WEBVTT 00:00:00.000 --> 00:00:01.700 I've been thinking a lot about the friction of 00:00:01.700 --> 00:00:04.700 creation. You know, you have this genuinely brilliant 00:00:04.700 --> 00:00:08.179 idea in your head, but the tools just completely 00:00:08.179 --> 00:00:10.519 stand in your way. You open some massive application. 00:00:10.839 --> 00:00:13.080 Yeah, and the interface is incredibly overwhelming. 00:00:13.640 --> 00:00:16.460 Exactly. It drains your creative momentum instantly. 00:00:16.879 --> 00:00:19.899 Well, what if Google quietly buried its most 00:00:19.899 --> 00:00:23.199 powerful AI video engine inside a boring workspace 00:00:23.199 --> 00:00:25.480 app, and they just kind of forgot to tell anyone? 00:00:25.800 --> 00:00:28.440 That's exactly what we're exploring today. Welcome 00:00:28.440 --> 00:00:30.859 to the Deep Dive. We're unpacking a fascinating 00:00:30.859 --> 00:00:33.740 guide by Max Anne. He published this comprehensive 00:00:33.740 --> 00:00:37.659 breakdown back in April 2026. Right. It explores 00:00:37.659 --> 00:00:40.899 the hidden depths of Google vids. Specifically, 00:00:41.140 --> 00:00:44.219 it focuses on the stealthy integration of VO 00:00:44.219 --> 00:00:47.320 3 .1. It's a completely wild update. It fundamentally 00:00:47.320 --> 00:00:49.719 changes how we approach daily content creation. 00:00:50.060 --> 00:00:52.200 Our mission today is highly focused. It's also 00:00:52.200 --> 00:00:54.789 very practical. We're going to figure out. how 00:00:54.789 --> 00:00:56.689 to stop jumping between five different apps. 00:00:56.890 --> 00:00:58.829 Making one simple video shouldn't require endless 00:00:58.829 --> 00:01:01.350 open tabs. No, you really shouldn't. We're going 00:01:01.350 --> 00:01:03.250 to master this hidden all -in -one production 00:01:03.250 --> 00:01:05.870 studio. You're going to save an unbelievable 00:01:05.870 --> 00:01:09.890 amount of time. This workflow eliminates so many 00:01:09.890 --> 00:01:12.689 unnecessary technical hurdles. Let's start by 00:01:12.689 --> 00:01:16.069 acknowledging the painful status quo. Editing 00:01:16.069 --> 00:01:19.230 video right now feels incredibly fragmented. 00:01:19.329 --> 00:01:21.950 Oh, deeply frustrating. I call it the power outage 00:01:21.950 --> 00:01:24.540 feeling. You spend hours formatting a project 00:01:24.540 --> 00:01:27.739 across multiple tools, and then something breaks. 00:01:28.180 --> 00:01:30.799 You lose your entire creative flow. Exactly. 00:01:30.879 --> 00:01:33.239 You pull your voiceover from one random website. 00:01:33.359 --> 00:01:35.620 You generate background images in a totally different 00:01:35.620 --> 00:01:38.180 application. Right. You download all these intermediate 00:01:38.180 --> 00:01:40.400 files to your desktop. Then you try to smash 00:01:40.400 --> 00:01:42.359 them together in an editor. And the frame rates 00:01:42.359 --> 00:01:45.319 almost never match up perfectly. The aspect ratios 00:01:45.319 --> 00:01:47.840 are always slightly off. Yeah, you spend more 00:01:47.840 --> 00:01:50.859 time formatting than actually creating. So Google 00:01:50.859 --> 00:01:54.859 basically observed that exact friction and they 00:01:54.859 --> 00:01:57.400 quietly solved it. They built a stealthy solution 00:01:57.400 --> 00:02:00.140 right into their standard ecosystem. You literally 00:02:00.140 --> 00:02:02.400 just type vids .new into your browser. And you 00:02:02.400 --> 00:02:04.920 instantly enter a completely integrated video 00:02:04.920 --> 00:02:07.760 production environment. It feels incredibly familiar, 00:02:07.939 --> 00:02:10.979 mostly because it looks exactly like Google Docs. 00:02:11.000 --> 00:02:13.879 Right. Which brings us to the massive hidden 00:02:13.879 --> 00:02:17.860 update. As of April 2nd, 2026, personal accounts 00:02:17.860 --> 00:02:22.120 get something huge. You get 10 free VO 3 .1 generations 00:02:22.120 --> 00:02:25.020 every single month. Which is absolutely crazy 00:02:25.020 --> 00:02:27.539 to think about. That feature isn't even available 00:02:27.539 --> 00:02:30.099 in the free version of Gemini. No, it isn't. 00:02:30.120 --> 00:02:34.039 VO 3 .1 is their flagship high -end video generation 00:02:34.039 --> 00:02:36.229 model. Right. It understands. complex physics. 00:02:36.699 --> 00:02:39.080 And highly realistic lighting. And Google just 00:02:39.080 --> 00:02:42.180 buried their best video AI inside a workspace 00:02:42.180 --> 00:02:44.740 tool. So you can generate high quality clips 00:02:44.740 --> 00:02:47.419 right on the timeline. You can record voiceovers 00:02:47.419 --> 00:02:49.900 directly into the browser. You can edit the entire 00:02:49.900 --> 00:02:52.379 sequence together seamlessly. You can even export 00:02:52.379 --> 00:02:54.680 directly to Drive or YouTube from there. Right. 00:02:54.740 --> 00:02:57.020 And you do all of this without downloading a 00:02:57.020 --> 00:02:58.860 single intermediate file. You never have to clutter 00:02:58.860 --> 00:03:01.439 your desktop with random MP4s again. But I have 00:03:01.439 --> 00:03:04.020 to ask, does this actually replace professional 00:03:04.020 --> 00:03:07.430 editing software suites? Or is it just a sandbox 00:03:07.430 --> 00:03:09.969 for quick projects? Well, it's definitely not 00:03:09.969 --> 00:03:12.509 replacing high -end Hollywood editing suites 00:03:12.509 --> 00:03:14.770 today. You're not going to cut a feature film 00:03:14.770 --> 00:03:17.349 in Google Vids. Right. It's designed to replace 00:03:17.349 --> 00:03:21.210 the messy, fragmented everyday workflow. It handles 00:03:21.210 --> 00:03:24.150 corporate explainers, internal memos, and quick 00:03:24.150 --> 00:03:27.110 social media clips flawlessly. So it replaces 00:03:27.110 --> 00:03:29.449 the messy workflow, not the pro editing suites. 00:03:29.610 --> 00:03:32.289 Precisely. It makes everyday communication dramatically 00:03:32.289 --> 00:03:34.430 smoother. Now that we know where this hidden 00:03:34.430 --> 00:03:37.550 tool lives. we need to understand how to actually 00:03:37.550 --> 00:03:40.129 use it. Right, without creating a complete mess. 00:03:40.310 --> 00:03:44.370 Because AI can generate some truly chaotic, disconnected 00:03:44.370 --> 00:03:48.210 clips. Yeah, if you just let it run wild. We 00:03:48.210 --> 00:03:50.129 really need to master the core workflow here. 00:03:50.650 --> 00:03:53.310 How do we keep the generated footage looking 00:03:53.310 --> 00:03:56.129 intentional and cohesive? Well, you have to embrace 00:03:56.129 --> 00:03:58.789 a very specific core loop. You generate the clip, 00:03:58.909 --> 00:04:01.539 you preview it. and then you insert. The biggest 00:04:01.539 --> 00:04:03.960 mindset shift is treating the AI like a first 00:04:03.960 --> 00:04:06.659 draft. You can't expect absolute perfection on 00:04:06.659 --> 00:04:08.979 the very first try. Never. You have to iterate 00:04:08.979 --> 00:04:12.419 and shape the output gradually. It's a probabilistic 00:04:12.419 --> 00:04:15.120 model making educated guesses. Let me be a bit 00:04:15.120 --> 00:04:17.560 vulnerable here. I still wrestle with prompt 00:04:17.560 --> 00:04:20.180 drift myself. Oh, it's a massive headache for 00:04:20.180 --> 00:04:23.240 absolutely everyone using these tools. Just to 00:04:23.240 --> 00:04:26.939 define that quickly, it is when AI forgets what 00:04:26.939 --> 00:04:29.189 your character looks like over time. Yeah. Your 00:04:29.189 --> 00:04:32.089 first scene looks incredibly cinematic and perfectly 00:04:32.089 --> 00:04:34.990 lit. Your protagonist is wearing a specific blue 00:04:34.990 --> 00:04:37.529 coat. Right. Then your second scene features 00:04:37.529 --> 00:04:39.889 a completely different looking protagonist. The 00:04:39.889 --> 00:04:42.430 coat is red. The lighting is completely flat. 00:04:42.649 --> 00:04:45.949 Exactly. Max's guide highlights a brilliant pro 00:04:45.949 --> 00:04:48.290 trick to fix this. It's called the Animid Image 00:04:48.290 --> 00:04:51.110 Continuity Hack. This is where the workflow actually 00:04:51.110 --> 00:04:53.610 gets incredibly powerful. Yeah. You generate 00:04:53.610 --> 00:04:56.000 your first clip. You scrub to the very end of 00:04:56.000 --> 00:04:58.800 that timeline. You take a screenshot of the last 00:04:58.800 --> 00:05:01.000 frame of scene A. So you just capture the final 00:05:01.000 --> 00:05:03.540 moment of that specific visual. Then you upload 00:05:03.540 --> 00:05:05.819 that screenshot as the image prompt for scene 00:05:05.819 --> 00:05:08.779 B. You tell the AI exactly what needs to happen 00:05:08.779 --> 00:05:11.459 next. Like the polar bear leaps forward into 00:05:11.459 --> 00:05:14.420 the freezing water. Exactly. Because you provided 00:05:14.420 --> 00:05:17.439 the visual anchor of the first frame. The AI 00:05:17.439 --> 00:05:20.560 inherits the lighting, the character, and the 00:05:20.560 --> 00:05:23.399 exact style. It uses that image as the starting 00:05:23.399 --> 00:05:26.959 line. The guide compares this to stacking Lego 00:05:26.959 --> 00:05:29.199 blocks perfectly instead of just throwing them 00:05:29.199 --> 00:05:31.319 in a random pile. I love that analogy. I actually 00:05:31.319 --> 00:05:33.319 think of it like passing a baton in a relay race. 00:05:33.420 --> 00:05:36.019 Oh, that's good too. The second runner doesn't 00:05:36.019 --> 00:05:38.560 start from a complete standstill. They inherit 00:05:38.560 --> 00:05:41.019 the forward momentum of the first runner. That's 00:05:41.019 --> 00:05:44.550 a fantastic way to look at it. It creates a genuinely 00:05:44.550 --> 00:05:48.269 continuous visual narrative for the viewer. And 00:05:48.269 --> 00:05:50.709 here's another highly counterintuitive tip for 00:05:50.709 --> 00:05:53.329 this exact workflow. Skipping the transitions 00:05:53.329 --> 00:05:56.410 between these linked clips often looks much more 00:05:56.410 --> 00:05:58.550 natural. That's interesting. Why shouldn't we 00:05:58.550 --> 00:06:01.129 just use dissolve transitions to hide the imperfections 00:06:01.129 --> 00:06:03.870 between generations? Well, because dissolve transitions 00:06:03.870 --> 00:06:07.149 actually blend the pixels of both distinct frames. 00:06:07.310 --> 00:06:10.730 It artificially highlights the slight AI variations 00:06:10.730 --> 00:06:13.649 instead of hiding them. Oh, I see. A hard cut 00:06:13.649 --> 00:06:16.069 forces the human brain to naturally accept the 00:06:16.069 --> 00:06:18.730 visual continuity. Right. No transitions keeps 00:06:18.730 --> 00:06:20.930 the inherited frames feeling like one continuous 00:06:20.930 --> 00:06:23.790 shot. Exactly. It feels much more like a deliberate 00:06:23.790 --> 00:06:26.850 camera movement. So we've mastered the visual 00:06:26.850 --> 00:06:29.550 continuity of the generated clips. Yeah. Let's 00:06:29.550 --> 00:06:31.410 shift our focus to the most practical feature 00:06:31.410 --> 00:06:33.990 for daily work. This is something that surprisingly 00:06:33.990 --> 00:06:36.850 doesn't cost a single dime. The free tier of 00:06:36.850 --> 00:06:39.290 this tool is actually incredibly robust. It's 00:06:39.290 --> 00:06:41.449 not just a crippled trial version. It handles 00:06:41.449 --> 00:06:44.250 the core workflow effortlessly for most normal 00:06:44.250 --> 00:06:47.209 users. You get the voiceovers, the basic editing, 00:06:47.370 --> 00:06:50.370 and the stock media. And the AI voiceovers are 00:06:50.370 --> 00:06:52.800 actually surprisingly good for a... totally free 00:06:52.800 --> 00:06:55.980 tool. Yeah. They have this calm, measured teacher 00:06:55.980 --> 00:06:58.620 style that works perfectly for tutorials. But 00:06:58.620 --> 00:07:00.839 the guide points out a massive sleeper feature 00:07:00.839 --> 00:07:03.959 hidden in the menus. It's the slides to video 00:07:03.959 --> 00:07:05.980 conversion tool. This is the feature that will 00:07:05.980 --> 00:07:08.660 literally save people hundreds of hours every 00:07:08.660 --> 00:07:11.060 month. You just go up to the file menu and click 00:07:11.060 --> 00:07:14.740 convert slides. Two sec silence. Imagine turning 00:07:14.740 --> 00:07:17.720 a 50 -slide corporate deck into a polished narrated 00:07:17.720 --> 00:07:21.060 video in seconds. Right? It completely changes 00:07:21.060 --> 00:07:23.720 how we share boring information at work. You 00:07:23.720 --> 00:07:25.779 no longer have to record yourself reading endless 00:07:25.779 --> 00:07:28.720 bullet points. The AI automatically writes a 00:07:28.720 --> 00:07:31.399 cohesive connecting script for you. It adds highly 00:07:31.399 --> 00:07:34.060 natural narration. And layers subtle background 00:07:34.060 --> 00:07:36.939 music underneath. It even handles the visual 00:07:36.939 --> 00:07:39.720 transitions between the slides completely automatically. 00:07:40.040 --> 00:07:42.560 You just review the final output and make a few 00:07:42.560 --> 00:07:45.860 quick stylistic tweaks. But how does the AI actually 00:07:45.860 --> 00:07:49.699 generate a cohesive script if a slide deck only 00:07:49.699 --> 00:07:53.160 contains sparse bullet points? It uses the underlying 00:07:53.160 --> 00:07:55.540 language model to deeply analyze the overall 00:07:55.540 --> 00:07:57.579 context. It doesn't just read the individual 00:07:57.579 --> 00:08:00.079 words on the screen. It looks at the semantic 00:08:00.079 --> 00:08:02.560 flow of the entire presentation holistically. 00:08:02.879 --> 00:08:05.079 So if slide three presents a... major problem. 00:08:05.279 --> 00:08:08.220 Right. And slide four presents a specific solution. 00:08:08.540 --> 00:08:11.399 It naturally infers the missing spoken context 00:08:11.399 --> 00:08:14.019 between those sparse bullets. Got it. It reads 00:08:14.019 --> 00:08:16.220 the broader context and writes the connecting 00:08:16.220 --> 00:08:18.319 tissue automatically. Yeah. It acts like a really 00:08:18.319 --> 00:08:21.120 smart, highly trained presentation coach, sponsor. 00:08:21.500 --> 00:08:23.620 Automating those massive slide decks is going 00:08:23.620 --> 00:08:26.160 to save people countless hours of pure drudgery. 00:08:26.300 --> 00:08:28.939 Absolutely. But the paid tier attempts to automate 00:08:28.939 --> 00:08:31.480 the human presenter entirely. This brings us 00:08:31.480 --> 00:08:34.019 to the AI avatars and the digital ingredients. 00:08:34.340 --> 00:08:36.559 Let's unpack how these digital avatars actually 00:08:36.559 --> 00:08:39.320 function in practice. You start by selecting 00:08:39.320 --> 00:08:42.159 a digital spokesperson from their built -in library. 00:08:42.460 --> 00:08:44.940 Then you upload an image of your actual physical 00:08:44.940 --> 00:08:47.220 product. They call this product image an ingredient 00:08:47.220 --> 00:08:50.200 in the VID system. The digital avatar then naturally 00:08:50.200 --> 00:08:52.360 interacts with it, all while reading your written 00:08:52.360 --> 00:08:55.360 script. It's absolutely perfect for software 00:08:55.360 --> 00:08:58.440 unboxing videos or quick feature tutorials. But 00:08:58.440 --> 00:09:01.059 the guide emphasizes a very critical rule for 00:09:01.059 --> 00:09:03.580 using these avatars. You must always run the 00:09:03.580 --> 00:09:06.279 preview first before committing completely. Right. 00:09:06.419 --> 00:09:08.899 Rendering these complex videos takes significant 00:09:08.899 --> 00:09:12.379 time. And it uses up your monthly credits. You 00:09:12.379 --> 00:09:14.259 have to check the lip sync and the overall pacing 00:09:14.259 --> 00:09:16.360 first. You really want to ensure the delivery 00:09:16.360 --> 00:09:19.059 tone matches your actual written script? Right. 00:09:19.120 --> 00:09:22.059 And a serious topic requires a deliberately serious 00:09:22.059 --> 00:09:24.629 delivery. Once the preview looks perfect, then 00:09:24.629 --> 00:09:26.610 you finally render the whole thing. The guide 00:09:26.610 --> 00:09:29.889 also emphasizes a broader philosophy of skipping 00:09:29.889 --> 00:09:32.909 the blank canvas entirely. The blank canvas is 00:09:32.909 --> 00:09:35.250 where most creative projects completely die. 00:09:35.429 --> 00:09:38.009 It's incredibly intimidating. You stare at a 00:09:38.009 --> 00:09:41.820 blinking cursor and just feel... entirely paralyzed 00:09:41.820 --> 00:09:45.220 the unlimited possibilities are too much so you 00:09:45.220 --> 00:09:47.419 should always use the built -in templates to 00:09:47.419 --> 00:09:50.759 start your workflow the ai image generation includes 00:09:50.759 --> 00:09:54.840 these really helpful fully working example prompts 00:09:54.840 --> 00:09:57.440 you're editing a working prompt rather than starting 00:09:57.440 --> 00:10:00.100 from absolute zero it gives you immediate creative 00:10:00.100 --> 00:10:02.440 momentum to just keep moving forward you just 00:10:02.440 --> 00:10:05.679 swap out a few key nouns and verbs But do these 00:10:05.679 --> 00:10:07.860 avatars risk falling into the uncanny valley 00:10:07.860 --> 00:10:10.500 and alienating the viewer? They definitely can 00:10:10.500 --> 00:10:12.779 if you just use them in an empty digital void. 00:10:13.120 --> 00:10:15.899 The human brain quickly detects something slightly 00:10:15.899 --> 00:10:18.419 artificial and feels creeped out. But when you 00:10:18.419 --> 00:10:20.779 give them a real physical ingredient to hold, 00:10:20.940 --> 00:10:24.039 it forces the viewer's brain to focus on the 00:10:24.039 --> 00:10:26.899 actual tangible product. Yeah, so grounding them 00:10:26.899 --> 00:10:29.549 with real -world product photos. Bridges that 00:10:29.549 --> 00:10:32.649 artificial gap. Exactly. It psychologically anchors 00:10:32.649 --> 00:10:35.370 the digital avatar to our shared physical reality. 00:10:35.649 --> 00:10:38.029 We've covered the absolute magic of this new 00:10:38.029 --> 00:10:41.330 integrated workflow. It really is magic. But 00:10:41.330 --> 00:10:43.870 to make this a reliable and truly daily tool, 00:10:44.049 --> 00:10:46.669 we really have to talk about the very real technical 00:10:46.669 --> 00:10:49.309 limitations. These are the hidden guardrails 00:10:49.309 --> 00:10:52.250 that will definitely trip new users up. The 10 00:10:52.250 --> 00:10:55.850 free VO3 .1 generations are basically just a 00:10:55.850 --> 00:10:58.460 generous sampler. They're not an unlimited buffet 00:10:58.460 --> 00:11:00.539 for you to endlessly consume. They'll vanish 00:11:00.539 --> 00:11:02.720 incredibly quickly if you waste them thoughtlessly. 00:11:02.899 --> 00:11:05.620 You have to be highly intentional with every 00:11:05.620 --> 00:11:08.360 single text prompt you write. You also need to 00:11:08.360 --> 00:11:10.419 heavily adopt the good enough fast philosophy. 00:11:10.740 --> 00:11:13.659 This tool is built specifically for internal 00:11:13.659 --> 00:11:16.460 explainers and basic training videos. Right. 00:11:16.559 --> 00:11:19.039 It's fantastic for quick social clips and daily 00:11:19.039 --> 00:11:22.039 memos. It's absolutely not meant for high -end 00:11:22.039 --> 00:11:24.200 theatrical commercial production. The biggest 00:11:24.200 --> 00:11:26.799 mistake new users make is expecting perfect initial 00:11:26.799 --> 00:11:29.899 results. They type one lazy prompt and get genuinely 00:11:29.899 --> 00:11:32.580 angry when it fails. You have to think in iterative 00:11:32.580 --> 00:11:35.100 sequences to be truly successful here. You adjust 00:11:35.100 --> 00:11:37.919 the text, you tweak the settings, and you slowly 00:11:37.919 --> 00:11:40.039 refine the output. You're guiding the system. 00:11:40.360 --> 00:11:43.440 Let me challenge that workflow a bit. If treating 00:11:43.440 --> 00:11:47.159 the AI as a first draft means iterating, does 00:11:47.159 --> 00:11:50.019 the user actually end up spending more time tweaking 00:11:50.019 --> 00:11:53.690 than they would just filming themselves? Think 00:11:53.690 --> 00:11:56.549 about the sheer physical logistics of real -world 00:11:56.549 --> 00:11:59.389 filming. You have to set up lighting and test 00:11:59.389 --> 00:12:01.649 your audio levels. Yeah, that's true. You do 00:12:01.649 --> 00:12:04.590 seven different takes because you misspoke. Then 00:12:04.590 --> 00:12:08.250 you offload huge files from an SD card. Tweaking 00:12:08.250 --> 00:12:10.929 a prompt takes about 30 seconds of typing text. 00:12:11.169 --> 00:12:13.049 Makes sense. The iteration is still dramatically 00:12:13.049 --> 00:12:15.870 faster than setting up a real camera. Plus, you 00:12:15.870 --> 00:12:17.470 never have to worry about how your hair looks 00:12:17.470 --> 00:12:20.090 on camera. Let's synthesize the big idea from 00:12:20.090 --> 00:12:23.029 this entire comprehensive guide. The defining 00:12:23.029 --> 00:12:26.490 shift of 2026 is the ultimate end of the blank 00:12:26.490 --> 00:12:28.970 canvas. You no longer have to build every single 00:12:28.970 --> 00:12:32.090 creative element from scratch. Google Vids is 00:12:32.090 --> 00:12:34.309 essentially asking you to become a creative director. 00:12:34.549 --> 00:12:37.470 You're assembling digital ingredients and guiding 00:12:37.470 --> 00:12:39.629 the overall creative vision. You're no longer 00:12:39.629 --> 00:12:42.490 just a manual editor fighting with complex software 00:12:42.490 --> 00:12:45.190 constraints. It elevates your daily role from 00:12:45.190 --> 00:12:48.750 pure manual labor to high -level strategic curation. 00:12:48.830 --> 00:12:50.710 I want to leave you with a final provocative 00:12:50.710 --> 00:12:54.629 thought today. If every single internal memo, 00:12:54.889 --> 00:12:58.450 sales pitch and training deck can now be instantly 00:12:58.450 --> 00:13:01.049 converted into a high quality narrated video, 00:13:01.370 --> 00:13:04.769 what actually happens to the written word in 00:13:04.769 --> 00:13:07.110 the modern workplace? That's a really profound 00:13:07.110 --> 00:13:09.470 and slightly terrifying question to consider 00:13:09.470 --> 00:13:12.450 right now. Does writing become completely obsolete 00:13:12.450 --> 00:13:15.600 for daily corporate communication? Or does it 00:13:15.600 --> 00:13:17.720 force us to only write things that truly matter? 00:13:17.960 --> 00:13:20.220 It might genuinely elevate writing to a much 00:13:20.220 --> 00:13:22.840 higher, highly premium format. We highly encourage 00:13:22.840 --> 00:13:25.600 you to open up vids .new today. Generate just 00:13:25.600 --> 00:13:27.519 one single clip and see how it actually feels. 00:13:27.620 --> 00:13:29.320 See how far you get before hitting a creative 00:13:29.320 --> 00:13:32.000 wall. You'll be deeply surprised by how highly 00:13:32.000 --> 00:13:34.480 intuitive it actually is. It turns out the most 00:13:34.480 --> 00:13:37.700 powerful video engine wasn't flashy. It was quietly 00:13:37.700 --> 00:13:41.279 buried inside a boring workspace app all along. 00:13:42.159 --> 00:13:42.899 Arturo Music.