find, 00:00:39.630 --> 00:00:42.409 well, basically a blank index card with two arrows 00:00:42.409 --> 00:00:44.310 pointing in radically different directions. Yeah, 00:00:44.429 --> 00:00:46.969 it forces an immediate halt. Like you arrived 00:00:46.969 --> 00:00:49.429 looking for a definition, and instead, you're 00:00:49.429 --> 00:00:52.009 handed this mandatory choice. Exactly. It's just 00:00:52.009 --> 00:00:54.210 the text of a Wikipedia disambiguation page. 00:00:54.429 --> 00:00:57.710 So today's mission is, how do we extract deep 00:00:57.710 --> 00:01:00.390 meaning from a mere handful of words? Because 00:01:00.390 --> 00:01:02.969 the source material we have today isn't an encyclopedic 00:01:02.969 --> 00:01:05.870 breakdown at all. It's purely a structural routing 00:01:05.870 --> 00:01:09.650 mechanism. Right. And our goal is to deeply analyze 00:01:09.650 --> 00:01:12.829 the specific language used on this tiny page 00:01:12.829 --> 00:01:17.010 to understand this fundamental split in how humans 00:01:17.010 --> 00:01:19.730 approach mathematical and logical proofs. It 00:01:19.730 --> 00:01:22.909 really exposes the raw skeletal structure of 00:01:22.909 --> 00:01:26.689 how a specific scientific discipline has fractured, 00:01:26.930 --> 00:01:28.930 essentially. It totally does. OK, let's unpack 00:01:28.930 --> 00:01:32.000 this. starting with just the sheer ontological 00:01:32.000 --> 00:01:34.719 shock of a disambiguation page even existing 00:01:34.719 --> 00:01:36.959 for this specific term. Yeah, it's quite the 00:01:36.959 --> 00:01:38.620 realization. Because, I mean, before we even 00:01:38.620 --> 00:01:40.719 look at the specific types of theorem provers, 00:01:40.739 --> 00:01:42.980 we have to look at the structure itself. My first 00:01:42.980 --> 00:01:45.659 assumption when I see a disambiguation page is 00:01:45.659 --> 00:01:48.739 administrative pedantry. Right. Wikipedia loves 00:01:48.739 --> 00:01:51.599 its obsessive categorization. Exactly. It just 00:01:51.599 --> 00:01:53.680 seems like the system is splitting hairs over 00:01:53.680 --> 00:01:56.319 software naming conventions just to keep its 00:01:56.319 --> 00:01:58.659 internal databases tidy, like standing at a fork. 00:01:58.569 --> 00:02:01.290 in a road where the sign is just too broad. So 00:02:01.290 --> 00:02:03.709 why wouldn't this just be one big umbrella article? 00:02:04.250 --> 00:02:06.870 What does the need for absolute disambiguation 00:02:06.870 --> 00:02:09.370 tell us about the nature of these tools? Well, 00:02:09.370 --> 00:02:11.490 what's fascinating here is that I'd argue it 00:02:11.490 --> 00:02:13.569 represents something much deeper than just, you 00:02:13.569 --> 00:02:16.090 know, database management. Oh, really? Yeah. 00:02:16.349 --> 00:02:20.090 In the architecture of human knowledge, a disambiguation 00:02:20.090 --> 00:02:23.210 page kind of acts as a quarantine zone for linguistic 00:02:23.210 --> 00:02:25.810 collision. A quarantine zone, wow. Right. Because 00:02:25.810 --> 00:02:28.469 if you look up a term and land on a unified article, 00:02:28.990 --> 00:02:32.189 it implies a unified consensus on what that concept 00:02:32.189 --> 00:02:35.169 actually is. Like, if two tools are just variations 00:02:35.169 --> 00:02:37.750 of the same underlying philosophy, they share 00:02:37.750 --> 00:02:40.419 a page. just get different subheadings. Exactly. 00:02:40.759 --> 00:02:43.659 The fact that this page isn't just offering synonyms, 00:02:43.780 --> 00:02:46.620 the fact that theorem prover requires a hard 00:02:46.620 --> 00:02:49.039 fork means the term itself has become dangerously 00:02:49.039 --> 00:02:51.659 ambiguous in the field of formal logic. So the 00:02:51.659 --> 00:02:54.120 text is forcing you to declare your intention 00:02:54.120 --> 00:02:56.439 before you're even allowed to proceed. Yeah, 00:02:56.439 --> 00:02:58.599 it requires the reader to make an immediate choice 00:02:58.599 --> 00:03:00.560 about what kind of logic system they're actually 00:03:00.560 --> 00:03:03.539 looking for. It's saying the umbrella term theorem 00:03:03.539 --> 00:03:07.039 prover is obsolete or at least functionally useless 00:03:07.039 --> 00:03:09.780 without a modifier. But this is where I get caught 00:03:09.780 --> 00:03:14.039 up, because both of these paths are aiming for 00:03:14.039 --> 00:03:17.180 the exact same destination, aren't they? Absolute 00:03:17.180 --> 00:03:19.639 mathematical truth. They are, yes. I mean, a 00:03:19.639 --> 00:03:22.240 theorem is a statement proven based on established 00:03:22.240 --> 00:03:24.919 axioms. It's literally the highest stakes in 00:03:24.919 --> 00:03:28.240 logic. So why wouldn't the tools built to find 00:03:28.240 --> 00:03:31.560 that truth belong under the same conceptual roof? 00:03:31.759 --> 00:03:33.819 Well, because the destination might be the same. 00:03:34.139 --> 00:03:36.960 But the mechanism of travel, and specifically 00:03:36.960 --> 00:03:40.400 humanity's role in that travel, is entirely irreconcilable 00:03:40.400 --> 00:03:43.099 between the two paths. Irreconcilable. OK. When 00:03:43.099 --> 00:03:46.460 dealing with absolute undeniable truth. The how 00:03:46.460 --> 00:03:49.120 is just as critical as the what. Right. And the 00:03:49.120 --> 00:03:52.060 two bullet points on this page describe two fundamentally 00:03:52.060 --> 00:03:54.819 incompatible philosophies of computation. OK. 00:03:54.919 --> 00:03:56.860 So let's walk down the first path provided by 00:03:56.860 --> 00:03:59.300 the text, which isolates the first of those philosophies. 00:03:59.400 --> 00:04:01.259 The automated approach. Right. The automated 00:04:01.259 --> 00:04:03.199 theorem prover. Just three words on the page. 00:04:03.800 --> 00:04:06.870 But attaching the word automated. to the concept 00:04:06.870 --> 00:04:10.069 of proving mathematical truth. It implies this 00:04:10.069 --> 00:04:12.409 massive shift in agency. It absolutely does. 00:04:12.669 --> 00:04:14.949 My initial read here, giving an analogy for you 00:04:14.949 --> 00:04:16.709 listening, is that we're talking about a machine 00:04:16.709 --> 00:04:20.170 operating as a self -sufficient agent. Is an 00:04:20.170 --> 00:04:22.810 automated prover, like putting a mathematical 00:04:22.810 --> 00:04:25.850 problem into a microwave, hitting a button and 00:04:25.850 --> 00:04:28.069 just waiting for the beep? I actually love that 00:04:28.069 --> 00:04:30.750 analogy, yes. If we connect this to the bigger 00:04:30.750 --> 00:04:33.810 picture, the word automated, in this context, 00:04:34.149 --> 00:04:37.189 describes a specific mechanistic approach. It 00:04:37.189 --> 00:04:40.069 implies independence and self -sufficiency. So 00:04:40.069 --> 00:04:42.250 the machine takes the reins entirely. Exactly. 00:04:42.949 --> 00:04:44.790 When you're trying to prove a theorem, you're 00:04:44.790 --> 00:04:47.389 essentially standing in a maze of infinite logical 00:04:47.389 --> 00:04:50.189 dimensions. Every step you take from your starting 00:04:50.189 --> 00:04:53.800 axioms branches into, like, thousands of impossible 00:04:53.800 --> 00:04:55.620 subsequent steps. And most of those are dead 00:04:55.620 --> 00:04:57.759 ends, right? The vast majority. Dead ends are 00:04:57.759 --> 00:05:00.300 infinite loops. So an automated prover is an 00:05:00.300 --> 00:05:02.480 algorithm deployed into that maze. Without a 00:05:02.480 --> 00:05:05.060 map. Just hitting a button and walking away without 00:05:05.060 --> 00:05:07.399 any human intervention. Right. It has the rules 00:05:07.399 --> 00:05:09.939 of movement, but no intuition about the layout. 00:05:10.170 --> 00:05:13.589 it has to rely on pure computational brute force. 00:05:13.610 --> 00:05:16.470 Feeling its way around blindly in the dark? Yeah, 00:05:16.509 --> 00:05:19.230 blindly feeling its way through the logical branches 00:05:19.230 --> 00:05:23.329 until it either hits a wall or miraculously stumbles 00:05:23.329 --> 00:05:25.569 upon the path that leads to the final proven 00:05:25.569 --> 00:05:29.189 state. The critical takeaway from automated is 00:05:29.189 --> 00:05:32.110 the absolute exclusion of the human mind from 00:05:32.110 --> 00:05:35.269 that internal navigation process. Which, I mean, 00:05:35.269 --> 00:05:37.649 that seems entirely backward to how humans actually 00:05:37.649 --> 00:05:40.240 do math. Oh, completely. We don't just blindly 00:05:40.240 --> 00:05:43.139 test every possible combination of numbers until 00:05:43.139 --> 00:05:45.639 something works. We have an intuitive grasp of 00:05:45.639 --> 00:05:48.420 the shape of a problem, you know. We see patterns. 00:05:48.579 --> 00:05:50.720 Yeah. We know instinctively that a certain approach 00:05:50.720 --> 00:05:53.399 is a total waste of time, so we just skip entire 00:05:53.399 --> 00:05:56.100 branches of the maze. And that is precisely the 00:05:56.100 --> 00:05:59.220 limitation inherent in the automated path. Machines 00:05:59.220 --> 00:06:01.779 do not possess intuition. They don't get that 00:06:01.779 --> 00:06:03.660 aha moment. Right. They don't have that sudden 00:06:03.660 --> 00:06:05.500 flash of insight where the answer just clicks. 00:06:05.579 --> 00:06:07.779 They have formal mechanics. So when you rely 00:06:07.779 --> 00:06:10.220 on the microwave approach, the automated theorem 00:06:10.220 --> 00:06:13.180 prover, you're trading the elegance of human 00:06:13.180 --> 00:06:16.480 intuition for the relentless, unforgiving speed 00:06:16.480 --> 00:06:19.800 of machine computation. OK. But if the machine 00:06:19.800 --> 00:06:22.459 is just blindly crunching through millions of 00:06:22.459 --> 00:06:25.449 permutations? What happens when it actually finds 00:06:25.449 --> 00:06:27.209 the proof? Well, that's the catch. Because I'd 00:06:27.209 --> 00:06:29.810 imagine the output is just a raw chronological 00:06:29.810 --> 00:06:33.129 log of every microscopic step it took. If an 00:06:33.129 --> 00:06:35.850 automated prover spits out a proof that's like 00:06:35.850 --> 00:06:38.889 10 ,000 pages of dense machine -generated logic 00:06:38.889 --> 00:06:42.050 and no human mind can actually comprehend the 00:06:42.050 --> 00:06:45.089 narrative of the math. Does that even count as 00:06:45.089 --> 00:06:48.170 a proof? That is an incredibly deep philosophical 00:06:48.170 --> 00:06:50.329 crisis within the field right now. Seriously? 00:06:50.509 --> 00:06:53.050 Yeah. Because if we define a proof as something 00:06:53.050 --> 00:06:56.050 that convinces a human mind of a truth, then 00:06:56.050 --> 00:06:59.350 a massive incomprehensible machine output totally 00:06:59.350 --> 00:07:01.189 fails that definition. Right. Because we don't 00:07:01.189 --> 00:07:03.509 get it. We can't read it. But if we define a 00:07:03.509 --> 00:07:05.850 proof strictly as a continuous chain of valid 00:07:05.850 --> 00:07:08.889 logical derivations, regardless of human readability, 00:07:09.310 --> 00:07:11.550 then the machine has succeeded perfectly. Wow. 00:07:11.550 --> 00:07:15.069 So we get the absolute truth. But we are completely 00:07:15.069 --> 00:07:17.470 alienated from understanding why it's true. The 00:07:17.470 --> 00:07:19.750 machine essentially becomes an oracle. Exactly. 00:07:20.069 --> 00:07:22.709 We ask it a question, it descends into this mathematical 00:07:22.709 --> 00:07:25.550 realm we can't perceive, and it hands us an answer 00:07:25.550 --> 00:07:29.129 we just have to blindly trust. That's wow. Okay, 00:07:29.269 --> 00:07:31.149 but here's where it gets really interesting. 00:07:32.520 --> 00:07:35.399 Because we contrast that microwave approach of 00:07:35.399 --> 00:07:38.620 automation with the second entirely different 00:07:38.620 --> 00:07:41.100 path listed on the page. The interactive path. 00:07:41.220 --> 00:07:43.079 Right. The text for the second path represents 00:07:43.079 --> 00:07:47.000 this jarring linguistic shift. We move completely 00:07:47.000 --> 00:07:49.800 away from the cold industrial tone of automated 00:07:49.800 --> 00:07:52.879 theorem prover. The second bullet point explicitly 00:07:52.879 --> 00:07:55.810 defines it as a proof assistant. an interactive 00:07:55.810 --> 00:07:58.329 theorem prover. The phrasing here is deliberate 00:07:58.329 --> 00:08:00.949 and highly revealing. The primary term is proof 00:08:00.949 --> 00:08:03.110 assistant, and interactive theorem prover is 00:08:03.110 --> 00:08:05.810 relegated to just an explanatory clause. Calling 00:08:05.810 --> 00:08:08.569 a highly complex piece of computational logic 00:08:08.569 --> 00:08:11.050 an assistant, I mean, that feels like a radical 00:08:11.050 --> 00:08:13.629 demotion. It does, yeah. It immediately re -centers 00:08:13.629 --> 00:08:16.029 the human being in the power dynamic. This sounds 00:08:16.029 --> 00:08:18.550 less like an autopilot and much more like a highly 00:08:18.550 --> 00:08:21.870 intelligent copilot or like a complex spell checker 00:08:21.870 --> 00:08:24.279 for pure logic. I think spell checker for logic 00:08:24.279 --> 00:08:28.319 is spot on. Using assistant explicitly centers 00:08:28.319 --> 00:08:31.959 the human as the primary prover. If the automated 00:08:31.959 --> 00:08:34.500 prover is a solitary probe launched into the 00:08:34.500 --> 00:08:37.600 dark maze, the interactive proof assistant is 00:08:37.600 --> 00:08:40.279 a highly advanced drone hovering right over your 00:08:40.279 --> 00:08:42.039 shoulder while you walk through the maze yourself. 00:08:42.440 --> 00:08:45.379 OK, so I am the boss in this scenario. I'm the 00:08:45.379 --> 00:08:47.039 one doing the actual proving. You're back in 00:08:47.039 --> 00:08:49.200 the driver's seat. But this raises an important 00:08:49.200 --> 00:08:52.580 question, right? Because interactive requires 00:08:52.580 --> 00:08:55.740 a dialogue between human intuition and machine 00:08:55.740 --> 00:08:58.480 verification. So if the interaction is essentially 00:08:58.480 --> 00:09:01.539 a feedback loop, and I am the one deciding which 00:09:01.539 --> 00:09:04.639 way to turn in this logical maze, what is the 00:09:04.639 --> 00:09:06.820 actual utility of the assistant, why not just 00:09:06.820 --> 00:09:09.679 do the math on a whiteboard? Because human intuition, 00:09:10.019 --> 00:09:12.480 while incredibly powerful for seeing the big 00:09:12.480 --> 00:09:15.259 picture, is notoriously sloppy when it comes 00:09:15.259 --> 00:09:17.539 to formal rigor. Oh, totally. We skip steps all 00:09:17.539 --> 00:09:20.059 the time. Exactly. We make massive leaps of logic. 00:09:20.179 --> 00:09:22.600 We write things like, it is trivial to see that. 00:09:22.759 --> 00:09:25.200 And then we skip 20 steps of tedious calculation 00:09:25.200 --> 00:09:27.399 because we intuitively know the result holds. 00:09:27.519 --> 00:09:29.460 Right. We drop negative signs. We make subtle 00:09:29.460 --> 00:09:32.360 assumptions. We build entire theoretical towers 00:09:32.360 --> 00:09:34.500 on slightly fractured foundations. So we are 00:09:34.500 --> 00:09:37.419 creative, but we are really flawed executioners. 00:09:37.610 --> 00:09:40.350 Right. So the interactive theorem prover exists 00:09:40.350 --> 00:09:43.250 to bridge that gap between your creative intuition 00:09:43.250 --> 00:09:47.009 and the absolute unforgiving rigidity of formal 00:09:47.009 --> 00:09:50.690 logic. OK, I see. You, the human, provide the 00:09:50.690 --> 00:09:53.850 strategy. You tell the assistant, I want to prove 00:09:53.850 --> 00:09:57.009 this lemma by using this specific method. You 00:09:57.009 --> 00:09:59.169 provide the structural outline, and the machine's 00:09:59.169 --> 00:10:01.909 job is to audit every single microstep of your 00:10:01.909 --> 00:10:04.570 proposed path. It's basically acting as a structural 00:10:04.570 --> 00:10:07.710 engineer checking an architect's blueprint. Like, 00:10:08.169 --> 00:10:11.450 I draw the beautiful sweeping skyscraper and 00:10:11.450 --> 00:10:13.629 the machine tells me if the steel beams in my 00:10:13.629 --> 00:10:15.389 drawing can actually support the weight without 00:10:15.389 --> 00:10:18.070 collapsing. Perfect analogy. And if they can't, 00:10:18.090 --> 00:10:20.830 the machine violently stops you. It refuses to 00:10:20.830 --> 00:10:22.950 let you proceed to the next step of the proof 00:10:22.950 --> 00:10:24.889 until you have logically satisfied the current 00:10:24.889 --> 00:10:28.169 one. Oh, wow. It forces you to expand your human 00:10:28.169 --> 00:10:30.950 leaps of intuition into exhaustive, watertight 00:10:30.950 --> 00:10:32.889 sequences of formal code. I mean, that sounds 00:10:32.889 --> 00:10:35.350 like an incredibly frustrating, pedantic process. 00:10:35.590 --> 00:10:39.090 It is deeply demanding. But the result is a proof 00:10:39.090 --> 00:10:41.750 that has the strategic elegance of a human mind 00:10:41.750 --> 00:10:44.990 verified with the flawless mechanical certainty 00:10:44.990 --> 00:10:47.600 of a computer. But here is where the text of 00:10:47.600 --> 00:10:51.019 this disambiguation page provokes a serious challenge. 00:10:51.700 --> 00:10:54.899 If the interactive path requires the human to 00:10:54.899 --> 00:10:57.779 essentially spoon -feed the strategy to the machine, 00:10:59.080 --> 00:11:01.850 isn't calling the machine a theorem prover? a 00:11:01.850 --> 00:11:03.850 bit fraudulent. How do you mean? Well, it's not 00:11:03.850 --> 00:11:05.809 actually proving anything, right? It's just checking 00:11:05.809 --> 00:11:08.070 my work. It's a syntax checker for mathematics. 00:11:08.549 --> 00:11:11.210 Ah, well that tension is exactly why the text 00:11:11.210 --> 00:11:14.429 primarily labels it a poof assistant and only 00:11:14.429 --> 00:11:17.450 secondarily an interactive theorem prover. It 00:11:17.450 --> 00:11:20.070 acknowledges the blurring of the lines. Because 00:11:20.070 --> 00:11:23.070 in highly complex, modern proofs, the human mind 00:11:23.070 --> 00:11:25.909 physically cannot track the millions of intersecting 00:11:25.909 --> 00:11:29.090 variables required to ensure no logical fallacies 00:11:29.090 --> 00:11:31.049 have occurred. So we literally couldn't do it 00:11:31.049 --> 00:11:33.370 without them. Exactly. Without the machine's 00:11:33.370 --> 00:11:35.570 interactive verification, the human proof wouldn't 00:11:35.570 --> 00:11:37.909 be structurally sound. So the act of proving 00:11:37.909 --> 00:11:40.690 becomes a symbiotic event. Neither party can 00:11:40.690 --> 00:11:43.049 really claim full ownership of the final truth. 00:11:43.169 --> 00:11:46.350 We have to rely on a hybrid cognition. Yes. Which 00:11:46.350 --> 00:11:48.830 really makes the existence of this disambiguation 00:11:48.830 --> 00:11:52.090 page so much more profound. It isn't just routing 00:11:52.090 --> 00:11:55.769 web traffic. It is enforcing a strict taxonomy 00:11:55.769 --> 00:11:58.809 on how humanity chooses to grapple with the limits 00:11:58.809 --> 00:12:02.549 of its own intelligence. Do we hand over total 00:12:02.549 --> 00:12:05.230 autonomy and accept incomprehensible answers? 00:12:05.950 --> 00:12:08.809 Or do we retain control and submit ourselves 00:12:08.809 --> 00:12:12.549 to the exhausting, pedantic oversight of a machine 00:12:12.549 --> 00:12:15.250 auditor? And the fact that Wikipedia refuses 00:12:15.250 --> 00:12:17.529 to let these two concepts blend into a single 00:12:17.529 --> 00:12:20.049 article speaks of how deeply entrenched these 00:12:20.049 --> 00:12:22.250 two philosophical camps are within the fields 00:12:22.250 --> 00:12:24.289 of computer science and mathematics. Absolutely. 00:12:24.590 --> 00:12:26.470 And that entrenchment actually brings us to the 00:12:26.470 --> 00:12:29.710 meta -context, the map itself. The physical metadata 00:12:29.710 --> 00:12:31.649 of the Wikipedia page. Right, the administrative 00:12:31.649 --> 00:12:33.690 text hidden at the very bottom of the source. 00:12:33.789 --> 00:12:35.610 Because when you look at the specific details 00:12:35.610 --> 00:12:38.110 provided, the timestamps and the licensing, it 00:12:38.110 --> 00:12:40.350 paints a very specific picture of the current 00:12:40.350 --> 00:12:42.710 state of this technological landscape. Invisible 00:12:42.710 --> 00:12:44.970 staff -holding of the information, yeah. Let's 00:12:44.970 --> 00:12:48.850 look at it, the text notes. This page was last 00:12:48.850 --> 00:12:54.720 edited on 30 December 2019 at 15 .43 UTC. It 00:12:54.720 --> 00:12:57.700 also lists an old ID and notes it is licensed 00:12:57.700 --> 00:13:00.399 under a Creative Commons Attribution ShareAlike 00:13:00.399 --> 00:13:03.019 4 .0 license. The timestamp is definitely the 00:13:03.019 --> 00:13:05.100 most critical piece of data there. For sure. 00:13:05.460 --> 00:13:07.799 Because, okay, my immediate reaction to seeing 00:13:07.799 --> 00:13:10.740 a date from late 2019 on an article about artificial 00:13:10.740 --> 00:13:12.820 intelligence and computation is that the page 00:13:12.820 --> 00:13:15.600 is dead. Right. Because three years in AI is... 00:13:15.600 --> 00:13:17.759 A completely different geological era. In the 00:13:17.759 --> 00:13:19.679 fast -paced world of technology and mathematics, 00:13:20.159 --> 00:13:22.659 fields experience paradigm -shifting breakthroughs 00:13:22.659 --> 00:13:24.980 on a monthly basis. So what does this all mean? 00:13:25.179 --> 00:13:26.960 The fact that this page hasn't been touched since 00:13:26.960 --> 00:13:30.159 late 2019. Does that mean this dichotomy is entirely 00:13:30.159 --> 00:13:32.559 settled? Or has it just been abandoned by the 00:13:32.559 --> 00:13:34.860 community? I would interpret that silence entirely 00:13:34.860 --> 00:13:37.490 differently than neglect. Really? Yeah. In the 00:13:37.490 --> 00:13:39.610 architecture of Wikipedia, neglect usually happens 00:13:39.610 --> 00:13:42.230 on fringe, obscure topics. But theorem prover 00:13:42.230 --> 00:13:44.549 is a foundational node. It routes traffic for 00:13:44.549 --> 00:13:46.509 some of the most heavily researched and fiercely 00:13:46.509 --> 00:13:48.710 debated topics in modern computer science. So 00:13:48.710 --> 00:13:51.389 it gets a lot of eyes. Exactly. So if this page 00:13:51.389 --> 00:13:53.570 is sitting frozen, it isn't because of neglect. 00:13:53.929 --> 00:13:56.950 It's because it represents a rock solid, universally 00:13:56.950 --> 00:13:59.690 agreed upon structure in the field. So it's not 00:13:59.690 --> 00:14:02.490 a dead page. It's a settled treaty. Beautifully 00:14:02.490 --> 00:14:05.769 put. The edit histories of Wikipedia pages dealing 00:14:05.769 --> 00:14:08.759 with active, evolving technology are usually 00:14:08.759 --> 00:14:11.679 chaotic battlegrounds. Definitions are constantly 00:14:11.679 --> 00:14:13.980 rewritten, boundaries are blurred, concepts are 00:14:13.980 --> 00:14:17.460 merged. But amid all that churn, this specific 00:14:17.460 --> 00:14:19.860 fork in the road has remained crystalline and 00:14:19.860 --> 00:14:23.019 static for years. There is no active debate fighting 00:14:23.019 --> 00:14:25.500 to change this specific categorization on the 00:14:25.500 --> 00:14:27.639 platform. Nobody is fighting to redefine the 00:14:27.639 --> 00:14:30.320 landscape. The community agrees that the binary 00:14:30.320 --> 00:14:33.279 division between automated and interactive is 00:14:33.279 --> 00:14:35.690 permanent. Right. It implies that despite Despite 00:14:35.690 --> 00:14:38.490 all the rapid advancements in AI, these two distinct 00:14:38.490 --> 00:14:41.610 philosophies have not merged. The automated provers 00:14:41.610 --> 00:14:43.710 haven't suddenly gained human -like intuition, 00:14:43.909 --> 00:14:46.029 and the interactive assistants haven't eliminated 00:14:46.029 --> 00:14:49.090 the need for human strategic oversight. The border 00:14:49.090 --> 00:14:51.070 between these two countries of thought is firmly 00:14:51.070 --> 00:14:55.169 established. Firmly. And the text even includes 00:14:55.169 --> 00:14:57.629 a line about the interface itself warning the 00:14:57.629 --> 00:15:00.730 user. If an internal link incorrectly led you 00:15:00.730 --> 00:15:03.250 here, you may wish to change the link to point 00:15:03.250 --> 00:15:05.870 directly to the intended article. That is the 00:15:05.870 --> 00:15:08.289 system actively demanding precision from its 00:15:08.289 --> 00:15:10.409 own contributors. Exactly. Because it shows that 00:15:10.409 --> 00:15:12.230 even the people writing the encyclopedia get 00:15:12.230 --> 00:15:15.529 lazy. A writer drafting an article on mathematics 00:15:15.529 --> 00:15:18.950 might just hyperlink the phrase theorem prover. 00:15:19.100 --> 00:15:21.580 treating it as a generic catch -all tool. But 00:15:21.580 --> 00:15:24.620 this disambiguation page acts as a strict editor, 00:15:25.019 --> 00:15:27.340 catching that lazy link and forcing the writer 00:15:27.340 --> 00:15:30.220 to clarify. It rejects the ambiguity completely. 00:15:30.759 --> 00:15:33.000 It forces the human to acknowledge whether the 00:15:33.000 --> 00:15:34.980 tool they're writing about is a solitary machine 00:15:34.980 --> 00:15:37.899 navigating the dark or a human -machine symbiosis. 00:15:38.220 --> 00:15:41.019 And the inclusion of that Creative Commons attribution 00:15:41.019 --> 00:15:43.480 share -alike license is also a fascinating detail 00:15:43.480 --> 00:15:45.820 to ponder here. The irony is pretty hard to miss. 00:15:46.240 --> 00:15:50.220 It really is. We're talking about highly specialized, 00:15:50.759 --> 00:15:54.059 rigidly formal tools designed to uncover absolute 00:15:54.250 --> 00:15:57.730 objective mathematical truths. Yet the framework 00:15:57.730 --> 00:16:00.629 we use to define and categorize these tools is 00:16:00.629 --> 00:16:03.330 governed by a collaborative, open source human 00:16:03.330 --> 00:16:06.230 license. The objective truth is categorized by 00:16:06.230 --> 00:16:09.289 a messy, subjective human consensus. It highlights 00:16:09.289 --> 00:16:11.690 our ultimate reliance on human community. The 00:16:11.690 --> 00:16:13.690 tools themselves might be cold and mechanical, 00:16:13.830 --> 00:16:16.049 but the map we use to understand them is collectively 00:16:16.049 --> 00:16:18.909 drawn and maintained by us. So to summarize our 00:16:18.909 --> 00:16:21.590 journey today, we've taken a tiny snippet of 00:16:21.590 --> 00:16:24.169 routing text, basically a digital sticky note, 00:16:24.789 --> 00:16:27.269 and realized it outlines the two major ways humanity 00:16:27.269 --> 00:16:29.850 teams up with machines to solve complex logic. 00:16:30.549 --> 00:16:32.289 Either by handing over the keys completely with 00:16:32.289 --> 00:16:34.629 the automated path, or by working together with 00:16:34.629 --> 00:16:37.230 the Interactor Proof Assistant. And for you listening, 00:16:37.549 --> 00:16:39.269 this matters because it reflects our broader 00:16:39.269 --> 00:16:41.549 relationship with technology, right? And how 00:16:41.549 --> 00:16:43.450 much control we are willing to hand over on a 00:16:43.450 --> 00:16:46.570 daily basis. It really scales far beyond formal 00:16:46.570 --> 00:16:49.690 logic. The split mirrors almost every modern 00:16:49.690 --> 00:16:52.309 debate about automation, artificial intelligence, 00:16:52.509 --> 00:16:55.909 and human obsolescence. Are we building machines 00:16:55.909 --> 00:16:59.730 to replace our cognition or augment it? The disambiguation 00:16:59.730 --> 00:17:02.070 page tells us we're actively building both futures 00:17:02.070 --> 00:17:04.849 simultaneously, but keeping them strictly segregated. 00:17:05.099 --> 00:17:08.559 Which leaves us with a necessary and kind of 00:17:08.559 --> 00:17:11.279 provocative question about the future trajectory 00:17:11.279 --> 00:17:13.759 of that second path. The interactive path. Right. 00:17:13.960 --> 00:17:16.180 Because the entire premise of the interactive 00:17:16.180 --> 00:17:19.359 model relies on a power dynamic embedded in the 00:17:19.359 --> 00:17:21.559 word assistant. The human acts as the visionary 00:17:21.559 --> 00:17:23.680 and the machine is just the pedantic auditor. 00:17:23.980 --> 00:17:27.559 Exactly. But as computational power grows, that 00:17:27.559 --> 00:17:30.849 line begins to waver. Think about it. If an interactive 00:17:30.849 --> 00:17:33.670 theorem prover becomes advanced enough to analyze 00:17:33.670 --> 00:17:36.509 your logical leaps and then start suggesting 00:17:36.509 --> 00:17:38.450 the next step before you even think of it. Oh, 00:17:38.450 --> 00:17:41.109 wow. Right. If it begins identifying structural 00:17:41.109 --> 00:17:43.710 flaws in your strategy and automatically rewriting 00:17:43.710 --> 00:17:45.769 large sections of the proof to correct them, 00:17:46.190 --> 00:17:48.829 at what point of technological advancement does 00:17:48.829 --> 00:17:51.369 the assistant end up doing so much of the heavy 00:17:51.369 --> 00:17:54.309 lifting that the human actually becomes the assistant 00:17:54.309 --> 00:17:57.210 to the machine? That is a staggering realization. 00:17:57.470 --> 00:18:00.529 If the machine is predicting my intuition, verifying 00:18:00.529 --> 00:18:02.670 the math, and steering me away from dead ends, 00:18:03.549 --> 00:18:06.349 am I still the pilot? Or have I just become the 00:18:06.349 --> 00:18:08.789 biological input device for the machine's logic? 00:18:09.009 --> 00:18:11.549 Who is really proving the theorem? When the interactive 00:18:11.549 --> 00:18:14.190 tool becomes predictive rather than just reactive, 00:18:14.369 --> 00:18:16.210 we really have to ask ourselves that question. 00:18:16.589 --> 00:18:18.670 An incredible thought to leave on. So the next 00:18:18.670 --> 00:18:20.670 time you are searching for a concept online and 00:18:20.670 --> 00:18:23.109 you hit a disambiguation page, don't just quickly 00:18:23.109 --> 00:18:26.009 click the link you need and move on. Take a second 00:18:26.009 --> 00:18:28.029 to look at the blank space between those options. 00:18:28.349 --> 00:18:30.650 Think about why the human consensus decided those 00:18:30.650 --> 00:18:34.009 ideas could no longer share the same room. The 00:18:34.009 --> 00:18:35.809 deepest insights about how we understand the 00:18:35.809 --> 00:18:38.589 world aren't always found in the massive sprawling 00:18:38.589 --> 00:18:41.509 articles. Sometimes, they're hiding right there 00:18:41.509 --> 00:18:43.769 in the structural cracks, forcing you to choose 00:18:43.769 --> 00:18:44.170 a path.