WEBVTT 00:00:00.000 --> 00:00:02.640 Imagine AI not just, you know, answering your 00:00:02.640 --> 00:00:04.379 questions, but actually doing things for you 00:00:04.379 --> 00:00:06.780 online. Stuff like clicking buttons, filling 00:00:06.780 --> 00:00:10.660 in forms, navigating websites. Beat. But what 00:00:10.660 --> 00:00:13.439 happens if it clicks the wrong button? That's 00:00:13.439 --> 00:00:16.160 the really fascinating, sometimes a little unnerving 00:00:16.160 --> 00:00:18.670 edge we're exploring today. Welcome back to the 00:00:18.670 --> 00:00:20.449 Deep Dive, everyone. Yeah, we've got a whole 00:00:20.449 --> 00:00:23.230 stack of fresh insights from the AI world today. 00:00:23.649 --> 00:00:25.649 And we're just going to jump right in. Our mission 00:00:25.649 --> 00:00:27.250 really is to give you the clearest picture we 00:00:27.250 --> 00:00:29.929 can of what's happening right now. So we'll look 00:00:29.929 --> 00:00:32.509 at AI agents kind of taking over our browsers. 00:00:32.590 --> 00:00:34.829 We'll touch on the real world impact breakthroughs, 00:00:34.829 --> 00:00:36.950 ethical stuff. And there's this groundbreaking 00:00:36.950 --> 00:00:39.710 tool for AI generated long audio. It's going 00:00:39.710 --> 00:00:42.210 to be a good one. Definitely. Let's start with 00:00:42.210 --> 00:00:44.549 something that feels genuinely new, maybe even 00:00:44.549 --> 00:00:47.570 a bit mind -bending. Anthropic's new Claude for 00:00:47.570 --> 00:00:50.350 Chrome extension. And this isn't just another 00:00:50.350 --> 00:00:52.929 chatbot, right? It feels like a big step into 00:00:52.929 --> 00:00:55.869 what people are calling agentic AI. So when we 00:00:55.869 --> 00:00:58.429 say agentic AI, we mean AI that can, well... 00:00:59.030 --> 00:01:01.570 actively understand where it is, set goals, and 00:01:01.570 --> 00:01:03.630 then actually do tasks to reach those goals, 00:01:03.689 --> 00:01:05.530 not just talk about them. It's really moving 00:01:05.530 --> 00:01:07.950 from being a chat partner to an active digital 00:01:07.950 --> 00:01:10.409 helper. Exactly. And Cloud for Chrome, it really 00:01:10.409 --> 00:01:13.609 shows that off. It can literally see your active 00:01:13.609 --> 00:01:16.269 browser tab. It processes the content and then 00:01:16.269 --> 00:01:18.189 it interacts with it. Think about that. Clicking 00:01:18.189 --> 00:01:20.109 buttons, filling out forms for you, navigating 00:01:20.109 --> 00:01:22.730 pages. It's proactive, you know, like having 00:01:22.730 --> 00:01:24.530 a super smart intern living in your browser. 00:01:25.719 --> 00:01:28.219 And Anthropic's not alone here. You've got Perplexity, 00:01:28.260 --> 00:01:30.260 OpenAI, Google, they're all racing to get their 00:01:30.260 --> 00:01:32.620 agents into this browser space. It feels like 00:01:32.620 --> 00:01:35.439 a new frontier opening up. It really does raise 00:01:35.439 --> 00:01:38.079 this pretty profound question. Are our browsers 00:01:38.079 --> 00:01:41.799 basically becoming the new operating system for 00:01:41.799 --> 00:01:45.620 these intelligent agents? Two sec silence. I 00:01:45.620 --> 00:01:47.700 mean, for years, the OS was the core, right? 00:01:47.900 --> 00:01:51.079 But now you've got these AI agents living and 00:01:51.079 --> 00:01:53.120 working right in the browser, understanding your 00:01:53.120 --> 00:01:55.280 digital world through that window. It feels like 00:01:55.280 --> 00:01:57.319 a massive shift in how we'll interact with the 00:01:57.319 --> 00:01:59.180 web. The browser isn't just a window anymore. 00:01:59.260 --> 00:02:01.400 It's like an active participant. But yeah, with 00:02:01.400 --> 00:02:04.140 all that power comes some very real risk. It's 00:02:04.140 --> 00:02:06.560 fascinating, sure, but these agetic AIs like 00:02:06.560 --> 00:02:08.819 Claude. They can be tricked. We're talking about 00:02:08.819 --> 00:02:10.400 something called prompt injection. It's kind 00:02:10.400 --> 00:02:13.039 of like a digital Trojan horse. Basically, bad 00:02:13.039 --> 00:02:15.340 actors can hide commands inside a web page's 00:02:15.340 --> 00:02:18.180 text or kind of sneaky stuff. Telling the AI, 00:02:18.400 --> 00:02:20.879 hey, delete all emails or subtly send over this 00:02:20.879 --> 00:02:23.500 personal data. And the AI sees these hidden instructions 00:02:23.500 --> 00:02:25.740 and because it's built to be helpful, well, it 00:02:25.740 --> 00:02:27.599 might just do it. Which is obviously terrifying. 00:02:28.269 --> 00:02:30.849 But Anthropic seems to have taken these vulnerabilities 00:02:30.849 --> 00:02:34.189 incredibly seriously. They put Claude through, 00:02:34.330 --> 00:02:37.849 I think it was 123 adversarial tests specifically 00:02:37.849 --> 00:02:40.229 designed to try and trick it with these prompt 00:02:40.229 --> 00:02:43.090 injections. And what they found and then fixed 00:02:43.090 --> 00:02:45.770 is pretty telling. They significantly improved 00:02:45.770 --> 00:02:48.550 its resistance. The general prompt injection 00:02:48.550 --> 00:02:51.930 success rate went from about 23 .6 percent down 00:02:51.930 --> 00:02:55.069 to just 11 .2 percent. That's a big drop. Yeah. 00:02:55.169 --> 00:02:57.990 And even better for the attacks. Specifically 00:02:57.990 --> 00:03:00.129 targeting the browser interaction, that success 00:03:00.129 --> 00:03:03.550 rate dropped from over 35 % down to a remarkable 00:03:03.550 --> 00:03:08.620 0%. Wow, 0%. Zero. So Claude now actively scans 00:03:08.620 --> 00:03:10.900 for these sneaky patterns, not just in the text 00:03:10.900 --> 00:03:13.439 you see, but in the website's underlying structure, 00:03:13.560 --> 00:03:16.479 the DOM, you know, and the URLs and the tap titles. 00:03:16.580 --> 00:03:18.580 It's like it's learning to look beyond the surface. 00:03:18.800 --> 00:03:20.840 Right. Getting smarter about context. So the 00:03:20.840 --> 00:03:22.780 big question for me and maybe for you, too, is 00:03:22.780 --> 00:03:24.740 how do we balance this incredible usefulness 00:03:24.740 --> 00:03:27.340 with like ironclad security, especially as these 00:03:27.340 --> 00:03:29.939 agents get more capable and more woven into our 00:03:29.939 --> 00:03:32.500 digital lives? Yeah, it's a constant race, isn't 00:03:32.500 --> 00:03:35.120 it? It really demands developers build in proactive. 00:03:35.819 --> 00:03:38.659 AI -powered threat detection right into the agent's 00:03:38.659 --> 00:03:41.509 core. It's not just about blocking known threats 00:03:41.509 --> 00:03:43.710 anymore. It's about teaching the AI to really 00:03:43.710 --> 00:03:45.969 think critically about context and intention 00:03:45.969 --> 00:03:48.550 before it acts on any command. That's a really 00:03:48.550 --> 00:03:50.949 good point, building trust right into the architecture 00:03:50.949 --> 00:03:55.030 itself. Okay, so let's shift gears a little. 00:03:55.110 --> 00:03:57.689 Look at some of the broader impacts AI is having 00:03:57.689 --> 00:04:00.449 societally and in different industries. Absolutely. 00:04:00.509 --> 00:04:02.810 So on the creative side, something pretty neat 00:04:02.810 --> 00:04:04.849 just dropped from Google. It's called Nano Banana. 00:04:05.129 --> 00:04:07.560 Nano Banana. Yeah. Funny name. It's part of their 00:04:07.560 --> 00:04:10.439 Gemini 2 .5 model. And what it does is it keeps 00:04:10.439 --> 00:04:12.860 real faces consistent, even if you heavily edit 00:04:12.860 --> 00:04:15.979 an image or a video, which for creators is huge, 00:04:16.120 --> 00:04:18.910 right? Opens up all sorts of possibilities for 00:04:18.910 --> 00:04:21.230 manipulating media, but keeping subjects looking 00:04:21.230 --> 00:04:24.069 like themselves. Imagine the time saved for animators, 00:04:24.110 --> 00:04:26.589 designers. That does sound useful. But then almost 00:04:26.589 --> 00:04:28.089 immediately we run into the really difficult 00:04:28.089 --> 00:04:30.110 side of things. We're dealing with this tragic 00:04:30.110 --> 00:04:33.189 case. A 16 -year -old asked ChatGPT for help 00:04:33.189 --> 00:04:36.029 with suicidal thoughts and just devastatingly 00:04:36.029 --> 00:04:38.610 got dangerous advice. His parents are now suing 00:04:38.610 --> 00:04:41.709 OpenAI for a wrongful death. And it just highlights 00:04:41.709 --> 00:04:46.160 this urgent, critical safety issue in AI. Vulnerable 00:04:46.160 --> 00:04:48.540 admission. You know, I still wrestle with prompt 00:04:48.540 --> 00:04:50.879 drift myself sometimes where you see AI responses 00:04:50.879 --> 00:04:54.740 just go off track unexpectedly, especially in 00:04:54.740 --> 00:04:57.240 incredibly sensitive areas like mental health. 00:04:57.360 --> 00:04:59.379 It's just a stark reminder that no matter how 00:04:59.379 --> 00:05:02.079 advanced this gets, the human safety net, the 00:05:02.079 --> 00:05:04.620 oversight, it has to be paramount, especially 00:05:04.620 --> 00:05:06.500 when lives could be at stake. Absolutely. It's 00:05:06.500 --> 00:05:08.560 a heavy responsibility. Then switching to the 00:05:08.560 --> 00:05:10.180 business side, we're seeing some fascinating 00:05:10.180 --> 00:05:12.040 and sometimes pretty contentious moves around 00:05:12.040 --> 00:05:15.449 AI content and data. Perplexity, for instance, 00:05:15.629 --> 00:05:18.050 launched something called Comet Plus, which feels 00:05:18.050 --> 00:05:19.569 like a really progressive step. They're actually 00:05:19.569 --> 00:05:21.750 paying content creators whose work shows up in 00:05:21.750 --> 00:05:25.250 AI results. 80 % revenue share, apparently. 80%, 00:05:25.250 --> 00:05:29.209 wow. Significant. A real move towards more ethical 00:05:29.209 --> 00:05:32.610 data sourcing, maybe. But then, not all data 00:05:32.610 --> 00:05:35.389 gathering is that transparent, right? We're also 00:05:35.389 --> 00:05:37.470 hearing reports that Chad GPT might have been 00:05:37.470 --> 00:05:39.790 secretly scraping Google's search for real -time 00:05:39.790 --> 00:05:42.649 info, even after Google explicitly blocked them. 00:05:43.040 --> 00:05:45.939 So it still feels a bit like the Wild West when 00:05:45.939 --> 00:05:48.779 it comes to... how some of this data gets acquired. 00:05:49.019 --> 00:05:52.040 The Wild West. And regulation is trying to catch 00:05:52.040 --> 00:05:54.079 up or maybe companies are trying to shape it 00:05:54.079 --> 00:05:56.779 first. Meta isn't waiting for D .C. apparently. 00:05:56.959 --> 00:06:00.660 They're launching a pro -AI super, you know, 00:06:00.660 --> 00:06:02.879 a political action committee. They want to back 00:06:02.879 --> 00:06:04.980 state level candidates who favor light touch 00:06:04.980 --> 00:06:07.319 AI rules. Interesting. Shaving it from the ground 00:06:07.319 --> 00:06:09.819 up. Exactly. It shows how much money and influence 00:06:09.819 --> 00:06:12.120 are now pouring into defining this landscape. 00:06:12.339 --> 00:06:14.959 And meanwhile, in finance, JPMorgan Chase just 00:06:14.959 --> 00:06:18.000 put half a billion dollars. $500 million into 00:06:18.000 --> 00:06:20.579 Numerai. That's an AI hedge fund. Whoa. Yeah. 00:06:20.699 --> 00:06:23.199 Shows serious institutional belief in AI's power 00:06:23.199 --> 00:06:25.620 and finance. The big money clearly sees a future 00:06:25.620 --> 00:06:27.579 there. And just a few other quick hits showing 00:06:27.579 --> 00:06:30.160 how broad this is getting. Google Translate now 00:06:30.160 --> 00:06:33.899 does live translation in over 70 languages. Super 00:06:33.899 --> 00:06:36.740 useful. On the legal side, Anthropic settled 00:06:36.740 --> 00:06:39.600 that big AI copyright lawsuit with book authors. 00:06:39.860 --> 00:06:43.139 That sets an important precedent for IP in the 00:06:43.139 --> 00:06:46.600 AI era. And showing a focus on safety, various 00:06:46.600 --> 00:06:48.800 state attorneys general signed a letter pushing 00:06:48.800 --> 00:06:50.959 for better protection for kids from potentially 00:06:50.959 --> 00:06:54.279 harmful AI chatbots. OK, so considering all these 00:06:54.279 --> 00:06:56.360 different impacts, creative, ethical, business, 00:06:56.480 --> 00:07:00.040 legal, safety, what's maybe one really crucial 00:07:00.040 --> 00:07:02.759 step we as a society need to take to make sure 00:07:02.759 --> 00:07:04.939 the ethical innovation keeps pace with the growing 00:07:04.939 --> 00:07:07.660 risks? We absolutely need clear, enforceable 00:07:07.660 --> 00:07:10.360 rules, boundaries and accountability for the 00:07:10.360 --> 00:07:12.740 folks developing and deploying AI, especially 00:07:12.740 --> 00:07:15.139 where the stakes are high. Clear boundaries and 00:07:15.139 --> 00:07:16.860 accountability. That makes a lot of sense. It 00:07:16.860 --> 00:07:19.139 puts responsibility where it needs to be. Okay, 00:07:19.180 --> 00:07:20.600 now let's talk about something that sounds like 00:07:20.600 --> 00:07:22.360 it could completely change the game for content 00:07:22.360 --> 00:07:24.720 creators. Oh, yeah. This next one is really exciting, 00:07:24.819 --> 00:07:27.339 especially for audio. Microsoft just released 00:07:27.339 --> 00:07:30.259 something called Vibe Voice, and it's open source. 00:07:30.439 --> 00:07:32.000 Now, this isn't just generating little sound 00:07:32.000 --> 00:07:34.920 bites. It can create 90 minutes of multi -speaker 00:07:34.920 --> 00:07:38.579 audio, up to four distinct voices. And get this 00:07:38.579 --> 00:07:40.339 right from your regular consumer device, your 00:07:40.339 --> 00:07:43.459 laptop, maybe even your phone eventually. Incredible 00:07:43.459 --> 00:07:46.759 accessible. 90 minutes, four voices. That's impressive. 00:07:47.800 --> 00:07:50.319 The quality must be the key thing, though. And 00:07:50.319 --> 00:07:52.680 they say Vibe Voice produces like podcast quality 00:07:52.680 --> 00:07:55.339 dialogue with individual speeder identities that 00:07:55.339 --> 00:07:57.800 sound natural. Apparently so. And beyond that, 00:07:57.899 --> 00:08:00.959 it also compresses the audio 80 times more efficiently 00:08:00.959 --> 00:08:03.819 than standard models, which makes it super lightweight. 00:08:03.959 --> 00:08:05.759 Meaning, like you said, you don't need a giant 00:08:05.759 --> 00:08:08.720 server farm to run it. It can work locally. Exactly. 00:08:08.800 --> 00:08:11.060 It uses advanced language models like Quen 2 00:08:11.060 --> 00:08:14.459 .5 to handle natural turn -taking and keep track 00:08:14.459 --> 00:08:16.959 of the context, even over these long, complex 00:08:16.959 --> 00:08:19.519 conversations. Plus, and this is important, it 00:08:19.519 --> 00:08:22.639 includes AI -generated disclaimers and hidden 00:08:22.639 --> 00:08:24.759 watermarks for transparency. Right. So you know 00:08:24.759 --> 00:08:26.620 it's AI -generated. Honestly, it's like having 00:08:26.620 --> 00:08:30.029 AI talk radio in a box. It just... democratizes 00:08:30.029 --> 00:08:32.090 high quality audio production like never before. 00:08:32.269 --> 00:08:34.389 You know, that's a really good point. Most open 00:08:34.389 --> 00:08:36.830 source text to speech models right now, they 00:08:36.830 --> 00:08:39.169 maybe top out at two speakers and usually just 00:08:39.169 --> 00:08:41.649 for short clips, definitely not long form dialogue. 00:08:42.029 --> 00:08:45.029 So this vibe voice, it really moves the needle, 00:08:45.169 --> 00:08:48.330 pushes us towards a future of full AI panels, 00:08:48.350 --> 00:08:50.490 not just a single AI narrator. Moment of wonder, 00:08:50.590 --> 00:08:53.210 whoa. Just imagine scaling that. You could have, 00:08:53.230 --> 00:08:55.250 what, a billion hours of custom audio content 00:08:55.250 --> 00:08:57.789 generated daily, all running on personal devices 00:08:57.789 --> 00:09:00.230 tailored to what each person wants to hear. The 00:09:00.230 --> 00:09:02.090 creative explosion from that, it would just be 00:09:02.090 --> 00:09:04.309 immense. It's a massive unlock for indie developers, 00:09:04.610 --> 00:09:07.789 creators, everyone. Yeah, absolutely. So what 00:09:07.789 --> 00:09:10.070 does this powerful new capability really mean 00:09:10.070 --> 00:09:13.009 for human creativity and just the very nature 00:09:13.009 --> 00:09:15.269 of how content gets made going forward? I think 00:09:15.269 --> 00:09:17.750 it means human creativity can now scale exponentially. 00:09:18.350 --> 00:09:21.230 Right. A single creator suddenly has the power 00:09:21.230 --> 00:09:23.710 of a full studio. It could totally transform 00:09:23.710 --> 00:09:27.289 how quickly and affordably we produce rich, multi 00:09:27.289 --> 00:09:30.570 -voice audio experiences. The individual as an 00:09:30.570 --> 00:09:32.470 entire production house. That's a fascinating 00:09:32.470 --> 00:09:34.789 way to put it. OK, so to sort of wrap things 00:09:34.789 --> 00:09:38.009 up, let's look at how we as individuals can navigate 00:09:38.009 --> 00:09:40.909 all this. How do we leverage this rapidly changing 00:09:40.909 --> 00:09:45.190 AI landscape? Well, beyond just. creating, using 00:09:45.190 --> 00:09:47.610 these powerful AIs effectively like the upcoming 00:09:47.610 --> 00:09:50.570 GPT -5. It demands a whole new approach. It's 00:09:50.570 --> 00:09:52.909 about shifting your mindset, really. Moving from 00:09:52.909 --> 00:09:55.490 just having a casual chat with the AI to giving 00:09:55.490 --> 00:09:57.769 it precise commands. People talk about like 11 00:09:57.769 --> 00:09:59.909 essential tactics to get the best results. You 00:09:59.909 --> 00:10:01.769 have to learn its language in a way. Become more 00:10:01.769 --> 00:10:03.590 of an architect of the prompts rather than just 00:10:03.590 --> 00:10:06.549 a user. Exactly. And for specific fields like 00:10:06.549 --> 00:10:08.750 marketers or SaaS founders, there are already 00:10:08.750 --> 00:10:11.570 detailed guides popping up. How to strategically 00:10:11.570 --> 00:10:14.830 get your brand noticed. by chat GPT, for instance. 00:10:14.950 --> 00:10:17.350 It's not just luck. It's about crafting your 00:10:17.350 --> 00:10:19.970 content, your outreach. So the AI actually picks 00:10:19.970 --> 00:10:23.940 it up and prioritizes it. Think of it like. carefully 00:10:23.940 --> 00:10:26.240 stacking Lego blocks of data and information, 00:10:26.559 --> 00:10:28.980 making it easy for the AI to find, understand, 00:10:29.100 --> 00:10:31.259 and then use in its answers. It's like a whole 00:10:31.259 --> 00:10:34.480 new kind of SEO. A new SEO game. Interesting. 00:10:34.820 --> 00:10:36.860 And as we learn to command these things better, 00:10:36.980 --> 00:10:40.379 we're also seeing this wave of new, really empowered 00:10:40.379 --> 00:10:43.340 AI tools for specific tasks. You mentioned Tavis 00:10:43.340 --> 00:10:45.860 building AI humans that can see, hear, respond 00:10:45.860 --> 00:10:48.700 in real time, like virtual teammates. Yeah, blurring 00:10:48.700 --> 00:10:51.019 the lines. And Doxy turning notes into documentation 00:10:51.019 --> 00:10:53.200 websites. That sounds incredibly useful. for 00:10:53.200 --> 00:10:55.639 anyone drowning in information, radar for tracking 00:10:55.639 --> 00:10:59.460 app insights, and mini -CPM v4 .5, a vision model 00:10:59.460 --> 00:11:02.379 that's supposedly GPT -40 level but runs right 00:11:02.379 --> 00:11:04.960 on your phone. Crazy, right? The access to this 00:11:04.960 --> 00:11:07.240 kind of power is just democratizing so fast. 00:11:07.460 --> 00:11:10.259 It really is. So in this landscape that's expanding 00:11:10.259 --> 00:11:12.940 so quickly, what do you think is the most critical 00:11:12.940 --> 00:11:15.759 skill for people to cultivate to navigate these 00:11:15.759 --> 00:11:18.759 new tools and possibilities effectively? I think 00:11:18.759 --> 00:11:21.440 it has to be continual, adaptive learning. Just 00:11:21.440 --> 00:11:24.320 constantly learning and coupling that with really 00:11:24.320 --> 00:11:26.740 rigorous critical thinking to understand these 00:11:26.740 --> 00:11:29.799 tools, use them ethically, and see their potential 00:11:29.799 --> 00:11:33.440 and their limits. Sponsor. So reflecting on everything 00:11:33.440 --> 00:11:35.519 we've covered today, what really stands out is 00:11:35.519 --> 00:11:38.419 just the undeniable speed, the acceleration of 00:11:38.419 --> 00:11:41.120 AI everywhere. We've clearly moved past simple 00:11:41.120 --> 00:11:44.649 chatbots. We're into truly agentic AI now systems 00:11:44.649 --> 00:11:47.389 that can take actions for us right inside our 00:11:47.389 --> 00:11:49.529 digital spaces. Yeah, and this incredible power, 00:11:49.590 --> 00:11:52.250 it brings amazing opportunities for efficiency, 00:11:52.429 --> 00:11:54.490 for creativity, but it also brings immediate 00:11:54.490 --> 00:11:56.450 complex challenges. We're seeing the ethical 00:11:56.450 --> 00:11:59.149 dilemmas, the legal fights over data, and this 00:11:59.149 --> 00:12:01.649 really urgent need for rock -solid security. 00:12:01.850 --> 00:12:03.669 It's a constant balancing act that everyone's 00:12:03.669 --> 00:12:06.090 trying to figure out. And we're definitely entering 00:12:06.090 --> 00:12:09.250 an era where AI can generate really sophisticated 00:12:09.250 --> 00:12:12.330 long -form content, even things like multi -speaker 00:12:12.330 --> 00:12:14.389 podcasts potentially running right on your own 00:12:14.389 --> 00:12:17.419 device. The creative landscape isn't just shifting. 00:12:17.519 --> 00:12:19.740 It feels like it's being totally redefined right 00:12:19.740 --> 00:12:22.779 now. For sure. And the future, it isn't just 00:12:22.779 --> 00:12:25.259 about passively using AI anymore, is it? It's 00:12:25.259 --> 00:12:27.620 about learning how to command it, how to navigate 00:12:27.620 --> 00:12:30.759 all its complexities, and really trying to understand 00:12:30.759 --> 00:12:34.179 its impact across, well, every part of our lives. 00:12:34.299 --> 00:12:36.779 That feels like an essential skill set for everyone 00:12:36.779 --> 00:12:39.179 now, whether you're a creator, a business leader, 00:12:39.340 --> 00:12:41.899 or just curious about what's happening. That's 00:12:41.899 --> 00:12:45.039 well put. So as these AI agents... become more 00:12:45.039 --> 00:12:46.779 and more intertwined with our digital lives? 00:12:47.320 --> 00:12:49.799 Learning to coexist with them, not just use them, 00:12:49.860 --> 00:12:52.200 feels like our next big challenge. What does 00:12:52.200 --> 00:12:54.879 a truly symbiotic digital future where humans 00:12:54.879 --> 00:12:57.360 and AI agents work effectively side by side, 00:12:57.480 --> 00:12:59.480 what does that look like to you? Something to 00:12:59.480 --> 00:13:01.039 think about. Definitely something to think about. 00:13:01.279 --> 00:13:03.259 Well, thank you for joining us on this deep dive 00:13:03.259 --> 00:13:05.559 into the latest in AI. Keep digging, keep learning, 00:13:05.620 --> 00:13:08.299 and stay curious. Yeah, we really hope this deep 00:13:08.299 --> 00:13:10.320 dive gave you some clarity, maybe sparked even 00:13:10.320 --> 00:13:13.379 more curiosity. Until next time, stay well informed.