WEBVTT 00:00:00.000 --> 00:00:01.459 You know that feeling, right? Like you're just 00:00:01.459 --> 00:00:04.459 drowning in information. Oh, yeah. Articles, 00:00:04.500 --> 00:00:09.039 reports, videos, social media. It's just stuff 00:00:09.039 --> 00:00:11.359 everywhere. It is. It's overwhelming. And trying 00:00:11.359 --> 00:00:13.960 to feel genuinely well -informed like you've 00:00:13.960 --> 00:00:16.899 really got it without spending hours and hours 00:00:16.899 --> 00:00:20.079 sifting through it all. Man, it's tough. It really 00:00:20.079 --> 00:00:23.280 is. And that overwhelm can actually stop you 00:00:23.280 --> 00:00:25.460 from even starting sometimes. Totally. You see 00:00:25.460 --> 00:00:28.219 the pile and you just think, where do I even 00:00:28.219 --> 00:00:32.990 begin? Exactly. But what if, what if there were, 00:00:32.990 --> 00:00:35.770 you know, new tools coming out that could actually 00:00:35.770 --> 00:00:38.850 help cut through all that noise? Tools to help 00:00:38.850 --> 00:00:40.789 manage it. Yeah. Get you to the valuable insights 00:00:40.789 --> 00:00:43.409 after specifically from your stack of sources. 00:00:43.549 --> 00:00:45.070 It's kind of what we're diving into today. Okay. 00:00:45.130 --> 00:00:47.649 We've got this source that talks about two specific 00:00:47.649 --> 00:00:51.270 AI research tools, Perflexity and Notebook LM, 00:00:51.509 --> 00:00:54.409 and how like you can actually use them together. 00:00:54.630 --> 00:00:57.509 Ah, so a combo approach. Right. To process information 00:00:57.509 --> 00:00:59.619 you bring to the table. What's interesting there 00:00:59.619 --> 00:01:01.979 is the idea of not just one tool, but a workflow, 00:01:02.240 --> 00:01:04.680 you know, leveraging their different strengths 00:01:04.680 --> 00:01:08.120 to get deep into your material. Yeah. So the 00:01:08.120 --> 00:01:10.239 mission for this deep dive is really to unpack 00:01:10.239 --> 00:01:12.799 this source, right? To do it. To show you how, 00:01:12.900 --> 00:01:15.390 according to this article. These specific tools 00:01:15.390 --> 00:01:17.810 can help you take your sources, whether you're 00:01:17.810 --> 00:01:19.469 prepping for a meeting, learning something totally 00:01:19.469 --> 00:01:23.109 new, or doing research for really anything. Right, 00:01:23.209 --> 00:01:25.230 any kind of knowledge work. Gather more around 00:01:25.230 --> 00:01:27.989 it, analyze your core material, and ultimately 00:01:27.989 --> 00:01:31.189 boost whatever project you're working on by finding 00:01:31.189 --> 00:01:33.870 those key insights within that material. It sounds 00:01:33.870 --> 00:01:38.329 like it's about efficiency, but also focus. Getting 00:01:38.329 --> 00:01:40.069 insights from the documents you actually care 00:01:40.069 --> 00:01:42.230 about. Exactly. Without getting lost in the wider 00:01:42.230 --> 00:01:45.909 web or having the AI, well, make stuff up. Right. 00:01:46.010 --> 00:01:48.489 The hallucination problem. And the core concept 00:01:48.489 --> 00:01:52.549 this source teases is pretty simple, but I think 00:01:52.549 --> 00:01:56.030 powerful. One tool for casting a wide net, helping 00:01:56.030 --> 00:01:57.829 you gather sources. Yeah, as a discovery phase. 00:01:58.109 --> 00:02:01.129 And the other for really focused analysis using 00:02:01.129 --> 00:02:03.299 only the material you brought in. Which directly 00:02:03.299 --> 00:02:06.099 tackles that major issue with large language 00:02:06.099 --> 00:02:09.699 models, the, you know, hallucinations. Definitely. 00:02:09.780 --> 00:02:12.340 When they just invent answers based on their 00:02:12.340 --> 00:02:14.780 massive training data instead of sticking to 00:02:14.780 --> 00:02:17.000 what's actually in front of them. Ah, the dreaded 00:02:17.000 --> 00:02:19.960 hallucinations. Okay, so the idea here, according 00:02:19.960 --> 00:02:22.599 to the source, is you use a tool like Perplexity 00:02:22.599 --> 00:02:25.650 for that initial broad sweep. Uh -huh. Finding 00:02:25.650 --> 00:02:27.990 reports, articles, maybe even discussions on 00:02:27.990 --> 00:02:31.169 forums, videos. It's like super fast at discovering 00:02:31.169 --> 00:02:34.030 potentially relevant sources online. Right. But 00:02:34.030 --> 00:02:36.629 then when you need to analyze your specific documents 00:02:36.629 --> 00:02:39.229 reliably. That's where the second tool comes 00:02:39.229 --> 00:02:42.550 in. Notebook LM. Yes. Notebook LM is designed, 00:02:42.789 --> 00:02:45.689 supposedly, to answer only from the sources you 00:02:45.689 --> 00:02:48.069 import. It builds a sort of knowledge base just 00:02:48.069 --> 00:02:50.189 from your files or links. Okay, so that's the 00:02:50.189 --> 00:02:52.590 key difference then. Perplexity helps you find 00:02:52.590 --> 00:02:54.449 stuff out there, maybe even organize potential 00:02:54.449 --> 00:02:57.250 sources. Yeah, the gathering part. But when you 00:02:57.250 --> 00:03:00.849 need to analyze your stack, the reports, the 00:03:00.849 --> 00:03:02.449 articles, the notes you've collected reliably, 00:03:02.870 --> 00:03:05.669 Notebook LM is meant to be the laser -focused 00:03:05.669 --> 00:03:09.639 one. Right. Supposedly minimizing those hallucinations 00:03:09.639 --> 00:03:12.000 because it's strictly grounded in your stuff. 00:03:12.219 --> 00:03:14.120 It doesn't look outside that. Okay, that makes 00:03:14.120 --> 00:03:17.219 sense. Exactly. So when you combine them, you 00:03:17.219 --> 00:03:20.219 get the best of both worlds potentially. Broad 00:03:20.219 --> 00:03:22.080 data collection from perplexity to help build 00:03:22.080 --> 00:03:25.699 your source list. Uh -huh. And then that focused... 00:03:26.379 --> 00:03:29.240 hopefully accurate analysis from Notebook LM 00:03:29.240 --> 00:03:32.379 grounded purely in those specific sources you 00:03:32.379 --> 00:03:34.939 selected. Right. And the source says this makes 00:03:34.939 --> 00:03:37.139 them suitable for all sorts of projects, academic, 00:03:37.379 --> 00:03:40.219 business, creative. Seems pretty versatile, you 00:03:40.219 --> 00:03:42.580 know? Yeah, I can see that. Use the wide net 00:03:42.580 --> 00:03:44.620 first. Find the good stuff that's out there. 00:03:44.719 --> 00:03:47.240 Then bring the important pieces into the, like, 00:03:47.259 --> 00:03:50.099 secure analysis zone. The walled garden of your 00:03:50.099 --> 00:03:51.819 own information. Yeah, that only talks about 00:03:51.819 --> 00:03:54.759 your material. Uh -huh. The source actually goes 00:03:54.759 --> 00:03:56.840 through some specific examples of how this combined 00:03:56.840 --> 00:03:59.219 approach can help you pull insights from different 00:03:59.219 --> 00:04:01.219 types of source stacks you might have. Okay, 00:04:01.280 --> 00:04:03.860 sounds good. Let's unpack a few. What's the first 00:04:03.860 --> 00:04:06.639 one? Market trend analysis. Let's say you've 00:04:06.639 --> 00:04:08.599 been tasked with understanding what's happening 00:04:08.599 --> 00:04:11.500 in a specific industry. Okay. Maybe you've collected 00:04:11.500 --> 00:04:13.819 a bunch of recent reports and articles using 00:04:13.819 --> 00:04:16.420 something like perplexity. Right, like researching. 00:04:17.620 --> 00:04:20.800 responsible AI trends. I've got these PDFs from 00:04:20.800 --> 00:04:22.939 consulting firms, maybe some research papers 00:04:22.939 --> 00:04:27.240 I found. Perfect. So you used a tool, maybe perplexity, 00:04:27.240 --> 00:04:29.939 to find those high quality sources searching 00:04:29.939 --> 00:04:33.000 for reports from Deloitte McKinsey published 00:04:33.000 --> 00:04:35.459 in the last year, let's say. Got it. Gathered 00:04:35.459 --> 00:04:38.579 those links, downloaded the PDFs, that initial 00:04:38.579 --> 00:04:42.060 search and collection phase. Exactly. Then you 00:04:42.060 --> 00:04:44.779 bring those specific reports and papers into 00:04:44.779 --> 00:04:47.699 Notebook LM. Okay. Upload them. Right. Now, instead 00:04:47.699 --> 00:04:50.100 of reading all 200 pages across three different 00:04:50.100 --> 00:04:52.540 reports, you can ask it questions based only 00:04:52.540 --> 00:04:54.259 on those documents. And it's supposed to tell 00:04:54.259 --> 00:04:55.980 you what your sources say and not just a general 00:04:55.980 --> 00:04:58.519 web answer. Precisely. Based on the stack of 00:04:58.519 --> 00:05:01.220 reports you provided on responsible AI, Notebook 00:05:01.220 --> 00:05:03.740 LM could tell you, for example, that your sources 00:05:03.740 --> 00:05:06.139 consistently highlight increased adoption and 00:05:06.139 --> 00:05:08.759 the growing importance of human in the loop oversight 00:05:08.759 --> 00:05:11.680 as major shifts. And crucially, it should point 00:05:11.680 --> 00:05:13.639 to the specific sections in your documents where 00:05:13.639 --> 00:05:16.389 it found that info. Oh, well, so I wouldn't just 00:05:16.389 --> 00:05:18.970 get a generic AI answer. I'd get an answer from 00:05:18.970 --> 00:05:23.029 my specific reports like Deloitte mentions increased 00:05:23.029 --> 00:05:27.230 adoption on page 15, driven by regulatory pressures, 00:05:27.310 --> 00:05:31.189 while McKinsey discusses human in the loop on 00:05:31.189 --> 00:05:34.449 page 32, emphasizing ethical considerations. 00:05:35.290 --> 00:05:37.269 Something like that. Yeah, exactly. It's grounding 00:05:37.269 --> 00:05:39.670 the insights strictly in what your documents 00:05:39.670 --> 00:05:42.470 actually say. OK. You can also ask it based on 00:05:42.470 --> 00:05:44.850 those reports. What's driving companies to invest 00:05:44.850 --> 00:05:48.110 in this topic? Right. And it would pull out motivations 00:05:48.110 --> 00:05:50.949 mentioned in your sources like regulatory pressures 00:05:50.949 --> 00:05:54.350 or maybe market demand. If those drivers are 00:05:54.350 --> 00:05:56.329 present in the reports, you fed it. And if my 00:05:56.329 --> 00:05:58.430 sources don't mention market demand? Notebook 00:05:58.430 --> 00:06:00.350 LM shouldn't invent it. It should basically say 00:06:00.350 --> 00:06:02.449 that's not mentioned in the provided documents. 00:06:02.649 --> 00:06:04.269 That's the key difference, right? It won't guess. 00:06:04.410 --> 00:06:07.810 It just says, nope, not in this stack. Okay. 00:06:08.029 --> 00:06:10.470 Market trend analysis using curated sources. 00:06:10.949 --> 00:06:13.430 That seems powerful for getting specific answers 00:06:13.430 --> 00:06:15.670 from your research. What about another example? 00:06:15.790 --> 00:06:18.189 How about product enhancement research? Imagine 00:06:18.189 --> 00:06:20.740 you're a product team. You've collected a bunch 00:06:20.740 --> 00:06:23.800 of user reviews for competing products, maybe 00:06:23.800 --> 00:06:26.459 related ones too. Yeah, scrape from Keptera or 00:06:26.459 --> 00:06:29.339 G2, read some threads from Reddit, watch some 00:06:29.339 --> 00:06:32.259 YouTube reviews, that kind of thing. Exactly. 00:06:32.699 --> 00:06:36.639 You've got this pile of user feedback and competitor 00:06:36.639 --> 00:06:40.420 reviews for, say, email marketing tools. You 00:06:40.420 --> 00:06:42.660 want to enhance yours. Okay, got it. I've got 00:06:42.660 --> 00:06:46.759 my pile. Great. So you've used tools. maybe Perplexity's 00:06:46.759 --> 00:06:49.439 focus search on review sites or Reddit. To gather 00:06:49.439 --> 00:06:51.259 those specific reviews and feedback, you got 00:06:51.259 --> 00:06:53.040 the text, the links, whatever you collected. 00:06:53.259 --> 00:06:55.379 Found the good stuff. Check. Then you bring all 00:06:55.379 --> 00:06:57.959 that user feedback data into Notebook LM. Right. 00:06:58.680 --> 00:07:01.360 Now you can start asking it to synthesize insights 00:07:01.360 --> 00:07:03.600 across all those individual reviews and pieces 00:07:03.600 --> 00:07:06.060 of feedback you imported. Like what are the common 00:07:06.060 --> 00:07:08.279 pain points customers mentioned across all these 00:07:08.279 --> 00:07:11.379 reviews? Exactly. Based only on the stack of 00:07:11.379 --> 00:07:13.579 reviews you imported, Notebook LM could tell 00:07:13.579 --> 00:07:16.199 you that across the board, users seem frustrated 00:07:16.199 --> 00:07:19.459 with technical glitches during setup. Or maybe 00:07:19.459 --> 00:07:22.740 limitations in the email design templates or 00:07:22.740 --> 00:07:25.980 slow customer support. It's summarizing themes 00:07:25.980 --> 00:07:28.420 present in your specific source material. Or 00:07:28.420 --> 00:07:31.939 I could ask, what specific features do users 00:07:31.939 --> 00:07:35.560 consistently mention as missing or desired to 00:07:35.560 --> 00:07:38.139 identify opportunities based on that feedback 00:07:38.139 --> 00:07:40.180 stack? Totally. And here's where it gets really 00:07:40.180 --> 00:07:42.180 interesting, according to the source. OK. You 00:07:42.180 --> 00:07:45.079 can upload your own product's website page, maybe 00:07:45.079 --> 00:07:48.579 the main features page, into Notebook LM alongside 00:07:48.579 --> 00:07:51.420 all that competitor review data you gathered. 00:07:51.620 --> 00:07:53.980 Oh, wow. So my stuff and their stuff. And then 00:07:53.980 --> 00:07:56.439 what? And then you can ask Notebook LM, grounded 00:07:56.439 --> 00:08:14.100 in your product page. Whoa. That's pretty cool. 00:08:14.220 --> 00:08:16.740 Getting AI to give you concrete suggestions based 00:08:16.740 --> 00:08:18.399 on your stuff and the research you collected. 00:08:18.500 --> 00:08:20.220 It's not just summarizing. It's helping with 00:08:20.220 --> 00:08:23.069 the analysis and... Like application. It really 00:08:23.069 --> 00:08:25.709 streamlines that phase, yeah. Another example 00:08:25.709 --> 00:08:28.610 they give is audience research. Ah, understanding 00:08:28.610 --> 00:08:30.709 your target audience. Okay, say I've got a stack 00:08:30.709 --> 00:08:34.809 of surveys, reports, articles about parents' 00:08:34.950 --> 00:08:37.090 needs for child care services. Perfect example. 00:08:37.629 --> 00:08:39.889 So you use something to find those recent surveys 00:08:39.889 --> 00:08:42.330 about working parents' needs, maybe research 00:08:42.330 --> 00:08:45.029 reports using specific site operators and perplexity 00:08:45.029 --> 00:08:48.090 to focus on academic or organizational sources. 00:08:48.409 --> 00:08:51.169 Right. Found the relevant PDFs and links. You've 00:08:51.169 --> 00:08:53.129 collected them. Built up that knowledge base 00:08:53.129 --> 00:08:55.389 about the audience. Got it. Right. Now you bring 00:08:55.389 --> 00:08:57.750 all those surveys, reports, and studies into 00:08:57.750 --> 00:09:00.769 Notebook LM. Okay. And you can ask it to synthesize 00:09:00.769 --> 00:09:02.710 what your specific sources say about this audience. 00:09:03.070 --> 00:09:05.070 Like, what are the biggest challenges parents 00:09:05.070 --> 00:09:07.289 mention about child care? According to these 00:09:07.289 --> 00:09:10.669 surveys I uploaded. Precisely. Based only on 00:09:10.669 --> 00:09:13.110 the documents you imported, Notebook LM could 00:09:13.110 --> 00:09:15.009 tell you that your sources highlight availability 00:09:15.009 --> 00:09:18.990 issues, maybe concerns about quality, or a perceived 00:09:18.990 --> 00:09:21.730 lack of learning programs as major challenges. 00:09:22.070 --> 00:09:25.250 It extracts these key points directly from your 00:09:25.250 --> 00:09:28.769 research. Or I could ask. What are the primary 00:09:28.769 --> 00:09:31.389 considerations when parents choose a child care 00:09:31.389 --> 00:09:33.970 service, according to these reports? To pull 00:09:33.970 --> 00:09:36.590 out factors like cost, location, convenience, 00:09:36.929 --> 00:09:40.769 maybe branding. Again, only if my sources actually 00:09:40.769 --> 00:09:43.309 mention them. Exactly. And just like the product 00:09:43.309 --> 00:09:45.649 research example, you can upload your own child 00:09:45.649 --> 00:09:48.450 care business website into Notebook LM alongside 00:09:48.450 --> 00:09:51.690 all that audience research. OK, so bring in my 00:09:51.690 --> 00:09:53.889 own site again. With the research this time. 00:09:54.049 --> 00:09:58.110 Yep. And then Ask Notebook LM, grounded in your 00:09:58.110 --> 00:10:00.549 site content and all that audience research you 00:10:00.549 --> 00:10:03.230 gathered, suggests the top three changes to our 00:10:03.230 --> 00:10:05.350 homepage messaging based on this research. Okay. 00:10:05.429 --> 00:10:07.649 It might suggest things like prominently featuring 00:10:07.649 --> 00:10:10.450 information about flexible hours. If the research 00:10:10.450 --> 00:10:12.669 you provided indicates that's a significant need 00:10:12.669 --> 00:10:14.850 parents mention, but your current site doesn't 00:10:14.850 --> 00:10:16.929 really highlight it. Wow. And I could even ask 00:10:16.929 --> 00:10:19.789 maybe what needs highlighted in their research 00:10:19.789 --> 00:10:22.220 are not currently mentioned on our website. could 00:10:22.220 --> 00:10:25.299 emphasize. To find things like, I don't know, 00:10:25.360 --> 00:10:27.700 learning philosophy or staff qualifications. 00:10:27.919 --> 00:10:30.600 If my research says parents care about that, 00:10:30.679 --> 00:10:33.220 but my site isn't featuring it prominently, that's 00:10:33.220 --> 00:10:35.879 wild. It is. It's like having an AI consultant 00:10:35.879 --> 00:10:38.919 analyze your website against market research 00:10:38.919 --> 00:10:41.179 you provided. It really is. Audience research 00:10:41.179 --> 00:10:43.580 seems like a really strong use case for this 00:10:43.580 --> 00:10:45.679 kind of focused analysis on your own material. 00:10:45.960 --> 00:10:48.059 Yeah, definitely. What about content creation? 00:10:48.320 --> 00:10:51.480 Like planning a podcast, maybe? Yeah. OK, say 00:10:51.480 --> 00:10:54.820 I've got. Reviews of successful leadership podcasts, 00:10:55.159 --> 00:10:58.100 maybe some transcripts of popular episodes. Perfect 00:10:58.100 --> 00:11:00.960 example. So you've used tools to find those top 00:11:00.960 --> 00:11:03.779 leadership podcasts, maybe gathered reviews from 00:11:03.779 --> 00:11:06.159 sites like Listen Notes, perhaps even downloaded 00:11:06.159 --> 00:11:08.320 some episode transcripts or show descriptions. 00:11:08.419 --> 00:11:10.649 Right. Got my raw material on the competitive 00:11:10.649 --> 00:11:12.610 landscape and what's already out there. Exactly. 00:11:12.629 --> 00:11:15.110 Now, you bring all those reviews, show descriptions, 00:11:15.269 --> 00:11:17.009 transcripts, whatever source material you've 00:11:17.009 --> 00:11:20.149 gathered into Notebook LM. Okay. And you ask 00:11:20.149 --> 00:11:23.230 it to analyze that specific content. So I could 00:11:23.230 --> 00:11:26.230 ask it, how do these specific podcast shows describe 00:11:26.230 --> 00:11:28.809 their ideal listeners? Based on their show descriptions 00:11:28.809 --> 00:11:31.850 and reviews that I put in. Precisely. Grounded 00:11:31.850 --> 00:11:35.009 only in the material you imported, it could synthesize 00:11:35.009 --> 00:11:37.470 descriptions of target audiences, specific job 00:11:37.470 --> 00:11:40.129 roles, experience levels, company types mentioned 00:11:40.129 --> 00:11:42.809 across the shows you provided. It's analyzing 00:11:42.809 --> 00:11:45.090 my competitive intelligence file, essentially. 00:11:45.429 --> 00:11:50.629 And I can ask it to compare the positioning of 00:11:50.629 --> 00:11:53.169 these shows. What unique angles do they emphasize? 00:11:53.870 --> 00:11:56.350 Based on my input. Totally. Based on your stack 00:11:56.350 --> 00:11:58.789 of show descriptions and reviews, Notebook LM 00:11:58.789 --> 00:12:01.090 could identify how different shows are carving 00:12:01.090 --> 00:12:03.389 out their niche. Which helps me figure out how 00:12:03.389 --> 00:12:05.850 to position my leadership podcast to stand out. 00:12:06.049 --> 00:12:09.029 Exactly. To stand out in that specific competitive 00:12:09.029 --> 00:12:11.990 set you researched, it seems super practical 00:12:11.990 --> 00:12:14.889 for content creators analyzing their space. That 00:12:14.889 --> 00:12:17.850 is super practical. Okay, makes sense. Finally, 00:12:17.870 --> 00:12:20.149 there was a fifth use case mentioned. Yeah, learning 00:12:20.149 --> 00:12:21.909 new subjects. Let's say you're diving into something 00:12:21.909 --> 00:12:24.590 totally new, like... Mm -hmm. Marketing psychology. 00:12:25.149 --> 00:12:27.809 Oh, this is big for a lot of us. Just learning 00:12:27.809 --> 00:12:31.470 something complex from scratch. Okay, I've collected 00:12:31.470 --> 00:12:33.769 a bunch of articles, maybe some lecture transcripts, 00:12:33.769 --> 00:12:36.129 a few relevant podcast episodes on marketing 00:12:36.129 --> 00:12:39.850 psychology. Great. So you've used tools, maybe 00:12:39.850 --> 00:12:42.730 searching perplexity for a title, marketing psychology, 00:12:42.970 --> 00:12:45.570 articles from trustworthy sites, finding relevant 00:12:45.570 --> 00:12:48.070 videos or podcast episodes, and collecting that 00:12:48.070 --> 00:12:50.690 material. You've got your learning stack. Got 00:12:50.690 --> 00:12:53.519 my library built. ready to learn right now you 00:12:53.519 --> 00:12:55.759 bring all those articles transcripts maybe links 00:12:55.759 --> 00:13:00.139 to videos notebook lm can process video transcripts 00:13:00.139 --> 00:13:03.059 too apparently into notebook lm okay this is 00:13:03.059 --> 00:13:05.460 where it can help you learn the material you 00:13:05.460 --> 00:13:08.440 gathered now does it just summarize it all It 00:13:08.440 --> 00:13:10.419 seems it can do more than just a simple summary. 00:13:10.600 --> 00:13:12.919 The source says Notebook LM has built -in features 00:13:12.919 --> 00:13:15.179 to generate things like a study guide and an 00:13:15.179 --> 00:13:17.580 FAQ based only on your imported sources. Oh, 00:13:17.600 --> 00:13:19.879 wow. You just click generate, apparently. So 00:13:19.879 --> 00:13:21.820 it can actually structure the material I found, 00:13:21.899 --> 00:13:24.559 not just spit back text? Yes. It's organizing 00:13:24.559 --> 00:13:27.340 and summarizing the key concepts within your 00:13:27.340 --> 00:13:30.519 specific learning materials. You could also ask 00:13:30.519 --> 00:13:33.559 it for, say, a beginner -friendly overview using 00:13:33.559 --> 00:13:35.980 simple language and real examples from the sources 00:13:35.980 --> 00:13:38.580 you provided. That sounds way better than just 00:13:38.580 --> 00:13:40.899 reading a dense textbook that might cover stuff 00:13:40.899 --> 00:13:43.100 I don't need right now. I'm getting a personalized 00:13:43.100 --> 00:13:45.720 summary based on the specific materials I curated. 00:13:45.899 --> 00:13:48.899 Totally. You could ask it to summarize the top 00:13:48.899 --> 00:13:51.340 marketing psychology principles and how to apply 00:13:51.340 --> 00:13:53.879 them in marketing based on these sources. It 00:13:53.879 --> 00:13:56.740 might explain principles like scarcity or anchoring 00:13:56.740 --> 00:13:59.919 bias with examples pulled directly from the articles 00:13:59.919 --> 00:14:02.340 or transcripts you imported. And I could ask. 00:14:02.799 --> 00:14:05.179 Show me examples from these documents of how 00:14:05.179 --> 00:14:07.179 companies use these psychological principles 00:14:07.179 --> 00:14:10.720 in business, like using scarcity to boost sales, 00:14:10.840 --> 00:14:12.620 if one of my articles discussed that specific 00:14:12.620 --> 00:14:14.960 tactic. Uh -huh. And again, you could upload 00:14:14.960 --> 00:14:16.820 your own website, let's say that child care business 00:14:16.820 --> 00:14:19.519 site again, just for fun. Okay, bring the child 00:14:19.519 --> 00:14:22.720 care site back in. And ask, how can these psychology 00:14:22.720 --> 00:14:25.340 principles, as explained in these sources I'm 00:14:25.340 --> 00:14:28.070 learning from, be applied to this website? It 00:14:28.070 --> 00:14:30.029 might suggest highlighting limited enrollment 00:14:30.029 --> 00:14:33.029 based on a scarcity principle explained in your 00:14:33.029 --> 00:14:35.190 learning material and showing how that applies 00:14:35.190 --> 00:14:37.129 specifically to a business like yours described 00:14:37.129 --> 00:14:39.429 on your site. That's amazing. So it's not just 00:14:39.429 --> 00:14:41.950 explaining concepts from my sources, but helping 00:14:41.950 --> 00:14:43.789 me apply them immediately to my own context. 00:14:44.009 --> 00:14:47.769 Yeah. That connects the learning to action directly 00:14:47.769 --> 00:14:51.059 from the material I chose. That's powerful. It 00:14:51.059 --> 00:14:53.440 really seems to be. It's leveraging AI to go 00:14:53.440 --> 00:14:55.539 beyond just finding information, but helping 00:14:55.539 --> 00:14:58.820 you analyze and synthesize insights from your 00:14:58.820 --> 00:15:01.580 specific stack and maybe even apply them. Absolutely. 00:15:01.600 --> 00:15:05.320 So wrapping this up then, the core idea is that 00:15:05.320 --> 00:15:08.879 tools like Perplexity and Notebook LM, when used 00:15:08.879 --> 00:15:12.059 together in this kind of workflow, they offer 00:15:12.059 --> 00:15:14.259 a potentially powerful way to cut through that 00:15:14.259 --> 00:15:17.039 information overload by focusing on your material. 00:15:17.360 --> 00:15:19.860 Yes. Perplexity helps you broaden your search 00:15:19.860 --> 00:15:22.019 and discover potential sources, cast that wide 00:15:22.019 --> 00:15:25.000 net. But Notebook LM seems key for that focused, 00:15:25.220 --> 00:15:27.759 hopefully hallucination -resistant analysis grounded 00:15:27.759 --> 00:15:30.159 purely in the reports, articles, reviews, or 00:15:30.159 --> 00:15:32.159 whatever documents you select and import. It 00:15:32.159 --> 00:15:34.240 really highlights the synergy, right? Use a wide 00:15:34.240 --> 00:15:36.519 net to gather, then take your specific catch 00:15:36.519 --> 00:15:39.200 and do a deep dive into that. relying only on 00:15:39.200 --> 00:15:40.980 what's in front of you in that tool. Exactly. 00:15:41.100 --> 00:15:43.820 It's about cutting down the noise and getting 00:15:43.820 --> 00:15:46.200 straight to the important stuff, grounded in 00:15:46.200 --> 00:15:48.899 the materials you curated for your project or 00:15:48.899 --> 00:15:52.639 your learning goal. So for you listening, think 00:15:52.639 --> 00:15:54.080 about your own projects, maybe your learning 00:15:54.080 --> 00:15:56.379 goals, maybe that stack of articles or reports 00:15:56.379 --> 00:15:59.559 sitting on your desk right now. Where could this 00:15:59.559 --> 00:16:03.019 kind of structured AI -assisted workflow, finding 00:16:03.019 --> 00:16:06.440 broader context than analyzing your specific 00:16:06.440 --> 00:16:09.639 documents for insights, help you cut down the 00:16:09.639 --> 00:16:12.259 noise? get you straight to those key takeaways 00:16:12.259 --> 00:16:14.600 or action items from the material you actually 00:16:14.600 --> 00:16:16.740 care about. It feels like a potentially significant 00:16:16.740 --> 00:16:19.659 shift in how we can approach information overload, 00:16:20.100 --> 00:16:22.700 you know, allowing us to really leverage the 00:16:22.700 --> 00:16:25.379 specific research we've already gathered or curated. 00:16:25.759 --> 00:16:28.259 Totally. And maybe here's something to ponder 00:16:28.259 --> 00:16:30.679 as these AI tools get smarter and more integrated 00:16:30.679 --> 00:16:33.179 into our workflows. Okay. How does our own role 00:16:33.179 --> 00:16:35.779 as the researcher or the learner really evolve? 00:16:36.480 --> 00:16:39.200 Are we just becoming like users of these tools, 00:16:39.419 --> 00:16:42.039 button pushers? Interesting question. Or does 00:16:42.039 --> 00:16:44.879 leveraging them effectively, especially this 00:16:44.879 --> 00:16:47.419 ability that the source highlights to ground 00:16:47.419 --> 00:16:50.019 the analysis in our own chosen sources actually 00:16:50.019 --> 00:16:52.929 make us. Better critical thinkers. You mean better 00:16:52.929 --> 00:16:55.009 at curating quality sources in the first place. 00:16:55.149 --> 00:16:57.409 Yeah, and better at synthesizing complex information, 00:16:57.750 --> 00:17:00.370 asking the right questions to pull out the truly 00:17:00.370 --> 00:17:02.690 valuable insights from our material. It forces 00:17:02.690 --> 00:17:04.890 you to be more deliberate, perhaps, about what 00:17:04.890 --> 00:17:08.109 you feed it and what you ask it. Maybe. Something 00:17:08.109 --> 00:17:09.950 to think about as you explore this stuff yourself.