WEBVTT 00:00:00.000 --> 00:00:01.879 Okay, hey there, and welcome back to the Deep 00:00:01.879 --> 00:00:05.400 Dive. Hey. Today we are just jumping into this 00:00:05.400 --> 00:00:08.380 stack of material you send over articles, research 00:00:08.380 --> 00:00:10.599 notes, a whole bunch of different stuff. And 00:00:10.599 --> 00:00:13.560 it shows us something really interesting. It's 00:00:13.560 --> 00:00:17.199 how AI, you know, the tech that feels like it's 00:00:17.199 --> 00:00:19.699 totally focused on the future. Right. Is actually 00:00:19.699 --> 00:00:22.980 giving us these wild new insights into the ancient 00:00:22.980 --> 00:00:27.280 world, like thousands of years ago ancient. And 00:00:27.280 --> 00:00:29.920 also obviously still totally shaping the tools 00:00:29.920 --> 00:00:33.000 we use right now today. Right. It's kind of fascinating, 00:00:33.159 --> 00:00:37.280 you know, seeing AI's reach span from like uncovering 00:00:37.280 --> 00:00:39.820 secrets from millennia ago to what's getting 00:00:39.820 --> 00:00:42.520 launched this week. Totally. Our mission here 00:00:42.520 --> 00:00:45.619 is to really just unpack the key nuggets from 00:00:45.619 --> 00:00:48.179 this material, figure out what's happening, and 00:00:48.179 --> 00:00:51.280 most importantly, why it actually matters to 00:00:51.280 --> 00:00:54.219 you. Totally. And yeah, there's some surprising 00:00:54.219 --> 00:00:57.020 bits in here. Like apparently... AI might actually 00:00:57.020 --> 00:01:00.420 change the timeline of ancient history, which, 00:01:00.520 --> 00:01:02.619 I don't know, that just sounds kind of mind -bending 00:01:02.619 --> 00:01:04.760 when you think about it. It really does. It highlights 00:01:04.760 --> 00:01:09.060 the capability of these models to find patterns 00:01:09.060 --> 00:01:11.060 and data that, you know, was previously just 00:01:11.060 --> 00:01:13.579 too complex or too large for us to easily process. 00:01:13.939 --> 00:01:16.200 So, yeah, let's dive in. We're starting in the 00:01:16.200 --> 00:01:20.980 way, way back past. Okay. And specifically with 00:01:20.980 --> 00:01:25.200 the Dead Sea Scrolls. Ah, yeah. This is one of 00:01:25.200 --> 00:01:27.299 the most significant findings highlighted in 00:01:27.299 --> 00:01:30.140 your sources. This new research shows how AI 00:01:30.140 --> 00:01:33.760 is actually redating some of these scroll fragments. 00:01:34.099 --> 00:01:36.120 Redating them? Yeah, suggesting they might be 00:01:36.120 --> 00:01:38.780 older than scholars previously thought. Okay, 00:01:38.799 --> 00:01:40.659 real quick, just for anyone who maybe isn't totally 00:01:40.659 --> 00:01:43.000 up to speed, what are the Dead Sea Scrolls? Oh, 00:01:43.000 --> 00:01:46.659 good point. They are essentially fragments of 00:01:46.659 --> 00:01:50.359 ancient Hebrew and Aramaic writings over 2 ,000 00:01:50.359 --> 00:01:52.859 years old. They were found near the Dead Sea 00:01:52.859 --> 00:01:55.579 back in the 1940s. 40s and 50s, obviously. They 00:01:55.579 --> 00:01:58.579 give us this incredible window into Jewish life 00:01:58.579 --> 00:02:00.620 culture and religious thought from that time. 00:02:00.799 --> 00:02:03.099 But the big challenge, as the sources explain, 00:02:03.340 --> 00:02:05.620 is that most of the fragments don't have specific 00:02:05.620 --> 00:02:08.000 dates written on them. So scholars have to use 00:02:08.000 --> 00:02:10.879 other methods to estimate their age. Right. So 00:02:10.879 --> 00:02:13.000 you've got this priceless historical source, 00:02:13.120 --> 00:02:16.400 but nailing down the exact timeline is really 00:02:16.400 --> 00:02:19.379 tricky. Exactly. So, okay, how does AI help with 00:02:19.379 --> 00:02:22.500 that? That's where this machine learning model 00:02:22.500 --> 00:02:26.210 called Enoch. comes in. EMAC. Yeah. The research 00:02:26.210 --> 00:02:29.169 describes how it was trained using two main types 00:02:29.169 --> 00:02:31.990 of data. First, the traditional radiocarbon 14 00:02:31.990 --> 00:02:34.509 dating data. Right. The standard stuff. Which 00:02:34.509 --> 00:02:37.150 is the standard scientific way to date organic 00:02:37.150 --> 00:02:39.969 material like the parchment. And then second, 00:02:40.129 --> 00:02:42.650 information about the handwriting itself. The 00:02:42.650 --> 00:02:44.650 handwriting. Yeah. The geometric and structural 00:02:44.650 --> 00:02:47.129 features like the size, the shape, the curvature 00:02:47.129 --> 00:02:49.449 of the letters. So it's not just about how old 00:02:49.449 --> 00:02:52.110 the paper is, but like the unique style of the 00:02:52.110 --> 00:02:53.830 person who is writing. that's kind of neat yeah 00:02:53.830 --> 00:02:56.189 exactly it's combining the physical age with 00:02:56.189 --> 00:02:59.610 the nuances of the script okay and the name enoch 00:02:59.610 --> 00:03:01.870 the source and point out is you know named after 00:03:01.870 --> 00:03:05.150 this ancient prophet symbolizing that blend of 00:03:05.150 --> 00:03:08.310 ancient wisdom and modern science nice touch 00:03:08.310 --> 00:03:10.469 and they're super clear the tool is designed 00:03:10.469 --> 00:03:13.509 to assist scholars to help refine their estimates 00:03:13.509 --> 00:03:17.090 not to completely replace the human experts right 00:03:17.090 --> 00:03:20.750 it's a tool not a replacement gotcha okay so 00:03:21.199 --> 00:03:24.159 What did Enoch find? Did it just confirm what 00:03:24.159 --> 00:03:26.500 everyone already thought? Well, mostly it did, 00:03:26.580 --> 00:03:28.800 which is good validation. The sources say there 00:03:28.800 --> 00:03:32.180 was about an 80 % match rate between Enoch's 00:03:32.180 --> 00:03:34.500 dating predictions and the human expert's previous 00:03:34.500 --> 00:03:36.860 estimates. Okay, 80 % is pretty high. That's 00:03:36.860 --> 00:03:39.479 a pretty strong agreement. But here's where it 00:03:39.479 --> 00:03:41.680 gets really interesting. Oh, tell me. What happened 00:03:41.680 --> 00:03:44.680 with the other 20 %? That remaining 20%. Enoch 00:03:44.680 --> 00:03:47.460 consistently dated those scrolls older than expected. 00:03:47.759 --> 00:03:50.729 Older? How much older? Sometimes by as much as 00:03:50.729 --> 00:03:53.729 a century. Whoa, 100 years. That's a significant 00:03:53.729 --> 00:03:55.270 difference when you're talking about documents 00:03:55.270 --> 00:03:57.990 from over 2 ,000 years ago. It really is. Like, 00:03:57.990 --> 00:03:59.830 is there a specific example? Yeah, the sources 00:03:59.830 --> 00:04:03.009 mention scroll 4Q114, which contains parts of 00:04:03.009 --> 00:04:05.229 the Book of Daniel. Okay. Human scholars had 00:04:05.229 --> 00:04:09.590 previously dated it to around 165 BCE. Enoch's 00:04:09.590 --> 00:04:13.030 range, though, puts it between 230 and 160 BCE. 00:04:13.669 --> 00:04:16.009 Hmm, so it could be 165, but it could also be... 00:04:16.839 --> 00:04:19.279 quite a bit earlier. Exactly. So while it does 00:04:19.279 --> 00:04:22.699 overlap with the human date, it definitely opens 00:04:22.699 --> 00:04:25.959 up the possibility that this particular text, 00:04:26.139 --> 00:04:28.740 or at least that fragment, could be older, potentially 00:04:28.740 --> 00:04:31.439 significantly older. So it's nudging the earliest 00:04:31.439 --> 00:04:34.779 possible origin. Back in time. Right. And Enoch 00:04:34.779 --> 00:04:37.060 found something even more specific that's impactful 00:04:37.060 --> 00:04:40.300 for scholars. Instances of what they call Herodian 00:04:40.300 --> 00:04:43.199 scripts. Okay. Like King Herod. Yeah. A handwriting 00:04:43.199 --> 00:04:45.779 style previously believed to be characteristic 00:04:45.779 --> 00:04:48.660 of, you know, the time of King Herod actually 00:04:48.660 --> 00:04:51.360 appearing in fragments that the AI dated before 00:04:51.360 --> 00:04:54.670 Herod's reign. before. Wow. Which suggests that 00:04:54.670 --> 00:04:56.730 the transition to this specific writing style 00:04:56.730 --> 00:04:59.310 might have happened earlier than previously understood. 00:04:59.649 --> 00:05:01.970 Okay. So AI is looking at these incredibly subtle 00:05:01.970 --> 00:05:04.310 details of handwriting, combining it with the 00:05:04.310 --> 00:05:06.610 science of carbon dating. And it's basically 00:05:06.610 --> 00:05:09.560 saying, hey. Maybe we need to tweak the historical 00:05:09.560 --> 00:05:12.399 timeline a bit based on this. Pretty much, yeah. 00:05:12.540 --> 00:05:15.360 And the why it matters point here is huge as 00:05:15.360 --> 00:05:18.319 the research highlights. Applying AI to ancient 00:05:18.319 --> 00:05:21.420 script analysis like this can fundamentally reshape 00:05:21.420 --> 00:05:23.519 our understanding of things like cultural evolution, 00:05:23.920 --> 00:05:26.899 how widespread literacy was in different periods, 00:05:27.139 --> 00:05:30.000 and the overall historical context of these texts. 00:05:30.319 --> 00:05:32.819 That makes sense. If texts are older, it changes 00:05:32.819 --> 00:05:35.779 when and how we think certain ideas, beliefs, 00:05:35.899 --> 00:05:38.910 or even just writing styles. spread or developed. 00:05:39.009 --> 00:05:41.110 It's kind of like the future is shedding light 00:05:41.110 --> 00:05:43.709 on the deep past to make it clearer. That's a 00:05:43.709 --> 00:05:46.370 really cool way to think about it. OK, so we 00:05:46.370 --> 00:05:49.089 see AI helping us understand these ancient, ancient 00:05:49.089 --> 00:05:52.810 writings. But I mean, AI is also just moving 00:05:52.810 --> 00:05:55.449 at like lightning speed right now, creating tools 00:05:55.449 --> 00:05:57.990 and driving research that feels. Totally sci 00:05:57.990 --> 00:06:00.389 -fi. Absolutely. The sources really emphasize 00:06:00.389 --> 00:06:03.029 this rapid pace. We go from ancient history, 00:06:03.149 --> 00:06:06.189 you know, to the sort of daily deluge of new 00:06:06.189 --> 00:06:07.930 AI development. It's almost hard to keep up. 00:06:08.029 --> 00:06:10.490 Yeah. Let's shift gears then and look at some 00:06:10.490 --> 00:06:12.889 of the examples of current AI tools and research 00:06:12.889 --> 00:06:14.470 that the sources cover, because it's a pretty 00:06:14.470 --> 00:06:16.870 diverse landscape. It really is. One area that 00:06:16.870 --> 00:06:18.949 felt particularly significant in the material 00:06:18.949 --> 00:06:22.310 is AI's impact in science and medicine. They 00:06:22.310 --> 00:06:25.730 talk about this new AI architecture called bioreason. 00:06:26.050 --> 00:06:28.379 Bioreason. Okay, so what exactly is that? What's 00:06:28.379 --> 00:06:31.600 kind of groundbreaking here is it combines a 00:06:31.600 --> 00:06:34.759 DNA language model with a standard large language 00:06:34.759 --> 00:06:37.860 model, an LLM. Like chat GPT type things? Yeah, 00:06:37.920 --> 00:06:40.199 like, you know, the kind that powers chatbots. 00:06:40.420 --> 00:06:43.259 Oh, so it's like it can understand the language 00:06:43.259 --> 00:06:46.620 of DNA in our normal human language? Yeah, exactly. 00:06:46.879 --> 00:06:49.699 The sources explain that this allows the LLM 00:06:49.699 --> 00:06:53.790 part to... Reason over genomic data, not just 00:06:53.790 --> 00:06:56.490 text data, but the actual genetic code. Okay. 00:06:56.610 --> 00:06:59.389 They tested it on standard biological benchmarks, 00:06:59.769 --> 00:07:01.949 data sets, essentially designed to connect genetic 00:07:01.949 --> 00:07:04.790 changes to disease pathways and biological functions. 00:07:05.149 --> 00:07:07.089 And how did it do? Was it like any good? The 00:07:07.089 --> 00:07:09.269 sources report a significant gain in performance, 00:07:09.449 --> 00:07:12.410 about 15 % better than using models that only 00:07:12.410 --> 00:07:16.290 use the LLM or only use the DNA data. 15 % better. 00:07:16.370 --> 00:07:18.769 That's pretty substantial. It really shows the 00:07:18.769 --> 00:07:21.139 power of tightly integrating these. different 00:07:21.139 --> 00:07:23.060 types of information. Okay, can you give me a 00:07:23.060 --> 00:07:26.040 concrete example of what this bioreason can actually 00:07:26.040 --> 00:07:29.000 do? Something specific. Sure. The sources use 00:07:29.000 --> 00:07:31.860 this example query. What is the impact of a specific 00:07:31.860 --> 00:07:35.500 genetic change, a PFN1 variant found on chromosome 00:07:35.500 --> 00:07:39.060 17? Okay. BioReason was able to predict amyotrophic 00:07:39.060 --> 00:07:41.920 lateral sclerosis, you know, ALS. ALS, okay. 00:07:42.019 --> 00:07:44.459 But it didn't just give ALS as an answer. The 00:07:44.459 --> 00:07:46.819 really powerful part is that it generated a step 00:07:46.819 --> 00:07:49.600 -by -step, 10 -step actually, mechanistic explanation 00:07:49.600 --> 00:07:52.800 of how that specific genetic variant is linked 00:07:52.800 --> 00:07:55.160 to ALS. Wow. Okay, so it's not just spinning 00:07:55.160 --> 00:07:57.399 out a prediction. It's actually showing its work, 00:07:57.500 --> 00:08:00.470 like the biological process behind it. Precisely. 00:08:00.470 --> 00:08:03.189 And the sources really emphasize that this ability 00:08:03.189 --> 00:08:06.649 to provide a reasoned explanation is essential. 00:08:06.949 --> 00:08:10.310 I bet. Especially for things like building trust 00:08:10.310 --> 00:08:14.230 in clinical diagnostics or speeding up drug discovery 00:08:14.230 --> 00:08:16.970 pipelines or even just generating automated hypotheses 00:08:16.970 --> 00:08:19.949 for researchers to explore. It could potentially 00:08:19.949 --> 00:08:22.629 significantly accelerate genetic and medical 00:08:22.629 --> 00:08:25.550 research. So we've gone from dating ancient texts 00:08:25.550 --> 00:08:28.290 by analyzing handwriting and carbon to predicting 00:08:28.290 --> 00:08:30.790 serious diseases by reading DNA and explaining 00:08:30.790 --> 00:08:34.490 the biology. That is quite the range. It really 00:08:34.490 --> 00:08:36.710 is. And then, yeah, your sources dive into the 00:08:36.710 --> 00:08:39.549 sheer volume of new tools and applications that 00:08:39.549 --> 00:08:42.190 are just launching right now. Yeah. It feels 00:08:42.190 --> 00:08:44.610 like, you know, every other day there's something 00:08:44.610 --> 00:08:47.169 new. It really does feel like that. The material 00:08:47.169 --> 00:08:49.730 carters a bunch. Like the recent OpenAI updates, 00:08:49.970 --> 00:08:53.460 you can now... record and transcribe audio right 00:08:53.460 --> 00:08:55.840 there in ChatGPT, right? And it can automatically 00:08:55.840 --> 00:08:58.720 summarize meetings and add timestamps. Plus, 00:08:58.799 --> 00:09:00.419 they added Google Drive integration, which is 00:09:00.419 --> 00:09:02.620 pretty useful. Yeah, those are pretty practical 00:09:02.620 --> 00:09:05.519 additions for everyday use, things people actually 00:09:05.519 --> 00:09:09.649 need. There are these new players popping up, 00:09:09.669 --> 00:09:12.269 too, like this French startup, H -Company. Oh, 00:09:12.269 --> 00:09:14.269 yeah. The source has mentioned they just launched 00:09:14.269 --> 00:09:18.450 three new AI agents, one that's reportedly 92 00:09:18.450 --> 00:09:21.389 % successful at tasks and cheaper than comparable 00:09:21.389 --> 00:09:24.690 models. 92%. Wow. And another that sounds pretty 00:09:24.690 --> 00:09:28.100 advanced, described as a super agent. Which suggests 00:09:28.100 --> 00:09:30.879 we're seeing AI move towards more autonomous, 00:09:30.940 --> 00:09:34.539 specialized helpers, not just general chat interfaces. 00:09:34.899 --> 00:09:38.059 Right. More like agents doing things. And for 00:09:38.059 --> 00:09:40.740 folks interested in creative stuff or like making 00:09:40.740 --> 00:09:43.679 money online, the sources highlight AI creator 00:09:43.679 --> 00:09:46.879 Rory Flynn showing this neat trick to make VO3 00:09:46.879 --> 00:09:49.899 video clips longer using just text prompts. Oh, 00:09:49.899 --> 00:09:53.850 cool. And there are guides on. Using tools like 00:09:53.850 --> 00:09:56.529 Genspark and VO3 specifically for beginners looking 00:09:56.529 --> 00:09:59.029 to offer social media services or, you know, 00:09:59.029 --> 00:10:01.049 build an online income. Yeah. Demonstrating how 00:10:01.049 --> 00:10:03.090 these tools are being immediately leveraged for 00:10:03.090 --> 00:10:05.809 practical, even entrepreneurial purposes. It's 00:10:05.809 --> 00:10:08.429 not just, you know, abstract tech. Totally. There's 00:10:08.429 --> 00:10:10.509 also just a flood of other interesting tools 00:10:10.509 --> 00:10:13.789 mentioned. ID browser for spotting trends and 00:10:13.789 --> 00:10:16.799 startup ideas. Okay. Spine research. to generate 00:10:16.799 --> 00:10:19.419 quick research reports. UISnapper, which can 00:10:19.419 --> 00:10:21.980 turn a screenshot of a user interface into AI 00:10:21.980 --> 00:10:24.539 prompts to rebuild it. That's useful. UISnapper. 00:10:24.720 --> 00:10:27.460 Yeah. And even niche things like Job for Agent, 00:10:27.600 --> 00:10:29.259 which is apparently a job board specifically 00:10:29.259 --> 00:10:32.299 for autonomous AI agents to find tasks. A job 00:10:32.299 --> 00:10:34.600 board for AIs. Okay, that's interesting. Right. 00:10:34.799 --> 00:10:38.259 And RecipeSnap, which can turn a photo of what's 00:10:38.259 --> 00:10:41.070 in your fridge. into recipe suggestions. I mean, 00:10:41.070 --> 00:10:42.649 that's pretty useful. I know, right? I could 00:10:42.649 --> 00:10:44.789 totally use Recipe Snap. I always stare blankly 00:10:44.789 --> 00:10:47.169 into the fridge. Me too. Yeah. And then you have 00:10:47.169 --> 00:10:50.149 the bigger players integrating AI, like HubSpot 00:10:50.149 --> 00:10:53.450 connecting their CRM data to ChatGPT. Right, 00:10:53.509 --> 00:10:56.169 making existing tools smarter. Or Notebook LM, 00:10:56.429 --> 00:10:58.629 adding a new feature for shared notebooks. And 00:10:58.629 --> 00:11:01.190 the sources note that OpenAI gave internet access 00:11:01.190 --> 00:11:03.830 back to its Codex agent, which is designed for 00:11:03.830 --> 00:11:07.210 software engineering tasks. Codex is back online. 00:11:07.490 --> 00:11:10.070 And Mistral launched their own coding client 00:11:10.070 --> 00:11:12.990 called Vibe, which is clearly aimed at competing 00:11:12.990 --> 00:11:16.070 with GitHub Copilot. So, yeah, a lot happening 00:11:16.070 --> 00:11:18.250 on the coding front, too. It's not just about 00:11:18.250 --> 00:11:20.889 discovering ancient secrets or medical breakthroughs. 00:11:20.889 --> 00:11:24.210 It's also totally changing how we build, create 00:11:24.210 --> 00:11:27.240 work and just like. get things done. Definitely. 00:11:27.500 --> 00:11:30.740 The sheer volume of new tools feels like a huge 00:11:30.740 --> 00:11:32.740 part of the landscape described in this material. 00:11:32.879 --> 00:11:34.960 Absolutely. And navigating this landscape brings 00:11:34.960 --> 00:11:37.059 up some important questions and challenges, which 00:11:37.059 --> 00:11:40.620 the sources touch on as well. Like, is AI going 00:11:40.620 --> 00:11:43.419 to make coding skills obsolete or make people 00:11:43.419 --> 00:11:45.440 lazy? That's a question a lot of people have. 00:11:45.580 --> 00:11:47.580 Right. It's a genuine concern, isn't it? If you're 00:11:47.580 --> 00:11:50.340 constantly using an AI assistant to write code 00:11:50.340 --> 00:11:53.039 for you, does it risk dulling your own fundamental 00:11:53.039 --> 00:11:55.539 skills? Yeah. The material acknowledges that 00:11:55.539 --> 00:11:58.899 concern and presents perspectives on how AI should 00:11:58.899 --> 00:12:01.899 be used smartly as a tool to enhance your skills 00:12:01.899 --> 00:12:05.220 and achieve deeper mastery rather than just relying 00:12:05.220 --> 00:12:07.860 on it as a crutch. Yeah, that makes sense. You 00:12:07.860 --> 00:12:09.340 have to learn how to use the tool to make yourself 00:12:09.340 --> 00:12:11.740 better, not just like delegate everything to 00:12:11.740 --> 00:12:14.360 it. Right. What about the maybe less positive 00:12:14.360 --> 00:12:19.059 side, like AI washing? Ah, yes. The sources bring 00:12:19.059 --> 00:12:22.370 up the concept of AI washing, which is... Essentially, 00:12:22.429 --> 00:12:25.549 when companies use the hype around AI to, you 00:12:25.549 --> 00:12:27.710 know, attract investment, boost their image or 00:12:27.710 --> 00:12:30.809 sometimes even justify cutting jobs without actually 00:12:30.809 --> 00:12:33.049 having genuine or sophisticated AI capabilities. 00:12:33.230 --> 00:12:35.870 The builder AI example mentioned is quite striking, 00:12:36.009 --> 00:12:39.330 where humans in India were reportedly posing 00:12:39.330 --> 00:12:43.169 as AI to fulfill tasks. Posing as AI. Yeah. It 00:12:43.169 --> 00:12:45.429 highlights that not everything labeled AI is 00:12:45.429 --> 00:12:47.990 necessarily the real deal. So you have to be 00:12:47.990 --> 00:12:50.590 pretty critical about the claims you hear. Buyer 00:12:50.590 --> 00:12:53.320 beware. kind of that's definitely and then on 00:12:53.320 --> 00:12:55.399 a totally different note the sources also present 00:12:55.399 --> 00:12:58.379 some really optimistic viewpoints like the google 00:12:58.379 --> 00:13:01.039 deep mind ceo's belief that ai will actually 00:13:01.039 --> 00:13:05.100 make us less selfish over time and that um artificial 00:13:05.100 --> 00:13:08.159 general intelligence agi the really smart ai 00:13:08.159 --> 00:13:11.120 yeah human level ai could lead to radical abundance 00:13:11.120 --> 00:13:13.580 and solve these fundamental underlying root node 00:13:13.580 --> 00:13:16.240 problems and you know maybe the next five to 00:13:16.240 --> 00:13:20.080 ten years wow it's a pretty uh A pretty bold 00:13:20.080 --> 00:13:22.000 vision for the near future, five to 10 years. 00:13:22.120 --> 00:13:24.559 It is. And the sources just present that as like 00:13:24.559 --> 00:13:27.299 a possibility or a perspective, right? Not a 00:13:27.299 --> 00:13:29.820 guaranteed future. Right. Okay. It's one possible 00:13:29.820 --> 00:13:32.399 take. Exactly. It's one viewpoint included in 00:13:32.399 --> 00:13:34.740 the material highlighting the wide range of beliefs 00:13:34.740 --> 00:13:36.840 and predictions out there about AI's ultimate 00:13:36.840 --> 00:13:39.799 impact on society. And in terms of like actual 00:13:39.799 --> 00:13:42.669 industry stuff happening. The sources mentioned 00:13:42.669 --> 00:13:45.429 Snorkel AI's recent fundraising, $100 million, 00:13:45.690 --> 00:13:48.730 which puts their valuation at $1 .3 billion. 00:13:49.149 --> 00:13:54.049 Wow. Big money. That just shows significant investment 00:13:54.049 --> 00:13:56.129 is still flowing into companies building specific 00:13:56.129 --> 00:13:59.250 AI products. It does. And the material also touches 00:13:59.250 --> 00:14:04.389 on the emerging legal and questions related to 00:14:04.389 --> 00:14:06.450 data. Oh, yeah, the data stuff. Like the Reddit 00:14:06.450 --> 00:14:09.149 lawsuit against Anthropic, which is related to 00:14:09.149 --> 00:14:11.870 the AI model using Reddit's data for training, 00:14:12.029 --> 00:14:14.350 allegedly without permission or compensation. 00:14:15.049 --> 00:14:18.490 These kinds of cases are just starting to, you 00:14:18.490 --> 00:14:21.090 know, define the legal and ethical boundaries 00:14:21.090 --> 00:14:23.950 around data usage and inaugural property in the 00:14:23.950 --> 00:14:26.649 age of these massive language models. Yeah, that 00:14:26.649 --> 00:14:30.240 whole area. is incredibly complex and starting 00:14:30.240 --> 00:14:32.279 to be figured out. It feels like the law is trying 00:14:32.279 --> 00:14:34.700 to catch up with the tech. Very much so. And 00:14:34.700 --> 00:14:36.320 there was even a little note in there about the 00:14:36.320 --> 00:14:39.460 AI Fire Academy trivia. Just a quick shout out 00:14:39.460 --> 00:14:41.059 again. Yeah, just a small mention from the source. 00:14:41.100 --> 00:14:42.519 It's kind of showing the different communities 00:14:42.519 --> 00:14:45.440 and aspects that make up the current AI world. 00:14:45.700 --> 00:14:48.340 Little bits of culture around it. Exactly. So, 00:14:48.340 --> 00:14:51.000 wow. We've really gone from uncovering secrets 00:14:51.000 --> 00:14:54.399 buried for thousands of years thanks to AI looking 00:14:54.399 --> 00:14:57.440 at ancient handwriting. to wrestling with brand 00:14:57.440 --> 00:15:00.240 new legal questions about data, and even hearing 00:15:00.240 --> 00:15:03.120 visions of a future with AGI -driven abundance. 00:15:03.519 --> 00:15:06.639 It really brings us back to the big question, 00:15:06.740 --> 00:15:09.360 why does all of this, this incredibly diverse 00:15:09.360 --> 00:15:12.779 set of developments, matter to you listening 00:15:12.779 --> 00:15:15.860 right now? That's the core point, isn't it? It 00:15:15.860 --> 00:15:18.580 matters because AI isn't just some abstract futuristic 00:15:18.580 --> 00:15:22.159 concept anymore. Right. It's fundamentally changing 00:15:22.159 --> 00:15:24.740 how we discover things, how we work, how we create, 00:15:24.899 --> 00:15:28.500 how we interact with information and the world 00:15:28.500 --> 00:15:30.240 around us. Right. Whether you're a researcher 00:15:30.240 --> 00:15:32.919 using something like BioReason, a creative looking 00:15:32.919 --> 00:15:35.679 for new tools to express ideas, a student trying 00:15:35.679 --> 00:15:38.080 to process a ton of information, or just navigating 00:15:38.080 --> 00:15:41.159 daily life, AI is impacting the tools available 00:15:41.159 --> 00:15:43.830 to you. It could potentially shift job markets 00:15:43.830 --> 00:15:45.669 and the skills needed. And it's changing how 00:15:45.669 --> 00:15:48.309 information itself is found, processed and understood. 00:15:48.529 --> 00:15:51.769 It's really about staying informed about this 00:15:51.769 --> 00:15:54.389 evolving landscape, seeing the potential applications, 00:15:54.570 --> 00:15:57.029 being aware of the challenges like AI washing 00:15:57.029 --> 00:16:00.929 and recognizing how AI is being applied in ways 00:16:00.929 --> 00:16:03.960 that might genuinely surprise you. Like, you 00:16:03.960 --> 00:16:07.139 know, rewriting ancient timelines. It definitely 00:16:07.139 --> 00:16:09.139 feels like we're just scratching the surface 00:16:09.139 --> 00:16:11.600 of what AI can do, both for uncovering things 00:16:11.600 --> 00:16:13.340 that were previously hidden and for building 00:16:13.340 --> 00:16:15.220 things that have never existed before. Totally 00:16:15.220 --> 00:16:17.600 agreed. It feels like just the beginning. OK, 00:16:17.679 --> 00:16:19.720 so we've taken this deep dive today looking at 00:16:19.720 --> 00:16:24.039 how AI is. unlocking secrets in the Dead Sea 00:16:24.039 --> 00:16:27.639 Scrolls by analyzing everything from radiocarbon 00:16:27.639 --> 00:16:30.659 data to these tiny details in handwriting. Yeah, 00:16:30.720 --> 00:16:33.440 the Enoch stuff. We saw AI reasoning over DNA. 00:16:33.980 --> 00:16:37.019 with bioreason to predict diseases and even offer 00:16:37.019 --> 00:16:39.460 these step -by -step explanations for complex 00:16:39.460 --> 00:16:41.279 medical connections. Which is pretty amazing. 00:16:41.580 --> 00:16:44.559 And we touched on this huge wave of new tools 00:16:44.559 --> 00:16:47.220 that are changing how we code, create, do research, 00:16:47.460 --> 00:16:50.159 and even how we manage what's in our refrigerators 00:16:50.159 --> 00:16:52.159 with things like RecipeSnap. It's been a journey 00:16:52.159 --> 00:16:54.500 for sure, from the deepest past to the immediate 00:16:54.500 --> 00:16:57.139 future, all in one go. Yeah. And here's just 00:16:57.139 --> 00:17:00.220 like a final thought for you to... Maybe mull 00:17:00.220 --> 00:17:02.320 over. Think about how these examples we discussed, 00:17:02.580 --> 00:17:05.400 whether it's AI redating ancient scrolls based 00:17:05.400 --> 00:17:08.680 on really subtle script variations or predicting 00:17:08.680 --> 00:17:11.720 a complex disease based on intricate patterns 00:17:11.720 --> 00:17:14.900 in DNA. They all show AI's pretty incredible 00:17:14.900 --> 00:17:17.740 power to find meaningful connections and patterns 00:17:17.740 --> 00:17:20.380 in data that was just previously too vast, too 00:17:20.380 --> 00:17:22.420 complex or too nuanced for humans to process 00:17:22.420 --> 00:17:24.910 at scale. Right. Seeing the unseen connections. 00:17:25.049 --> 00:17:27.650 Exactly. What does this fundamental ability of 00:17:27.650 --> 00:17:30.509 AI to reveal these hidden structures across wildly 00:17:30.509 --> 00:17:32.529 different areas, history, biology, user interfaces, 00:17:32.730 --> 00:17:35.170 whatever mean for how we'll understand and interact 00:17:35.170 --> 00:17:39.690 with everything going forward? It's kind of mind 00:17:39.690 --> 00:17:40.130 -blowing, right?