WEBVTT 00:00:00.000 --> 00:00:01.740 You know, for the most powerful computers we 00:00:01.740 --> 00:00:04.200 have, the absolute limit, it used to be just 00:00:04.200 --> 00:00:06.519 13 seconds. 13 seconds, yeah. That's how long 00:00:06.519 --> 00:00:08.679 they could stay running before, well, physics 00:00:08.679 --> 00:00:10.779 basically shut them down. It sounds like a long 00:00:10.779 --> 00:00:14.060 time in maybe pure computation terms, but in 00:00:14.060 --> 00:00:16.699 reality, nothing. Imagine trying to run anything 00:00:16.699 --> 00:00:18.579 serious knowing the whole thing's going to collapse. 00:00:18.780 --> 00:00:22.399 Exactly. That limit, atom loss, it was the ceiling 00:00:22.399 --> 00:00:25.320 for quantum, but... the sources we looked at 00:00:25.320 --> 00:00:27.760 today they show this huge harvard breakthrough 00:00:27.760 --> 00:00:31.000 right a physics team built a system a quantum 00:00:31.000 --> 00:00:34.460 system that actually repairs itself mid -calculation 00:00:34.460 --> 00:00:37.600 whoa okay that changes everything seriously we're 00:00:37.600 --> 00:00:40.640 shifting from these like super delicate lab demos 00:00:40.640 --> 00:00:43.039 like can you miss it yeah to a real possibility 00:00:43.039 --> 00:00:46.719 of like always on quantum servers that's the 00:00:46.719 --> 00:00:49.530 pivot For the real world, for commercial use. 00:00:49.789 --> 00:00:51.950 Huge. Welcome to the Deep Dive. We're going to 00:00:51.950 --> 00:00:54.170 pull on three big threads from the sources you 00:00:54.170 --> 00:00:56.710 shared with us. And the theme that kind of ties 00:00:56.710 --> 00:00:59.789 them together is this hunt for reliability, whether 00:00:59.789 --> 00:01:02.469 that's securing AI for businesses or keeping 00:01:02.469 --> 00:01:04.569 these quantum machines actually running. Yeah, 00:01:04.609 --> 00:01:06.709 and our roadmap really shows how deep that goes. 00:01:06.870 --> 00:01:09.780 First up, we're going to unpack... IBM's big 00:01:09.780 --> 00:01:12.939 move, very strategic one, I think, into open 00:01:12.939 --> 00:01:16.560 source AI with Granite 4 .0. Okay. Then second, 00:01:16.620 --> 00:01:20.180 we'll look at this boom in AI agents and agentic 00:01:20.180 --> 00:01:22.700 systems and what they're actually becoming. You 00:01:22.700 --> 00:01:24.819 know, the practical tools for businesses. And 00:01:24.819 --> 00:01:27.120 finally, yeah, we dig into that quantum breakthrough, 00:01:27.340 --> 00:01:29.040 the self -healing machine, the one that gets 00:01:29.040 --> 00:01:32.219 rid of atom loss, making these really long computations 00:01:32.219 --> 00:01:34.599 possible, like truly possible for the first time. 00:01:34.640 --> 00:01:37.920 Right. OK, let's start with Granite. So IBM rolled 00:01:37.920 --> 00:01:41.280 out this Granite 4 .0 family and their positioning 00:01:41.280 --> 00:01:44.459 seems really clear. It's hybrid, it's enterprise 00:01:44.459 --> 00:01:48.400 ready and crucially open source. They don't seem 00:01:48.400 --> 00:01:51.239 to be chasing the, you know, biggest model trophy. 00:01:51.879 --> 00:01:54.459 more like the most deployable one. That distinction, 00:01:54.519 --> 00:01:56.260 it's everything here. This isn't just another, 00:01:56.379 --> 00:01:59.120 hey, look, a big model announcement. It's a strategic 00:01:59.120 --> 00:02:03.319 play designed for the messy reality of corporate 00:02:03.319 --> 00:02:06.359 IT. They even made it light enough to run on 00:02:06.359 --> 00:02:10.840 pretty cheap GPUs, which for companies wanting 00:02:10.840 --> 00:02:13.659 to use it internally, maybe on the edge, that's 00:02:13.659 --> 00:02:15.900 critical. And the performance numbers are... 00:02:16.189 --> 00:02:18.330 Honestly, kind of surprising for the size. Like 00:02:18.330 --> 00:02:21.169 the 3 billion parameter granite model, it actually 00:02:21.169 --> 00:02:24.210 beats their own older 8 billion parameter model 00:02:24.210 --> 00:02:26.150 in some key areas. Yeah, beats its big brother. 00:02:26.289 --> 00:02:29.090 Right. That's pure efficiency. And it matters 00:02:29.090 --> 00:02:33.090 because, okay, get this, 70 % less memory needed 00:02:33.090 --> 00:02:36.770 to run. 70%. Wow. So businesses can run this 00:02:36.770 --> 00:02:39.849 on hardware they maybe already have or cheaper 00:02:39.849 --> 00:02:43.080 stuff. Right now. And here's something really 00:02:43.080 --> 00:02:45.319 interesting for anyone dealing with lots of data. 00:02:45.419 --> 00:02:48.879 Okay. The core design has linear scaling, meaning 00:02:48.879 --> 00:02:51.219 if you feed it more input, like a huge document, 00:02:51.280 --> 00:02:53.879 massive context, it actually gets faster, not 00:02:53.879 --> 00:02:55.740 slower. Wait, faster with more input. That's 00:02:55.740 --> 00:02:58.000 counterintuitive. I know. But it's a huge win 00:02:58.000 --> 00:03:00.539 for companies just drowning in documents and 00:03:00.539 --> 00:03:03.439 data they need to process. Beat. Big deal. And 00:03:03.439 --> 00:03:05.639 beyond just the efficiency, the sources really 00:03:05.639 --> 00:03:08.479 hammered the security and compliance side. It's 00:03:08.479 --> 00:03:11.300 licensed under Apache 2 .0, which means it's 00:03:11.300 --> 00:03:13.620 properly open. None of those weird restrictive 00:03:13.620 --> 00:03:16.219 clauses we've seen pop up elsewhere. Right. But 00:03:16.219 --> 00:03:18.879 maybe the real kicker for big regulated companies. 00:03:19.159 --> 00:03:22.800 This is the first major open model that has governance 00:03:22.800 --> 00:03:26.699 baked right in from the start. It's ISO 42001 00:03:26.699 --> 00:03:30.139 certified. ISO 42001. Yeah. Think of it like 00:03:30.139 --> 00:03:33.129 a giant pre -approved security blanket. For companies 00:03:33.129 --> 00:03:36.009 where compliance is everything. Okay, hold on, 00:03:36.030 --> 00:03:38.250 though. That ISO certification, that sounds like 00:03:38.250 --> 00:03:39.930 a lot of hoops to jump through, a lot of overhead. 00:03:40.250 --> 00:03:43.629 Is that really the most critical feature here, 00:03:43.689 --> 00:03:46.210 or is it maybe more marketing? No, I think it's 00:03:46.210 --> 00:03:48.030 genuinely critical. I mean, yes, the efficiency 00:03:48.030 --> 00:03:50.650 for edge stuff is also vital, absolutely. But 00:03:50.650 --> 00:03:52.490 for a lot of enterprises, their legal team will 00:03:52.490 --> 00:03:55.270 just shut down any open model project if those 00:03:55.270 --> 00:03:58.340 governance guardrails aren't there. Period. So 00:03:58.340 --> 00:04:00.280 IBM is basically saying, look, we did the really 00:04:00.280 --> 00:04:02.659 hard compliance paperwork. So your lawyers don't 00:04:02.659 --> 00:04:04.919 kill your AI plans. They're jumping into that 00:04:04.919 --> 00:04:07.719 gap, that uncertainty left by, say, Meta's licensing. 00:04:07.979 --> 00:04:10.860 So the goal isn't really to be the most powerful 00:04:10.860 --> 00:04:14.560 model, but the most trustworthy, the vetted, 00:04:14.639 --> 00:04:17.740 secure, efficient option compared to maybe Lama 00:04:17.740 --> 00:04:20.439 or Quinn. So companies can build agents, but 00:04:20.439 --> 00:04:23.050 still keep control. Exactly. Keep control of 00:04:23.050 --> 00:04:25.209 the whole stack. That makes a lot of sense. Okay, 00:04:25.250 --> 00:04:26.790 let's shift gears then. Let's talk about those 00:04:26.790 --> 00:04:28.769 agents. Right, agents. We keep saying the word. 00:04:28.810 --> 00:04:32.829 So a simple definition. An agentic system is 00:04:32.829 --> 00:04:36.420 AI that's designed to do things. to perform actions, 00:04:36.620 --> 00:04:39.620 not just answer a question you type in. They're 00:04:39.620 --> 00:04:42.060 proactive. And this whole trend is moving so 00:04:42.060 --> 00:04:44.259 fast. I mean, just to show how serious this is, 00:04:44.360 --> 00:04:46.240 one of the sources mentioned a senior Google 00:04:46.240 --> 00:04:50.060 engineer just dropped this huge 424 -page document. 00:04:50.420 --> 00:04:53.100 Wow. Free to download called Agentic Design Patterns. 00:04:53.180 --> 00:04:55.540 It lays out how to actually build these things 00:04:55.540 --> 00:04:57.339 properly. That tells you a lot about where the 00:04:57.339 --> 00:04:59.620 focus is shifting. Structure. And the competition 00:04:59.620 --> 00:05:02.079 is fierce. OpenAI, they're releasing something 00:05:02.079 --> 00:05:04.379 called an agent builder, which looks like a direct 00:05:04.379 --> 00:05:07.120 challenge. to tools people already use, like 00:05:07.120 --> 00:05:11.420 Zapier or N8n for workflow automation. Ah, interesting. 00:05:11.600 --> 00:05:13.079 Yeah, I mean, people have been building these 00:05:13.079 --> 00:05:15.339 sorts of wrappers around LLMs for a bit now. 00:05:15.459 --> 00:05:18.699 But when the big platform providers start building 00:05:18.699 --> 00:05:21.810 integrated tools themselves... That's a major 00:05:21.810 --> 00:05:23.910 shakeup for those existing companies, a real 00:05:23.910 --> 00:05:26.170 threat. It does feel like a big leap, though, 00:05:26.230 --> 00:05:28.069 going from just, you know, typing a prompt into 00:05:28.069 --> 00:05:31.149 a box to setting up these complex multi -step 00:05:31.149 --> 00:05:34.209 agent things. I bet a lot of people, even if 00:05:34.209 --> 00:05:36.509 they use AI daily, find that sequence part tricky. 00:05:37.050 --> 00:05:39.810 That's fair. And it points to the challenge right 00:05:39.810 --> 00:05:42.410 now. I'll admit it's still complicated even for 00:05:42.410 --> 00:05:44.709 me sometimes. I still wrestle with prompt drift 00:05:44.709 --> 00:05:47.129 myself sometimes, you know, making those really 00:05:47.129 --> 00:05:50.029 complex multi -step prompts work reliably. Yeah. 00:05:50.540 --> 00:05:52.480 Yeah. It's not always smooth sailing yet. Right. 00:05:52.699 --> 00:05:54.519 Which is why we've got to focus on the practical 00:05:54.519 --> 00:05:57.019 wins first. Exactly. And we found some good stuff 00:05:57.019 --> 00:05:58.980 in the sources on that, like how to use single 00:05:58.980 --> 00:06:01.879 prompts, just one command, in tools like Google's 00:06:01.879 --> 00:06:04.139 Notebook LM to get really specific structured 00:06:04.139 --> 00:06:06.879 stuff out quickly. Cuts through that complexity. 00:06:07.279 --> 00:06:09.339 Yeah. That's where the immediate value is, I 00:06:09.339 --> 00:06:12.220 think. Like imagine getting a full meeting summary 00:06:12.220 --> 00:06:15.360 from a 90 -minute recording with one prompt. 00:06:15.480 --> 00:06:18.100 Or creating a study guide that pulls info from 00:06:18.100 --> 00:06:21.259 three different documents. Or drafting a whole 00:06:21.259 --> 00:06:24.660 30 -day content plan. Boom. Done. These are real, 00:06:24.699 --> 00:06:26.800 actionable things you can do with one command 00:06:26.800 --> 00:06:30.699 that save, like, actual hours. And we're seeing 00:06:30.699 --> 00:06:33.399 these tools pop up right in the daily workflow 00:06:33.399 --> 00:06:36.639 now. Google's apparently put its AI coding agent. 00:06:37.069 --> 00:06:40.350 jewels, they call it, right into terminals and 00:06:40.350 --> 00:06:43.209 slack. Yeah. So it's moving out of the lab into 00:06:43.209 --> 00:06:45.329 the places you actually work. And the money's 00:06:45.329 --> 00:06:47.970 following the practical use cases, too. Like 00:06:47.970 --> 00:06:51.610 Inspiron raised $100 million for an AI platform 00:06:51.610 --> 00:06:54.790 that helps improve care and efficiency in senior 00:06:54.790 --> 00:06:58.129 living facilities. Wow. $100 million. Yeah. Agents 00:06:58.129 --> 00:06:59.750 are going straight to where the real operational 00:06:59.750 --> 00:07:02.569 headaches are, solving actual problems like staffing, 00:07:02.709 --> 00:07:05.240 task management. concrete stuff. So it's really 00:07:05.240 --> 00:07:07.759 that shift from the model answering us to the 00:07:07.759 --> 00:07:09.980 model doing stuff for us. That difference feels 00:07:09.980 --> 00:07:12.639 like the core of this next wave in business automation. 00:07:12.939 --> 00:07:15.079 How fast do you think we'll see this agent tech 00:07:15.079 --> 00:07:17.439 move? Yeah. You know, from just being rappers 00:07:17.439 --> 00:07:20.139 to actually handling complex tasks without a 00:07:20.139 --> 00:07:22.699 human watching every step. Hmm. That's the big 00:07:22.699 --> 00:07:26.060 question. But given the speed things are moving, 00:07:26.160 --> 00:07:29.160 I think automating specific complex business 00:07:29.160 --> 00:07:33.079 tasks feels, well, imminent. It's coming fast. 00:07:33.420 --> 00:07:36.279 OK, let's pull back a bit now. Look at some of 00:07:36.279 --> 00:07:38.120 the big industry headlines that kind of give 00:07:38.120 --> 00:07:40.040 us the vibe of the whole market, the push and 00:07:40.040 --> 00:07:42.220 pull happening. Well, the biggest drama, obviously, 00:07:42.360 --> 00:07:44.939 is still the open AI and Musk thing. Open AI 00:07:44.939 --> 00:07:47.620 really fired back at his lawsuit over trade secrets. 00:07:47.699 --> 00:07:49.660 Oh, yeah. What'd they say? Called it basically 00:07:49.660 --> 00:07:51.740 harassment, a tactic just meant to slow them 00:07:51.740 --> 00:07:54.720 down. So, yeah, that feud is definitely still 00:07:54.720 --> 00:07:58.259 simmering. Hotly. Meanwhile, despite all that 00:07:58.259 --> 00:08:01.259 noise, the money just keeps pouring in. OpenAI 00:08:01.259 --> 00:08:04.079 apparently just officially became the world's 00:08:04.079 --> 00:08:06.040 most valuable private company. Yeah, which just 00:08:06.040 --> 00:08:08.480 underscores the insane level of investment happening 00:08:08.480 --> 00:08:10.839 right now. It is this really intense mix, isn't 00:08:10.839 --> 00:08:13.439 it? Like, massive valuations on one side, and 00:08:13.439 --> 00:08:15.980 on the other you've got Sam Altman himself warning 00:08:15.980 --> 00:08:18.680 the whole AI industry might be heading for a, 00:08:18.720 --> 00:08:21.540 quote, spectacular implosion. That's a pretty 00:08:21.540 --> 00:08:23.519 stark warning. Especially coming from the guy 00:08:23.519 --> 00:08:25.339 leading the most valuable company in the space. 00:08:25.560 --> 00:08:28.560 It's... Yeah. High stakes. What it tells me, 00:08:28.560 --> 00:08:30.879 looking at these top stories, is that everything's 00:08:30.879 --> 00:08:33.500 moving at once. The deep infrastructure stuff 00:08:33.500 --> 00:08:36.080 and the user experience stuff. Like, look at 00:08:36.080 --> 00:08:40.159 NVIDIA. They launched this AI aerial tool, uses 00:08:40.159 --> 00:08:44.700 their GPUs to actively boost 5G and even 6G networks. 00:08:44.879 --> 00:08:46.700 Yeah, they're basically plugging AI straight 00:08:46.700 --> 00:08:50.519 into the core global communication grid. making 00:08:50.519 --> 00:08:52.580 sure they stay dominant at that hardware level. 00:08:52.740 --> 00:08:54.860 And then on the user side, you see Apple making 00:08:54.860 --> 00:08:57.419 big moves, apparently looking outside the company 00:08:57.419 --> 00:09:00.379 for a new AI chief. Which signals they know they 00:09:00.379 --> 00:09:02.759 need fresh talent, maybe need to play catch up 00:09:02.759 --> 00:09:05.200 fast in some areas. These are huge structural 00:09:05.200 --> 00:09:07.159 shifts. And then there's just some interesting 00:09:07.159 --> 00:09:09.799 little details too for flavor. Google pushed 00:09:09.799 --> 00:09:12.259 nano banana into full production. Nano banana. 00:09:12.500 --> 00:09:15.159 Yeah. Apparently it now supports 10 new aspect 00:09:15.159 --> 00:09:18.100 ratios. And this is big for creators. You can 00:09:18.100 --> 00:09:20.769 get just the image out. No need for text prompts 00:09:20.769 --> 00:09:24.490 mixed in. Image only output. Ah, OK. That's useful. 00:09:24.610 --> 00:09:26.490 And you can't ignore the cultural side either. 00:09:26.529 --> 00:09:29.250 The sources mentioned that viral trend recreating 00:09:29.250 --> 00:09:32.269 famous movie scenes. Oh, with Pikachu. Yeah. 00:09:32.389 --> 00:09:35.230 Like Batman or The Godfather, but starring Pikachu. 00:09:35.529 --> 00:09:38.929 It just shows how quickly this high quality generative 00:09:38.929 --> 00:09:42.610 AI stuff is hitting. like mainstream culture. 00:09:42.769 --> 00:09:45.070 It's accelerating. And speaking of accelerating, 00:09:45.269 --> 00:09:48.289 Meta's moving on hardware too. Users are apparently 00:09:48.289 --> 00:09:50.950 testing live navigation features right now on 00:09:50.950 --> 00:09:53.750 their AI Ray -Bans. Whoa. Yeah, imagine getting 00:09:53.750 --> 00:09:56.429 directions prompts right there in your glasses 00:09:56.429 --> 00:09:58.830 as you walk around. So you've got this crazy 00:09:58.830 --> 00:10:02.070 tension, right? Unbelievable valuations, serious 00:10:02.070 --> 00:10:05.100 warnings about a bubble, but... Underneath it 00:10:05.100 --> 00:10:07.759 all, the actual infrastructure is being rebuilt 00:10:07.759 --> 00:10:11.379 super fast. It's wild. Yeah. Does that title, 00:10:11.460 --> 00:10:14.240 most valuable private company, even mean that 00:10:14.240 --> 00:10:16.620 much when the founder is warning about a market 00:10:16.620 --> 00:10:18.940 implosion? Well, it really highlights the conflict, 00:10:19.019 --> 00:10:21.840 doesn't it? Massive investment hype running headlong 00:10:21.840 --> 00:10:24.019 into these deep worries about whether the industry's 00:10:24.019 --> 00:10:26.740 growth is sustainable or just moving too fast. 00:10:26.960 --> 00:10:29.019 Medroll sponsor, Reed Placeholder. Okay, this 00:10:29.019 --> 00:10:31.059 next piece, this quantum breakthrough, it feels 00:10:31.059 --> 00:10:33.899 genuinely like... Profound. We mentioned earlier 00:10:33.899 --> 00:10:36.139 how fragile quantum computers are. Right. The 00:10:36.139 --> 00:10:38.919 13 second limit. Yeah. For years, atom loss just 00:10:38.919 --> 00:10:41.259 meant even the best machines tapped out around 00:10:41.259 --> 00:10:44.440 13 seconds, which is why you couldn't do long, 00:10:44.440 --> 00:10:47.860 complex, continuous calculations. It was scientifically 00:10:47.860 --> 00:10:51.919 just impossible. Physics itself imposed that 00:10:51.919 --> 00:10:54.659 limit. But this Harvard team, it seems like they 00:10:54.659 --> 00:10:56.960 beat it by focusing on continuous operation, 00:10:57.259 --> 00:11:00.000 their solution, actively replacing the atoms 00:11:00.000 --> 00:11:03.600 that get lost in real time. mid -calculation, 00:11:03.600 --> 00:11:06.259 without messing up the quantum state. The way 00:11:06.259 --> 00:11:07.779 they do it is pretty brilliant. They use something 00:11:07.779 --> 00:11:10.360 called an optical lattice, basically. Super -controlled 00:11:10.360 --> 00:11:12.759 light beams holding atoms and combine it with 00:11:12.759 --> 00:11:15.000 optical tweezers. Tweezers made of light. Yeah. 00:11:15.259 --> 00:11:17.919 They grab fresh atoms and just sloth them into 00:11:17.919 --> 00:11:20.399 the empty spots left by the lost ones. And they 00:11:20.399 --> 00:11:22.139 can do it incredibly fast, injecting something 00:11:22.139 --> 00:11:24.480 like 300 ,000 atoms per second back into the 00:11:24.480 --> 00:11:27.080 system. Wow. So it's like a self -healing machine 00:11:27.080 --> 00:11:29.759 but at the atomic level? Exactly. It holds 3 00:11:29.759 --> 00:11:32.100 ,000 qubits, which is already impressive, and 00:11:32.100 --> 00:11:34.360 keeps that delicate quantum state going even 00:11:34.360 --> 00:11:36.240 while this atom refresh is happening constantly 00:11:36.240 --> 00:11:39.100 in the background. And that one change, it just 00:11:39.100 --> 00:11:41.080 unlocks everything that was blocked before, doesn't 00:11:41.080 --> 00:11:44.960 it? Totally. Suddenly. Long, complex calculations. 00:11:45.340 --> 00:11:47.960 They're actually possible. It gives us a clear 00:11:47.960 --> 00:11:51.940 path toward real quantum programs, programs that 00:11:51.940 --> 00:11:54.679 run like, you know, actual robust apps, not just 00:11:54.679 --> 00:11:57.639 these fragile 13 second demos. Yeah. Whoa. I 00:11:57.639 --> 00:12:00.399 mean, just imagine scaling that a billion queries 00:12:00.399 --> 00:12:02.179 maybe. Yeah. When you don't have to constantly 00:12:02.179 --> 00:12:04.740 restart the machine every 13 seconds, that continuous 00:12:04.740 --> 00:12:07.139 power, it just rewrites the whole commercial 00:12:07.139 --> 00:12:09.419 roadmap for quantum. Which is why people are 00:12:09.419 --> 00:12:12.059 calling it. a potential iPhone moment for quantum 00:12:12.059 --> 00:12:14.460 computing. You know, AWS is Harvard's partner 00:12:14.460 --> 00:12:17.600 on this, and you can bet Microsoft, Google, they're 00:12:17.600 --> 00:12:20.000 all watching this incredibly closely. Because 00:12:20.000 --> 00:12:22.320 if you solve the stability problem, you basically 00:12:22.320 --> 00:12:24.360 solve the commercialization problem. Yeah, given 00:12:24.360 --> 00:12:26.820 how fragile things have been. Is this mid -calculation 00:12:26.820 --> 00:12:28.720 repair system, is that the real key? The thing 00:12:28.720 --> 00:12:30.659 that finally unlocks commercial quantum computing? 00:12:30.879 --> 00:12:32.799 It feels like it has to be, right? Continuous 00:12:32.799 --> 00:12:35.200 operation seems like the absolute prerequisite 00:12:35.200 --> 00:12:38.259 for having always -on quantum servers that businesses 00:12:38.259 --> 00:12:41.009 can actually rely on. use widely. It moves it 00:12:41.009 --> 00:12:44.070 from physics lab curiosity to potential enterprise 00:12:44.070 --> 00:12:47.269 tool. Hashtag tag tag outro. OK, so let's pull 00:12:47.269 --> 00:12:49.649 this down. Three key nuggets we pulled out for 00:12:49.649 --> 00:12:52.149 you today. Go for it. First, enterprise ready, 00:12:52.250 --> 00:12:56.490 secure, open source AI is definitely here. IBM's 00:12:56.490 --> 00:12:59.250 Granite 4 .0 is leading that charge with its 00:12:59.250 --> 00:13:02.889 focus on efficiency and crucially, those compliance 00:13:02.889 --> 00:13:06.000 guardrails. Right. Second takeaway, AI agents 00:13:06.000 --> 00:13:07.899 are moving fast. They're automating professional 00:13:07.899 --> 00:13:10.960 work, going beyond simple chat into real tools 00:13:10.960 --> 00:13:13.860 like OpenAI's Agent Builder and those super useful 00:13:13.860 --> 00:13:17.399 single prompt commands for specific tasks. And 00:13:17.399 --> 00:13:20.360 third, quantum computing just took this absolutely 00:13:20.360 --> 00:13:22.440 massive leap. They solved the core instability 00:13:22.440 --> 00:13:26.320 problem that creates a real viable path towards 00:13:26.320 --> 00:13:29.759 continuous, always -on quantum power. Finally. 00:13:30.269 --> 00:13:31.990 It's such an interesting contrast when you put 00:13:31.990 --> 00:13:33.669 it all together from the sources, isn't it? On 00:13:33.669 --> 00:13:36.389 one hand, this huge necessary push for control, 00:13:36.450 --> 00:13:38.169 like the ISO certification for granite, making 00:13:38.169 --> 00:13:40.269 it safe to deploy. And on the other hand, this 00:13:40.269 --> 00:13:42.870 giant leap into almost uncontrolled power with 00:13:42.870 --> 00:13:45.090 these always -on self -healing quantum machines. 00:13:45.330 --> 00:13:47.590 But both roads are really aiming for the same 00:13:47.590 --> 00:13:49.970 thing, ultimate reliability. Yeah, it's that 00:13:49.970 --> 00:13:52.169 moment where fragility finally starts giving 00:13:52.169 --> 00:13:55.210 way to actual functionality. You can see it happening. 00:13:55.450 --> 00:13:57.490 It leaves us with maybe one last thought for 00:13:57.490 --> 00:14:01.149 you to chew on. If quantum machines can now repair 00:14:01.149 --> 00:14:04.690 themselves in real time, constantly to keep running, 00:14:04.850 --> 00:14:07.970 what processes, what systems maybe in your own 00:14:07.970 --> 00:14:10.909 work life are due for that kind of constant self 00:14:10.909 --> 00:14:14.230 -healing upgrade? That's good food for thought. 00:14:14.529 --> 00:14:16.509 Well, thanks for joining us for this deep dive 00:14:16.509 --> 00:14:18.309 into your sources. Thanks, everyone. We'll catch 00:14:18.309 --> 00:14:18.769 you next time.