WEBVTT 00:00:00.000 --> 00:00:02.560 You know, when you look at a modern city skyline, 00:00:03.060 --> 00:00:05.099 your eyes are just immediately drawn to the glass, 00:00:05.259 --> 00:00:08.419 right? The steel, the sweeping architectural 00:00:08.419 --> 00:00:10.740 lines of those massive skyscrapers. Oh, yeah. 00:00:10.839 --> 00:00:12.539 You admire the finished product. It's beautiful. 00:00:12.880 --> 00:00:15.419 Exactly. But what you really don't think about, 00:00:15.800 --> 00:00:17.820 what is completely invisible to you as you walk 00:00:17.820 --> 00:00:21.539 down the street, are the millions of tons of 00:00:21.539 --> 00:00:25.460 rebar and poured concrete buried deep underground. 00:00:25.579 --> 00:00:28.829 Right. The foundation. The foundation is rarely 00:00:28.829 --> 00:00:30.589 glamorous. It's definitely not the thing they 00:00:30.589 --> 00:00:32.369 put on the tourism postcards. No, definitely 00:00:32.369 --> 00:00:34.450 not. But if it cracks, the whole structure comes 00:00:34.450 --> 00:00:37.329 down. And today we are taking a deep dive into 00:00:37.329 --> 00:00:39.689 the digital equivalent of that buried concrete. 00:00:40.429 --> 00:00:43.670 Welcome to another deep dive. We spent time poring 00:00:43.670 --> 00:00:45.810 through an incredibly comprehensive Wikipedia 00:00:45.810 --> 00:00:48.270 article on open source software. And the mission 00:00:48.270 --> 00:00:51.820 today is to, well... to figure out a modern miracle. 00:00:52.000 --> 00:00:54.579 It really is a miracle. It is. We want to uncover 00:00:54.579 --> 00:00:56.979 how software that is given away entirely for 00:00:56.979 --> 00:00:59.479 free, like no price tag, no strings, somehow 00:00:59.479 --> 00:01:02.039 became the unseen trillion dollar backbone of 00:01:02.039 --> 00:01:04.739 the entire digital world. Yeah, and it really 00:01:04.739 --> 00:01:07.459 requires a massive conceptual shift to understand 00:01:07.459 --> 00:01:09.879 because, you know, most of the time you probably 00:01:09.879 --> 00:01:12.620 think of software as a physical product you buy 00:01:12.620 --> 00:01:17.090 in a box. or maybe a closed app ecosystem that 00:01:17.090 --> 00:01:20.329 a specific massive tech company owns and tightly 00:01:20.329 --> 00:01:22.890 controls. Right, like buying a video game or 00:01:22.890 --> 00:01:25.950 a word processor. Exactly. But the reality is 00:01:25.950 --> 00:01:28.250 that the digital world you interact with every 00:01:28.250 --> 00:01:32.109 single day actually runs on a massive decentralized 00:01:32.109 --> 00:01:35.189 global system of open collaboration. And when 00:01:35.189 --> 00:01:37.150 we say trillion dollar backbone, that is not 00:01:37.150 --> 00:01:39.590 an exaggeration. Right out of the gate, the data 00:01:39.590 --> 00:01:42.090 we're looking at drops just a staggering number. 00:01:42.109 --> 00:01:46.069 Oh, the 2020. Yes. According to a 2024 estimate, 00:01:46.689 --> 00:01:49.650 open source software brings an estimated $8 .8 00:01:49.650 --> 00:01:53.609 trillion in value to firms. $8 .8 trillion. It's 00:01:53.609 --> 00:01:56.189 mind boggling. It's wild. And if companies couldn't 00:01:56.189 --> 00:01:57.849 just plug and play this free code, they would 00:01:57.849 --> 00:01:59.609 have to spend three and a half times more than 00:01:59.609 --> 00:02:01.489 they currently do, literally just to keep the 00:02:01.489 --> 00:02:03.409 lights on. It's an astronomical figure. And to 00:02:03.409 --> 00:02:05.670 understand how this $8 .8 trillion ecosystem 00:02:05.670 --> 00:02:08.969 even functions today, we kind of have to look 00:02:08.969 --> 00:02:11.250 backward. Back to the beginning. Right. Because 00:02:11.250 --> 00:02:14.189 this entire phenomenon didn't start out as some, 00:02:14.189 --> 00:02:17.090 you know, savvy boardroom business strategy. 00:02:17.610 --> 00:02:19.629 It actually started out as a moral rebellion. 00:02:19.830 --> 00:02:22.490 OK, let's unpack this because the history here 00:02:22.490 --> 00:02:25.770 is. It's just fascinating. The research notes 00:02:25.770 --> 00:02:27.629 that back in the early days of computing, we're 00:02:27.629 --> 00:02:30.990 talking the 1950s and 60s, academic sharing of 00:02:30.990 --> 00:02:33.909 code was just the default. It was just how things 00:02:33.909 --> 00:02:35.849 were done. Yeah, programmers shared their work 00:02:35.849 --> 00:02:37.889 freely to learn from each other and build better 00:02:37.889 --> 00:02:42.189 machines. But then by the 1970s and 80s, a commercial 00:02:42.189 --> 00:02:45.270 software industry really emerged, and companies 00:02:45.270 --> 00:02:47.349 started locking things down. They realized there 00:02:47.349 --> 00:02:50.150 was money to be made. Exactly. They moved to 00:02:50.150 --> 00:02:52.689 what we call proprietary models, where they actually 00:02:52.689 --> 00:02:54.930 hid the underlying source code from the people 00:02:54.930 --> 00:02:57.990 buying it. And that lockdown completely fractured 00:02:57.990 --> 00:03:00.610 the culture of programming, which introduces 00:03:00.610 --> 00:03:03.009 us to a guy named Richard Stallman. Right, the 00:03:03.009 --> 00:03:05.789 rebel. Yeah. In 1984, Stallman took a massive 00:03:05.789 --> 00:03:08.789 stand. He actually quit his job at MIT. He was 00:03:08.789 --> 00:03:11.050 specifically because the collaborative programmer 00:03:11.050 --> 00:03:13.789 culture in his lab was being completely suffocated 00:03:13.789 --> 00:03:16.169 by this new proprietary lockdown. He couldn't 00:03:16.169 --> 00:03:18.689 handle the restrictions. Exactly. He couldn't 00:03:18.689 --> 00:03:20.310 share his code. He couldn't tinker with other 00:03:20.310 --> 00:03:23.349 people's code to fix basic problems. And he felt 00:03:23.349 --> 00:03:26.729 this restriction was fundamentally, morally wrong. 00:03:27.310 --> 00:03:30.870 So he decides to start a crusade. In 1985, he 00:03:30.870 --> 00:03:33.469 founds the Free Software Foundation and creates 00:03:33.469 --> 00:03:36.430 an operating system called GNU. Which is a huge 00:03:36.430 --> 00:03:38.330 undertaking. Oh, absolutely. And just for anyone 00:03:38.330 --> 00:03:40.430 listening who isn't, like a computer science 00:03:40.430 --> 00:03:43.050 major, an operating system is essentially the 00:03:43.050 --> 00:03:46.090 core set of instructions that tells your computer's 00:03:46.090 --> 00:03:48.509 hardware how to actually run your programs. It's 00:03:48.509 --> 00:03:50.909 the most foundational layer. Right, the concrete. 00:03:51.110 --> 00:03:54.479 Yes. And Stallman had a very specific ethos for 00:03:54.479 --> 00:03:57.259 this foundation. He famously argued that software 00:03:57.259 --> 00:04:00.780 should be free, as in free speech, not free beer. 00:04:01.020 --> 00:04:03.340 That phrase is just the perfect distillation 00:04:03.340 --> 00:04:05.919 of his philosophy. He meant it's entirely about 00:04:05.919 --> 00:04:07.900 the freedom to study the code, change the code, 00:04:08.020 --> 00:04:09.919 and distribute the code. It wasn't necessarily 00:04:09.919 --> 00:04:12.020 about the price tag at the register. The liberty 00:04:12.020 --> 00:04:14.780 to do what you want with it. Exactly. To legally 00:04:14.780 --> 00:04:17.300 enforce this philosophy, he created a fascinating 00:04:17.300 --> 00:04:20.639 legal mechanism called copy left, and he formalized 00:04:20.639 --> 00:04:22.720 it in something called the GNU General Public 00:04:22.720 --> 00:04:26.360 License, or GPL. Copy left? I mean, it's a great 00:04:26.360 --> 00:04:28.720 play on words, but how does that actually work 00:04:28.720 --> 00:04:31.620 in practice? Well, think of traditional copyright 00:04:31.620 --> 00:04:34.120 as a way to restrict who can use your work. Yeah, 00:04:34.259 --> 00:04:37.040 sure. All rights reserved. Right. CopyLeft flips 00:04:37.040 --> 00:04:39.939 that completely on its head. It legally requires 00:04:39.939 --> 00:04:42.699 that if you take someone's free software, use 00:04:42.699 --> 00:04:46.240 it, and modify it, you are legally bound to share 00:04:46.240 --> 00:04:49.100 your newly modified version under those exact 00:04:49.100 --> 00:04:51.500 same free terms. Oh, wow. So you can't just take 00:04:51.500 --> 00:04:53.879 it and hide it. Nope. You can't take free code, 00:04:54.100 --> 00:04:56.339 improve it, and then lock it away. It's basically 00:04:56.339 --> 00:04:59.120 a viral license designed to legally guarantee 00:04:59.120 --> 00:05:01.540 that software remains free forever. That makes 00:05:01.540 --> 00:05:05.139 perfect sense for a moral crusade. But reading 00:05:05.139 --> 00:05:07.019 through the history, there was a massive branding 00:05:07.019 --> 00:05:10.259 problem here. The word free absolutely terrified 00:05:10.259 --> 00:05:12.620 the business world in the 90s. Oh, yeah. Completely 00:05:12.620 --> 00:05:14.959 alienated them. Because businesses hear free, 00:05:15.100 --> 00:05:18.639 and they immediately think unreliable, unsupported, 00:05:18.800 --> 00:05:22.129 or worst of all, anti -profit. You did. Corporate 00:05:22.129 --> 00:05:23.870 executives just couldn't wrap their heads around 00:05:23.870 --> 00:05:25.790 relying on something that didn't have a massive 00:05:25.790 --> 00:05:28.350 company backing it with a warranty. So Stallman 00:05:28.350 --> 00:05:30.970 is essentially like a master chef, insisting 00:05:30.970 --> 00:05:34.170 every single diner gets his secret recipes, along 00:05:34.170 --> 00:05:36.230 with a philosophical lecture about the moral 00:05:36.230 --> 00:05:38.949 right to tweak those recipes. Right. But the 00:05:38.949 --> 00:05:41.610 business world just wanted the free food without 00:05:41.610 --> 00:05:43.930 the lecture. They wanted the benefits of shared 00:05:43.930 --> 00:05:47.029 code without the activism. Is that why open source 00:05:47.029 --> 00:05:49.550 kind of won out? That is a perfect way to look 00:05:49.550 --> 00:05:52.129 at it. And that fundamental disconnect is what 00:05:52.129 --> 00:05:56.189 caused a massive schism. In 1998, at a strategy 00:05:56.189 --> 00:05:58.949 session held by free software supporters, a woman 00:05:58.949 --> 00:06:01.269 named Christine Peterson suggested a new term, 00:06:01.870 --> 00:06:04.990 open source. A total rebrand. Exactly. They were 00:06:04.990 --> 00:06:07.069 actively looking for a term that would rebrand 00:06:07.069 --> 00:06:09.449 the entire movement. They wanted to drop the 00:06:09.449 --> 00:06:12.529 moral activism and highlight the pragmatic, highly 00:06:12.529 --> 00:06:15.230 efficient Business -friendly potential of sharing 00:06:15.230 --> 00:06:17.110 code. Just make it palatable for the boardroom. 00:06:17.290 --> 00:06:19.529 Yeah. And this single meeting led to the creation 00:06:19.529 --> 00:06:22.509 of the Open Source Initiative, or OSI. And Stallman 00:06:22.509 --> 00:06:25.709 absolutely hated this. He detested it. And he 00:06:25.709 --> 00:06:27.889 still does today. Yeah. Stallman views the term 00:06:27.889 --> 00:06:30.230 open source as a total abandonment of the moral 00:06:30.230 --> 00:06:32.829 fight. Really? How so? Well, for instance, he 00:06:32.829 --> 00:06:34.810 points out that under the open source methodology, 00:06:35.389 --> 00:06:37.730 you can develop highly restrictive software things 00:06:37.730 --> 00:06:41.129 like digital rights management or DRM. Oh, the 00:06:41.129 --> 00:06:43.790 digital locks on file? Exactly. Yeah. Locks that 00:06:43.790 --> 00:06:46.110 dictate how you can use a file. You can build 00:06:46.110 --> 00:06:48.790 those locks using open source methods, which 00:06:48.790 --> 00:06:51.629 completely violates Stallman's core concept of 00:06:51.629 --> 00:06:54.430 user freedom. So there's this deep philosophical 00:06:54.430 --> 00:06:57.490 fault line. You have free software, which is 00:06:57.490 --> 00:06:59.829 a rigid social and moral stance, and then open 00:06:59.829 --> 00:07:02.069 source, which is just a practical way to develop 00:07:02.069 --> 00:07:04.610 code faster. Right. And today, the community 00:07:04.610 --> 00:07:08.110 often blends them into the acronym FOSS free 00:07:08.110 --> 00:07:10.769 and open source software just to sort of keep 00:07:10.769 --> 00:07:13.810 the peace. Exactly. But semantic battles aside, 00:07:13.970 --> 00:07:16.439 the reality is that the business world embraced 00:07:16.439 --> 00:07:19.079 the open source label and it completely took 00:07:19.079 --> 00:07:21.139 off. Well, let's talk about how it took off because 00:07:21.139 --> 00:07:23.060 the actual mechanics of building the software 00:07:23.060 --> 00:07:25.360 is mind boggling. It really is. Here's where 00:07:25.360 --> 00:07:27.939 it gets really interesting because on paper, 00:07:28.560 --> 00:07:30.600 allowing anyone on the internet to just download 00:07:30.600 --> 00:07:33.180 your core code, mess with it and submit changes 00:07:33.180 --> 00:07:36.600 sounds like a recipe for absolute disaster. Complete 00:07:36.600 --> 00:07:39.399 chaos. How do you build an operating system with 00:07:39.399 --> 00:07:42.180 thousands of unpaid, unmanaged strangers without 00:07:42.180 --> 00:07:45.670 it turning into complete digital rubble? To explain 00:07:45.670 --> 00:07:48.110 that mechanical miracle, our research highlights 00:07:48.110 --> 00:07:51.930 a highly influential 1997 essay. It's by Eric 00:07:51.930 --> 00:07:54.410 S. Raymond, and it's called The Cathedral and 00:07:54.410 --> 00:07:58.129 the Bazaar. Okay. Raymond used two very distinct 00:07:58.129 --> 00:08:00.769 architectural metaphors to explain how software 00:08:00.769 --> 00:08:02.889 gets made. The cathedral being the traditional 00:08:02.889 --> 00:08:05.629 proprietary way. Yes. In traditional software 00:08:05.629 --> 00:08:08.279 engineering, The cathedral, everything is highly 00:08:08.279 --> 00:08:11.879 centralized. You have a few elite, highly skilled 00:08:11.879 --> 00:08:14.660 architects who carefully plan every single feature 00:08:14.660 --> 00:08:17.420 in total isolation. Behind closed doors. Exactly. 00:08:17.759 --> 00:08:19.899 The rules are strictly defined. Access to the 00:08:19.899 --> 00:08:22.319 code is heavily restricted behind corporate firewalls. 00:08:22.579 --> 00:08:24.480 And the final product is only unveiled to the 00:08:24.480 --> 00:08:26.540 public when they deem it completely finished 00:08:26.540 --> 00:08:29.480 and perfect. An open source is the bazaar? The 00:08:29.480 --> 00:08:31.720 bazaar is the exact opposite. It's decentralized, 00:08:31.839 --> 00:08:33.720 it looks entirely unstructured from the outside, 00:08:34.059 --> 00:08:36.240 roles burr together. Just people yelling and 00:08:36.240 --> 00:08:39.240 trading in an open market. Yeah. But the most 00:08:39.240 --> 00:08:41.419 crucial distinction is how it treats the people 00:08:41.419 --> 00:08:44.299 actually using the software. In the cathedral, 00:08:44.740 --> 00:08:47.940 users are just consumers. In the bazaar, users 00:08:47.940 --> 00:08:50.899 are actively treated as co -developers. Oh, wow. 00:08:51.000 --> 00:08:53.480 They are encouraged to download the beta, look 00:08:53.480 --> 00:08:56.299 at the code, submit their own fixes, and aggressively 00:08:56.299 --> 00:08:59.159 report bugs. But wait, my instinct is to push 00:08:59.159 --> 00:09:01.779 back on this entirely. If I have thousands of 00:09:01.779 --> 00:09:04.100 random people throwing code at a wall, how do 00:09:04.100 --> 00:09:06.299 they not just constantly overwrite each other's 00:09:06.299 --> 00:09:08.440 work and break the system faster? Sounds like 00:09:08.440 --> 00:09:11.039 a nightmare, right? It really does. The text 00:09:11.039 --> 00:09:13.360 talks about the bizarre model encouraging early 00:09:13.360 --> 00:09:16.980 and frequent releases of buggy software. Releasing 00:09:16.980 --> 00:09:19.759 buggy software to a chaotic crowd sounds terrible. 00:09:20.279 --> 00:09:22.259 How does that actually result in a more reliable 00:09:22.259 --> 00:09:24.460 end product? What's fascinating here is that 00:09:24.460 --> 00:09:26.320 it seems counterintuitive until you understand 00:09:26.320 --> 00:09:29.120 the underlying philosophy. Raymond dubbed it 00:09:29.120 --> 00:09:32.080 Linus's Law, named after Linus Torvalds, the 00:09:32.080 --> 00:09:33.799 creator of the Linux operating system. Okay, 00:09:33.860 --> 00:09:36.539 what's the law? The law states, given enough 00:09:36.539 --> 00:09:39.779 eyeballs, all bugs are shallow. Meaning, if enough 00:09:39.779 --> 00:09:42.580 people are staring at a problem, someone is bound 00:09:42.580 --> 00:09:45.960 to spot the solution. Exactly. Think about how 00:09:45.960 --> 00:09:48.600 a cathedral tests software. You have a small, 00:09:49.000 --> 00:09:51.840 dedicated, quality assurance team sitting in 00:09:51.840 --> 00:09:55.500 an office trying to guess and anticipate every 00:09:55.500 --> 00:09:57.879 possible way a user might break the program. 00:09:58.259 --> 00:10:00.259 Which is impossible. Right. They're limited by 00:10:00.259 --> 00:10:01.919 their own imagination and their own hardware. 00:10:01.940 --> 00:10:05.000 Right. But in the bizarre, when you release code 00:10:05.000 --> 00:10:09.159 early and frequently, Every single user's computer 00:10:09.159 --> 00:10:12.960 becomes a brand new, highly unique testing environment. 00:10:13.000 --> 00:10:15.460 Oh, I see. You instantly have millions of diverse 00:10:15.460 --> 00:10:17.740 machines running different hardware with different 00:10:17.740 --> 00:10:20.419 software configurations, acting as a massive, 00:10:20.659 --> 00:10:23.200 globally distributed QA team. So they find the 00:10:23.200 --> 00:10:26.159 blind spots instantly. Exactly. If there is a 00:10:26.159 --> 00:10:28.440 bug that only triggers when a specific graphics 00:10:28.440 --> 00:10:31.139 card interacts with a specific keystroke, one 00:10:31.139 --> 00:10:33.419 of those millions of users will find it instantly. 00:10:33.690 --> 00:10:36.710 So the sheer volume of diverse users mathematically 00:10:36.710 --> 00:10:39.549 overwhelms the small team of experts. But what 00:10:39.549 --> 00:10:41.690 about the chaos of actually combining all their 00:10:41.690 --> 00:10:43.850 fixes? That's where frequent integration comes 00:10:43.850 --> 00:10:45.769 in. They don't just dump all the code into one 00:10:45.769 --> 00:10:48.029 bucket. They use version control systems. Like 00:10:48.029 --> 00:10:50.110 a filter? Yeah, think of it like a master blueprint. 00:10:50.690 --> 00:10:52.950 You don't let anyone draw directly on the master. 00:10:53.370 --> 00:10:55.990 People make a copy, draw their suggested changes, 00:10:56.590 --> 00:10:59.250 and submit that copy for review. Ah, so someone 00:10:59.250 --> 00:11:02.350 is still steering the ship. Right. The core maintainers 00:11:02.350 --> 00:11:05.309 review those small bite -sized changes constantly 00:11:05.309 --> 00:11:08.090 and merge them in. By doing this continuously, 00:11:08.470 --> 00:11:11.029 they avoid the nightmare of trying to untangle 00:11:11.029 --> 00:11:14.070 thousands of overlapping bugs at the very end 00:11:14.070 --> 00:11:16.490 of a five -year project. The chaos is actually 00:11:16.490 --> 00:11:19.210 a highly calibrated filter. It crowdsources the 00:11:19.210 --> 00:11:21.210 peer review. That is brilliant. It really is. 00:11:21.289 --> 00:11:23.509 And it perfectly explains why the massive corporate 00:11:23.509 --> 00:11:25.629 cathedrals eventually realized they couldn't 00:11:25.629 --> 00:11:27.629 keep up. I mean, they simply couldn't hire enough 00:11:27.629 --> 00:11:30.090 testers to beat the bizarre, so they had to join 00:11:30.090 --> 00:11:33.100 it. Which brings us to this incredible economic 00:11:33.100 --> 00:11:36.399 paradox. Yes. The shift in the corporate attitude 00:11:36.399 --> 00:11:39.139 toward open source is one of the starkest pivots 00:11:39.139 --> 00:11:42.220 in tech history. Truly. I mean, in 2001, an executive 00:11:42.220 --> 00:11:44.659 at Microsoft literally referred to open source 00:11:44.659 --> 00:11:48.059 as an intellectual property destroyer. It was 00:11:48.059 --> 00:11:50.460 viewed as an existential threat to the entire 00:11:50.460 --> 00:11:53.019 software industry. But look at the data today. 00:11:53.240 --> 00:11:56.580 More than 90 % of companies use open -source 00:11:56.580 --> 00:11:59.039 software as a foundational component of their 00:11:59.039 --> 00:12:02.340 proprietary products. 90%. And Microsoft itself 00:12:02.340 --> 00:12:05.559 totally flipped. They now actively use it, they 00:12:05.559 --> 00:12:07.759 are one of the biggest contributors to it, and 00:12:07.759 --> 00:12:10.539 they've even integrated Linux into Windows. They 00:12:10.539 --> 00:12:13.860 realize they had to adapt or die. But this creates 00:12:13.860 --> 00:12:16.259 a fascinating puzzle for traditional economists. 00:12:16.559 --> 00:12:20.100 Oh, so. Standard economic theory usually focuses 00:12:20.100 --> 00:12:23.299 on the consumption of scarce resources. Like, 00:12:23.600 --> 00:12:26.740 if I eat an apple, you physically cannot eat 00:12:26.740 --> 00:12:29.019 that same apple. Right, supply and demand. But 00:12:29.019 --> 00:12:31.919 open source software operates as a public good. 00:12:32.200 --> 00:12:34.799 It is not scarce. If you download a web browser, 00:12:34.899 --> 00:12:36.840 it doesn't stop me from downloading the exact 00:12:36.840 --> 00:12:38.820 same browser. In fact, our research points out 00:12:38.820 --> 00:12:41.240 that open source actually increases in value 00:12:41.240 --> 00:12:43.159 the more people use it and contribute to it. 00:12:43.240 --> 00:12:45.179 It completely denies normal supply and demand. 00:12:45.200 --> 00:12:47.240 Because it relies on a totally different currency. 00:12:47.370 --> 00:12:50.750 reputation and network effects. Oh, clout. Exactly. 00:12:51.190 --> 00:12:52.950 A developer might spend a hundred hours writing 00:12:52.950 --> 00:12:56.049 code for free, but in return, they build a massive 00:12:56.049 --> 00:12:58.250 reputation among their peers globally. Which 00:12:58.250 --> 00:13:00.549 pays off later. Yeah, that reputation translates 00:13:00.549 --> 00:13:03.809 into lucrative job offers, consulting gigs and 00:13:03.809 --> 00:13:06.490 massive influence. The producer is essentially 00:13:06.490 --> 00:13:09.710 consuming the social benefit of their own free 00:13:09.710 --> 00:13:12.509 labor. It's an economy driven by clout and shared 00:13:12.509 --> 00:13:15.190 utility rather than, you know, direct point of 00:13:15.190 --> 00:13:18.299 sale transactions. Exactly. But for major corporations 00:13:18.299 --> 00:13:20.879 to play in this bazaar without getting sued into 00:13:20.879 --> 00:13:23.980 oblivion, there has to be some serious legal 00:13:23.980 --> 00:13:27.580 glue holding it all together. And the documentation 00:13:27.580 --> 00:13:30.539 breaks down the licensing into two main camps, 00:13:31.139 --> 00:13:34.820 permissive licenses and copy left licenses. Understanding 00:13:34.820 --> 00:13:36.980 the difference between these two is really the 00:13:36.980 --> 00:13:39.679 master key to seeing how corporations actually 00:13:39.679 --> 00:13:42.039 navigate this space today. Right. Both types 00:13:42.039 --> 00:13:44.460 grant you broad rights to use and modify the 00:13:44.460 --> 00:13:46.940 code. But they have violently different rules 00:13:46.940 --> 00:13:48.720 for what you are allowed to do after you make 00:13:48.720 --> 00:13:50.700 those changes Okay, let me try an analogy here 00:13:50.700 --> 00:13:53.159 to see if I have this right So a permissive license 00:13:53.159 --> 00:13:55.500 like the MIT or Apache licenses that are mentioned 00:13:55.500 --> 00:13:58.340 a lot in the source is like me saying Here is 00:13:58.340 --> 00:14:01.100 my grandmother's amazing award -winning cake 00:14:01.100 --> 00:14:03.490 recipe. Okay, I'm with you Do whatever you want 00:14:03.490 --> 00:14:05.389 with it. You can bake it. You can change the 00:14:05.389 --> 00:14:07.490 frosting. You can even put the final cake in 00:14:07.490 --> 00:14:09.690 a fancy box and sell it at a bakery for profit. 00:14:09.809 --> 00:14:13.090 Yes. And crucially, you don't own anything. And 00:14:13.090 --> 00:14:14.809 you don't have to tell anyone else how you made 00:14:14.809 --> 00:14:17.409 your specific version of the frosting. So no 00:14:17.409 --> 00:14:20.230 strings attached. Right. Permissive licenses. 00:14:20.809 --> 00:14:23.649 Let a company take open source code, build an 00:14:23.649 --> 00:14:26.909 app on top of it, and then make that final polished 00:14:26.909 --> 00:14:29.929 app completely proprietary and closed source. 00:14:30.240 --> 00:14:34.179 Got it. But a CopyLeft license -like Richard 00:14:34.179 --> 00:14:36.620 Stallman's GPL that we talked about earlier is 00:14:36.620 --> 00:14:38.379 different. That's like saying, here's the cake 00:14:38.379 --> 00:14:40.779 recipe, but if you add a secret ingredient to 00:14:40.779 --> 00:14:43.379 it to make it better, you are legally obligated 00:14:43.379 --> 00:14:46.259 to. To publish your new updated recipe for free 00:14:46.259 --> 00:14:50.179 for everyone else to use. Exactly. CopyLeft forces 00:14:50.179 --> 00:14:52.860 any derivative works to stay in the open source 00:14:52.860 --> 00:14:55.639 ecosystem. you cannot lock the new version back 00:14:55.639 --> 00:14:58.120 up. And this explains so much corporate behavior. 00:14:58.340 --> 00:15:00.519 I mean, if I'm a massive tech company, I am heavily 00:15:00.519 --> 00:15:02.159 favoring permissive licenses. As much as you 00:15:02.159 --> 00:15:05.659 are. It lets me take the incredible free foundation 00:15:05.659 --> 00:15:08.799 built by millions of volunteers in the bazaar, 00:15:09.059 --> 00:15:11.320 put a fresh coat of paint on it, and then lock 00:15:11.320 --> 00:15:14.299 it behind a proprietary paywall to sell. which 00:15:14.299 --> 00:15:16.860 leads to a growing and frankly controversial 00:15:16.860 --> 00:15:19.940 trend known as open sourcing. OK, what is that? 00:15:20.200 --> 00:15:22.220 Companies will release the bare bones foundation 00:15:22.220 --> 00:15:25.139 of their software under an open license, hoping 00:15:25.139 --> 00:15:27.860 to attract free volunteer developers to fix their 00:15:27.860 --> 00:15:30.740 bugs and improve the core code. Like Tom Sawyer 00:15:30.740 --> 00:15:33.360 getting people to paint the fence for him. Exactly 00:15:33.360 --> 00:15:36.139 like that. But they keep all the premium features 00:15:36.139 --> 00:15:38.600 closed off. It's a very one sided relationship. 00:15:38.720 --> 00:15:40.820 It sounds like they are basically mining the 00:15:40.820 --> 00:15:43.830 bizarre for free labor. It is a huge point of 00:15:43.830 --> 00:15:46.309 tension within the community today. Because if 00:15:46.309 --> 00:15:48.750 trillion -dollar corporations extract all this 00:15:48.750 --> 00:15:51.490 value without pumping enough resources like, 00:15:51.490 --> 00:15:53.669 you know, funding or developer time back into 00:15:53.669 --> 00:15:56.230 the ecosystem, the foundation could eventually 00:15:56.230 --> 00:15:58.190 crack. So what does this all mean? I mean, if 00:15:58.190 --> 00:16:00.210 this invisible architecture is running our banking 00:16:00.210 --> 00:16:03.210 systems, our hospitals, and our phones, and generating 00:16:03.210 --> 00:16:06.789 trillions in value, who exactly is pouring the 00:16:06.789 --> 00:16:09.490 concrete? It's a great question. Who are the 00:16:09.490 --> 00:16:12.909 actual human beings typing the code? And how 00:16:12.909 --> 00:16:15.190 are world governments reacting to the fact that 00:16:15.190 --> 00:16:18.330 their critical infrastructure relies on a decentralized 00:16:18.330 --> 00:16:20.350 bizarre? Well, if you look at the demographic 00:16:20.350 --> 00:16:23.409 data, it strips away a lot of the utopian myths. 00:16:23.730 --> 00:16:25.649 Yeah, it really does. You might picture this 00:16:25.649 --> 00:16:28.470 perfectly distributed global network, but contributors 00:16:28.470 --> 00:16:31.509 heavily cluster in specific established tech 00:16:31.509 --> 00:16:35.490 hubs. The U .S., China, Germany, India, and the 00:16:35.490 --> 00:16:37.649 UK dominate the contributions. Does that make 00:16:37.649 --> 00:16:40.509 sense? But the most jarring statistic is the 00:16:40.509 --> 00:16:43.480 massive disparity regarding gender. I'm completely 00:16:43.480 --> 00:16:45.799 stunned reading this. The data shows that in 00:16:45.799 --> 00:16:49.299 2002, only 1 .5 % of international open source 00:16:49.299 --> 00:16:52.360 developers were women. 1 .5%. It's nothing. And 00:16:52.360 --> 00:16:54.820 even tracking it across a massive timeline from 00:16:54.820 --> 00:16:58.600 2005 to 2021, that number only rose to 9 .8%. 00:16:58.600 --> 00:17:01.019 It is vastly unrepresentative of the general 00:17:01.019 --> 00:17:03.379 tech workforce. It is. And it directly challenges 00:17:03.379 --> 00:17:05.619 one of the community's most deeply held myths. 00:17:05.759 --> 00:17:09.200 Which one? The idea of a pure meritocracy. There's 00:17:09.200 --> 00:17:11.900 this pervasive belief in the bizarre that code 00:17:11.900 --> 00:17:14.240 is all that matters that because you are just 00:17:14.240 --> 00:17:17.099 the screen name on a forum Nobody cares who you 00:17:17.099 --> 00:17:19.079 are as long as your code compiles, right? It 00:17:19.079 --> 00:17:22.819 sounds very egalitarian But the research completely 00:17:22.819 --> 00:17:25.859 dismantles that myth completely when a contributor's 00:17:25.859 --> 00:17:28.339 gender is Identifiable as female in these forums 00:17:28.339 --> 00:17:31.920 or platforms. They face measurable systemic biases. 00:17:31.920 --> 00:17:34.880 It's in the data Yeah, they experience much longer 00:17:34.880 --> 00:17:37.599 feedback times when they submit code, their work 00:17:37.599 --> 00:17:40.259 is scrutinized far more heavily, and they have 00:17:40.259 --> 00:17:42.720 lower overall acceptance rates for their contributions 00:17:42.720 --> 00:17:45.259 compared to their male peers. And it goes beyond 00:17:45.259 --> 00:17:48.359 just code acceptance. They are frequently pushed 00:17:48.359 --> 00:17:50.880 by the community toward less technical housekeeping 00:17:50.880 --> 00:17:53.259 roles. Like what? Like manual testing, managing 00:17:53.259 --> 00:17:55.740 the community, or writing documentation. And 00:17:55.740 --> 00:17:57.559 this happens despite internal research showing 00:17:57.559 --> 00:18:00.039 they have equal productivity and capability in 00:18:00.039 --> 00:18:02.359 actual coding. That is incredibly frustrating. 00:18:02.519 --> 00:18:05.440 It is. So even in a supposedly open, borderless 00:18:05.440 --> 00:18:08.579 ecosystem, human biases are deeply hard -coded 00:18:08.579 --> 00:18:11.700 into the culture. It's a sobering reality. And 00:18:11.700 --> 00:18:14.099 that human element, the fact that this is built 00:18:14.099 --> 00:18:16.559 by real people with biases and vulnerabilities, 00:18:17.420 --> 00:18:20.519 becomes terrifying when you zoom out to the geopolitical 00:18:20.519 --> 00:18:22.759 level. Oh, absolutely. World governments have 00:18:22.759 --> 00:18:25.200 suddenly woken up to the fact that they are entirely 00:18:25.200 --> 00:18:28.119 dependent on this volunteer -maintained software. 00:18:28.559 --> 00:18:31.700 It's not just a hobbyist tool anymore. It is 00:18:31.700 --> 00:18:34.599 critical national infrastructure. And governments 00:18:34.599 --> 00:18:37.579 are just scrambling to figure out how to harness 00:18:37.579 --> 00:18:40.480 the economic power of the bizarre while actively 00:18:40.720 --> 00:18:42.859 protecting their national security from it. The 00:18:42.859 --> 00:18:45.160 examples here are fascinating. Look at France. 00:18:45.440 --> 00:18:48.339 They implemented a proactive policy to heavily 00:18:48.339 --> 00:18:50.960 favor free and open source software within their 00:18:50.960 --> 00:18:53.779 government agencies. Right. By simply driving 00:18:53.779 --> 00:18:56.059 demand toward open source, they generated an 00:18:56.059 --> 00:18:59.279 estimated 18 % increase in tech startups across 00:18:59.279 --> 00:19:02.559 the country. It acted as a massive economic engine. 00:19:02.819 --> 00:19:04.890 But then you have the defensive posture. like 00:19:04.890 --> 00:19:06.630 the United States Department of Defense. Oh, 00:19:06.630 --> 00:19:08.829 right. They utilize open source software extensively, 00:19:09.170 --> 00:19:11.269 but they have to screen it relentlessly. They 00:19:11.269 --> 00:19:13.650 evaluate the integrity of the code and they are 00:19:13.650 --> 00:19:15.849 specifically hunting for foreign governmental 00:19:15.849 --> 00:19:19.109 influence. Wait, I have to jump in here. If open 00:19:19.109 --> 00:19:22.089 source code is public and anyone can read it, 00:19:22.369 --> 00:19:25.089 how would a spy or a foreign government hide 00:19:25.089 --> 00:19:28.470 malicious code? Wouldn't one of those millions 00:19:28.470 --> 00:19:31.450 of eyeballs spot it immediately? Well, they don't 00:19:31.450 --> 00:19:33.849 attack the main fortress. They attack the supply 00:19:33.849 --> 00:19:36.460 chain. What do you mean? Modern software relies 00:19:36.460 --> 00:19:39.720 on dependencies. If you build an app, you don't 00:19:39.720 --> 00:19:42.180 write the code to render a JPEG image from scratch. 00:19:42.460 --> 00:19:44.000 Right. I guess I'd just use a tool for that. 00:19:44.240 --> 00:19:46.869 Exactly. You just plug in an existing open source 00:19:46.869 --> 00:19:49.009 component that does it for you. Your app is now 00:19:49.009 --> 00:19:51.710 dependent on that tiny piece of code. OK, I follow. 00:19:51.990 --> 00:19:55.130 Bad actors will target obscure, forgotten subcomponents 00:19:55.130 --> 00:19:57.549 that are maintained by maybe one or two burned 00:19:57.549 --> 00:20:00.630 out volunteers. If they can slip a vulnerability 00:20:00.630 --> 00:20:03.089 into that tiny component, it gets pulled into 00:20:03.089 --> 00:20:05.150 the massive cathedral projects automatically. 00:20:05.569 --> 00:20:08.769 Oh, wow. That is terrifying. It is. Which is 00:20:08.769 --> 00:20:11.170 why this is also becoming a weapon of technological 00:20:11.170 --> 00:20:14.019 independence. Look at China. The Huawei situation. 00:20:14.220 --> 00:20:18.000 Yes. In 2019, amidst global trade tensions, the 00:20:18.000 --> 00:20:20.839 Chinese telecom giant Huawei was blocked from 00:20:20.839 --> 00:20:23.359 using Google's proprietary Android ecosystem. 00:20:23.720 --> 00:20:26.259 Right. Completely cut off. In response, they 00:20:26.259 --> 00:20:28.400 accelerated the building of their own alternative 00:20:28.400 --> 00:20:31.819 operating system, Harmony OS. And they heavily 00:20:31.819 --> 00:20:35.019 relied on the open source model to bypass those 00:20:35.019 --> 00:20:37.299 international restrictions and rapidly build 00:20:37.299 --> 00:20:39.170 out their own infrastructure. And then you have 00:20:39.170 --> 00:20:41.069 countries like Germany who are realizing they 00:20:41.069 --> 00:20:43.630 just need to pay for the concrete. Exactly. They 00:20:43.630 --> 00:20:46.250 established the Sovereign Tech Fund. Right. Which 00:20:46.250 --> 00:20:49.369 uses government money to financially support 00:20:49.369 --> 00:20:52.210 the independent maintainers of these vital open 00:20:52.210 --> 00:20:55.529 source tools. They are literally paying to reinforce 00:20:55.529 --> 00:20:57.670 the invisible foundation before it collapses. 00:20:58.190 --> 00:21:00.509 If we connect this to the bigger picture, we 00:21:00.509 --> 00:21:03.509 are witnessing a massive paradigm shift in how 00:21:03.509 --> 00:21:06.190 global power operates. Governments are walking 00:21:06.190 --> 00:21:08.970 a razor -thin tightrope. They desperately need 00:21:08.970 --> 00:21:12.069 the incredible, rapid technological innovation 00:21:12.069 --> 00:21:15.329 that only comes from open, borderless collaboration. 00:21:15.430 --> 00:21:17.049 Because they can't afford to build their own 00:21:17.049 --> 00:21:19.849 cathedrals from scratch. Right. But they are 00:21:19.849 --> 00:21:22.470 terrified of what might be lurking in the shadows 00:21:22.470 --> 00:21:25.869 of the bazaar. They are acutely aware of the 00:21:25.869 --> 00:21:29.369 vulnerabilities. This tension between open collaboration 00:21:29.369 --> 00:21:32.890 and localized national security is going to define 00:21:32.890 --> 00:21:35.849 the next decade of technology. It really forces 00:21:35.849 --> 00:21:38.289 you to look at the digital world through a completely 00:21:38.289 --> 00:21:40.349 different lens. I mean, the next time you open 00:21:40.349 --> 00:21:43.450 your web browser or launch an app on your smartphone 00:21:43.450 --> 00:21:46.359 or even just read an article online. Take a moment 00:21:46.359 --> 00:21:48.680 to realize what you are actually touching. It's 00:21:48.680 --> 00:21:51.079 all connected. The invisible architecture of 00:21:51.079 --> 00:21:53.420 your digital life wasn't just built in a sleek 00:21:53.420 --> 00:21:56.319 corporate boardroom by people in suits. It was 00:21:56.319 --> 00:22:00.160 built in a chaotic, messy, global bazaar. It's 00:22:00.160 --> 00:22:03.160 an unlikely alliance of moral idealists, weekend 00:22:03.160 --> 00:22:05.700 hackers, and trillion dollar corporate giants 00:22:05.700 --> 00:22:08.220 all somehow agreeing to share the exact same 00:22:08.220 --> 00:22:10.619 digital bricks. And the most provocative detail 00:22:10.619 --> 00:22:12.859 from all our research suggests that this collaborative 00:22:12.859 --> 00:22:15.220 model might not stay confined to your screen. 00:22:15.119 --> 00:22:18.160 for much longer. There's a brief but fascinating 00:22:18.160 --> 00:22:21.079 mention of the term open source expanding into 00:22:21.079 --> 00:22:23.319 movements like open source ecology. Open source 00:22:23.319 --> 00:22:25.420 ecology, what does that actually look like in 00:22:25.420 --> 00:22:29.579 practice? It's the ambition to completely decentralize 00:22:29.579 --> 00:22:33.240 physical technology. It takes the exact same 00:22:33.240 --> 00:22:35.180 principles of the software bizarre, you know, 00:22:35.220 --> 00:22:38.119 open blueprints, shared schematics, collaborative 00:22:38.119 --> 00:22:41.460 problem -solving, and applies them to hardware. 00:22:41.640 --> 00:22:43.579 The hardware? Like what? We're talking about 00:22:43.579 --> 00:22:46.180 open source tractors, 3D printed housing, and 00:22:46.180 --> 00:22:48.740 medical equipment. Wow. So if the open source 00:22:48.740 --> 00:22:51.059 bazaar could disrupt the proprietary software 00:22:51.059 --> 00:22:54.240 industry to the tune of 8 .8 trillion dollars, 00:22:55.059 --> 00:22:57.500 what happens when we fully unleash that model 00:22:57.500 --> 00:22:59.619 on the physical world? It changes everything. 00:22:59.880 --> 00:23:01.880 Will we soon be downloading and tweaking the 00:23:01.880 --> 00:23:04.140 blueprints for our cars, our medical devices, 00:23:04.319 --> 00:23:06.579 or our homes just as easily as we download a 00:23:06.579 --> 00:23:08.789 new web browser today? The digital revolution 00:23:08.789 --> 00:23:10.589 you're holding in your hand right now might just 00:23:10.589 --> 00:23:13.150 be a preview of a much larger physical revolution. 00:23:13.529 --> 00:23:16.069 A physical foundation built by everyone for everyone. 00:23:16.390 --> 00:23:18.029 Something to think about the next time you log 00:23:18.029 --> 00:23:20.230 on. Until next time, keep diving deep.