Two Jones Explain How Google Leaves Tool Makers Jonesing for Data

WEBVTT
00:10.172 --> 00:16.719
[SPEAKER_01]: Hey everyone, welcome to webcology, here on www.fm, it's the 15th of October, 2026.
00:16.759 --> 00:21.745
[SPEAKER_01]: This Jim Hedger from DigitalOeys Media and Chrissy and Shackinger from Search for the Walls.
00:22.266 --> 00:22.846
[SPEAKER_03]: Hi everybody!
00:23.816 --> 00:39.906
[SPEAKER_01]: uh... we have in the studio uh... dicks and jones from inlinks and and wakay and uh... rion jones from uh... from razor fish and the developer of supercon uh... just the new guys know they're in the uh... the jones are sitting in the
00:41.267 --> 00:44.549
[SPEAKER_01]: in the green room, no relation to each other, but they're here.
00:44.929 --> 00:51.093
[SPEAKER_01]: For the jump further, Spank special thanks for our sponsors, audience key and a quick studio.
00:52.414 --> 00:58.258
[SPEAKER_01]: Again, big love, thank you for keeping the lights on and the microphones that tapes rolling and the microphone's going.
00:59.659 --> 01:03.721
[SPEAKER_01]: And before we go any further, we have sadness to report.
01:05.972 --> 01:24.562
[SPEAKER_01]: One of the industry OGs, one of the original SEOs, Nick Wilson, based in the UK, one of the first to introduce a news aggregator trying to separate all the noise out of the actual signal that SEOs ought to be listening to.
01:26.863 --> 01:30.146
[SPEAKER_01]: Nick died last weekend at home, he had a, after a long, bad, a cancer.
01:30.927 --> 01:33.930
[SPEAKER_01]: He was a beloved figure in the industry, as long as I can remember.
01:34.610 --> 01:37.773
[SPEAKER_01]: And again, it was his news aggregator that started.
01:38.834 --> 01:41.556
[SPEAKER_01]: So many people's mornings, he took a lot of the,
01:43.618 --> 01:51.401
[SPEAKER_01]: hype and the marketing bullshit and separated it out from real news that people reading news to read.
01:52.281 --> 02:07.328
[SPEAKER_01]: He continued to help people through the rest of his like two-and-a-half decade career and so, you know, continue to appear at conferences and speak, collaborate with others and the industry was, you know,
02:07.708 --> 02:11.471
[SPEAKER_01]: stunned and really sat in too much and learned that he he died over the weekend.
02:11.831 --> 02:15.113
[SPEAKER_01]: So Nick Wilson, you changed our world man.
02:15.133 --> 02:16.554
[SPEAKER_01]: Thank you so much and you'll be missed.
02:18.095 --> 02:29.062
[SPEAKER_03]: Yeah, so I said I didn't know Nick while I knew him through social, but I know a lot of friends in the UK especially overseas are just absolutely heartbreaking and we're really good friends with himself.
02:29.983 --> 02:33.645
[SPEAKER_03]: Our hearts got with everybody in industry who are missing Nick today.
02:35.985 --> 02:43.668
[SPEAKER_00]: I just like to echo, I mean, I knew I've known Nick for such a long time when I first knew me was still living in Russia.
02:44.688 --> 02:53.972
[SPEAKER_00]: He's not Russian at all, but anyway his wife is and he was always lovely and he, of course, had Yandex and his mind back at the days.
02:54.452 --> 03:00.735
[SPEAKER_00]: He actually ended up in Ireland, so he actually died in Ireland, so in Ireland they bury people very, very quickly
03:02.015 --> 03:17.719
[SPEAKER_00]: So I didn't get a chance to get over to the funeral, but he put himself pretty much a castle, you know, a mansion, and it's massive, and I never got to go there, which was a real pity.
03:17.819 --> 03:30.962
[SPEAKER_00]: It's going to be terrible with not having me in his, all the guys that talk partnership, where he was at the end, yeah, hard goes out to there and also to his wife and family as well.
03:31.962 --> 03:38.986
[SPEAKER_01]: Yeah, well Nick Wilson again, you'll you'll be missed.
03:40.066 --> 03:44.549
[SPEAKER_01]: Okay, so again, we have we have Dickson and Ryan waiting in the green room.
03:44.569 --> 03:49.971
[SPEAKER_01]: We've got a couple quicker industry announcements to go through and then we'll lead in to what we're going to be talking about today.
03:50.031 --> 03:54.514
[SPEAKER_01]: The great data poisoning from Google, but before we get there
03:55.629 --> 03:58.271
[SPEAKER_01]: Google had safety of every person 26 is now out.
03:58.291 --> 04:02.174
[SPEAKER_01]: It's available at a Google universe near you.
04:02.194 --> 04:06.297
[SPEAKER_03]: Yeah, there's a bunch of new features so people should get checked into it.
04:06.318 --> 04:09.360
[SPEAKER_03]: Jim and I don't do ads, so we don't really get in depth at all, so it's.
04:10.332 --> 04:12.073
[SPEAKER_01]: But you can't check it into it.
04:12.293 --> 04:12.493
[SPEAKER_01]: S.E.
04:12.533 --> 04:17.954
[SPEAKER_01]: Roundtable put a put an article up about it this morning, so check it out.
04:18.734 --> 04:19.575
[SPEAKER_01]: This is kind of cool.
04:19.955 --> 04:24.436
[SPEAKER_01]: Google's launched a video series for weird technical SEO questions.
04:24.536 --> 04:26.256
[SPEAKER_01]: The non-cononical cabinet.
04:27.837 --> 04:33.918
[SPEAKER_01]: This time, sir, starring Gary E's, and I'm not sure the other person was involved.
04:41.113 --> 05:05.630
[SPEAKER_03]: sort of like, we are to questions the effective, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh, uh,
05:10.440 --> 05:12.181
[SPEAKER_02]: awkward naming is what Gary does best.
05:12.281 --> 05:13.321
[SPEAKER_02]: Like that's his secret.
05:13.381 --> 05:14.742
[SPEAKER_02]: He's giving something awkward name.
05:14.782 --> 05:16.262
[SPEAKER_02]: In case you listen to it.
05:17.003 --> 05:17.623
[SPEAKER_03]: Yeah.
05:17.643 --> 05:18.903
[SPEAKER_01]: It's an inch near at heart, right?
05:19.764 --> 05:20.144
[SPEAKER_03]: Yeah.
05:20.244 --> 05:20.464
[SPEAKER_03]: Yeah.
05:20.484 --> 05:25.366
[SPEAKER_03]: Jim and I have a show on the show to let engineers name things.
05:25.426 --> 05:30.088
[SPEAKER_01]: Google Merchant Center is in juice, new store dress attribute for local inventories.
05:31.408 --> 05:38.251
[SPEAKER_01]: You'll be able to know what local store, if you have a chain, that stuff has been moved from or looked at in.
05:40.627 --> 05:42.292
[SPEAKER_01]: Oh.
05:42.392 --> 05:42.793
[SPEAKER_01]: Okay.
05:43.435 --> 05:43.676
[SPEAKER_01]: Um.
05:44.939 --> 05:47.345
[SPEAKER_01]: Try with Pack M-SEO plugin.
05:47.385 --> 05:48.027
[SPEAKER_01]: Make say to us.
05:49.275 --> 05:50.696
[SPEAKER_01]: Is this an advertisement?
05:50.816 --> 05:51.577
[SPEAKER_01]: It's absurdly.
05:51.617 --> 05:55.719
[SPEAKER_03]: No, no, no, no, it's somebody an industry who developed a tool to make things faster.
05:55.739 --> 05:57.701
[SPEAKER_03]: Roger is teaching you today.
05:57.881 --> 05:58.461
[SPEAKER_03]: John Henshah.
05:58.481 --> 05:59.602
[SPEAKER_03]: Do you remember John?
05:59.962 --> 06:00.362
[SPEAKER_01]: Oh, okay.
06:00.382 --> 06:01.543
[SPEAKER_01]: John, I know John Henshah.
06:01.823 --> 06:06.746
[SPEAKER_03]: Yeah, so it's a free plugin for people to try to make their site faster and of course we're Chris.
06:07.567 --> 06:10.169
[SPEAKER_03]: And he just released a new version with new features.
06:10.469 --> 06:14.872
[SPEAKER_03]: So I give a little shout out to because it helps me to say really fast.
06:15.492 --> 06:16.192
[SPEAKER_03]: I'm not recovered.
06:16.232 --> 06:16.833
[SPEAKER_03]: It has a story.
06:17.490 --> 06:22.314
[SPEAKER_01]: I'd like to talk to John simply about developing in the WordPress environment and what that feels like these days.
06:23.194 --> 06:23.975
[SPEAKER_03]: Yeah, that would be great.
06:23.995 --> 06:24.955
[SPEAKER_03]: We can talk to like that.
06:26.096 --> 06:26.276
[SPEAKER_01]: No.
06:28.078 --> 06:37.845
[SPEAKER_01]: But you know what we have a couple of experts who develop in a different environment, who develop SEO tools, site measurement tools,
06:40.214 --> 06:46.121
[SPEAKER_01]: generative AI awareness tools, et cetera, in the Googley environment.
06:47.242 --> 06:47.403
[SPEAKER_01]: And
06:50.375 --> 07:00.260
[SPEAKER_01]: Recently, this movement has been, Google's had issues with people taking their information and trying to provide analytics in the AI environment.
07:02.181 --> 07:11.185
[SPEAKER_01]: New analysis from Enterprise rank tracking firm nozzle claims Google's showing misleading search result pages to rank trackers in AI companies.
07:11.305 --> 07:16.128
[SPEAKER_01]: They're actually poisoning data that the AI companies are trying to try to collect.
07:17.356 --> 07:17.937
[SPEAKER_03]: Yeah.
07:17.957 --> 07:21.139
[SPEAKER_03]: So, people should read the nozzle article.
07:21.259 --> 07:23.701
[SPEAKER_03]: It's from, this is out.
07:23.721 --> 07:24.142
[SPEAKER_03]: Sorry, Jim.
07:24.182 --> 07:25.823
[SPEAKER_03]: Is it a search engine or a table of search engine?
07:25.843 --> 07:27.184
[SPEAKER_01]: Journal, this comes at a journal.
07:27.424 --> 07:36.272
[SPEAKER_03]: Yeah, search engine for the details, because nozzle, there's a link to the nozzle report and it's quite extensive with graphs and all sorts of information about how they did this study.
07:37.420 --> 07:43.225
[SPEAKER_03]: I look like about 88% of results are repeated across the board in regular search.
07:44.165 --> 07:51.511
[SPEAKER_03]: You know, the order of them, but in this case, I think it's 22% when they stapers are going over them.
07:51.911 --> 07:58.576
[SPEAKER_03]: I just think it's quite ironic that somebody is scraping everybody's sights and packaging them as an ad overview without a click.
07:59.317 --> 08:05.381
[SPEAKER_03]: It has now decided that people in the SEO industry can't have the information to say where their sights might be ranking.
08:06.574 --> 08:11.068
[SPEAKER_03]: You know, they're trying to support the AI people, but Google has, like,
08:12.399 --> 08:29.605
[SPEAKER_03]: Google has MIT engineers, Harvard engineers, Stanford engineers, like, couldn't Google find a better way to do this, where it's just not like basic, like it's so basic, it's so sparring, it's not even interesting way to do things, it's like, so basic, it's a little bit of bullying.
08:29.645 --> 08:34.706
[SPEAKER_01]: Well, it's a little bit of bullying, bullying tactics, like, oh my god, like, really?
08:35.686 --> 08:37.147
[SPEAKER_03]: Oh, it shocks the price.
08:37.447 --> 08:38.787
[SPEAKER_03]: I feel free to jump in, you guys.
08:38.807 --> 08:39.828
[SPEAKER_03]: There isn't better way.
08:40.508 --> 08:41.328
[SPEAKER_02]: There is a better way.
08:41.348 --> 08:42.048
[SPEAKER_02]: I'm just going to say it.
08:42.208 --> 08:43.789
[SPEAKER_02]: Give us a freaking API.
08:43.989 --> 08:45.069
[SPEAKER_02]: We know it exists, right?
08:45.109 --> 08:46.629
[SPEAKER_02]: It has access to it, right?
08:46.729 --> 08:51.570
[SPEAKER_02]: Everybody, you can, if you're a huge company with a deal with Google, you can get a search results API.
08:52.271 --> 08:52.771
[SPEAKER_02]: Guess what?
08:53.151 --> 08:58.592
[SPEAKER_02]: Everyone of these huge companies, some rush, bright edge, HRFs, open AI, perplexity.
08:59.012 --> 09:02.753
[SPEAKER_02]: They have the money to pay you for searched data and they probably will.
09:03.353 --> 09:04.413
[SPEAKER_02]: Give us any why.
09:04.933 --> 09:05.233
[SPEAKER_00]: Okay.
09:05.253 --> 09:07.494
[SPEAKER_03]: So I'm like, boy, go ahead.
09:09.148 --> 09:15.252
[SPEAKER_00]: Well, okay, so I've got this search in the Google search API, but it doesn't have AI overviews in it.
09:16.352 --> 09:21.515
[SPEAKER_00]: But there is one, and if you dig hard enough, you can find it.
09:22.256 --> 09:23.957
[SPEAKER_00]: But it doesn't cover AI overviews.
09:24.877 --> 09:26.799
[SPEAKER_00]: But I'm absolutely astounded.
09:26.819 --> 09:27.579
[SPEAKER_00]: They've taken this long.
09:27.599 --> 09:28.319
[SPEAKER_00]: I'll be honest with you.
09:33.146 --> 09:41.673
[SPEAKER_00]: The one, the one reason that my previous business, well, well, that, that, that, that, um, majestic didn't start going down the, uh, the, the rank check in route.
09:42.253 --> 09:48.478
[SPEAKER_00]: A big reason for that was eventually, we thought that the search engines would start blocking those results.
09:48.718 --> 09:50.800
[SPEAKER_00]: And I can't believe it's taken this long.
09:51.160 --> 09:52.581
[SPEAKER_00]: I mean, you know, I don't work.
09:52.681 --> 09:54.803
[SPEAKER_00]: It's still on a day to day basis anymore.
09:54.823 --> 09:58.205
[SPEAKER_00]: So I still have to do some stuff with the, the, the, the guys.
09:58.225 --> 10:01.648
[SPEAKER_00]: But, you know, uh, if I'd then we're going to take this long, I wouldn't have started.
10:02.068 --> 10:14.775
[SPEAKER_00]: I might have made some different decisions 20 years ago, but I think they're doing it in a really interesting way, actually, because they can, yeah, because they've got a dial.
10:15.275 --> 10:20.198
[SPEAKER_00]: Once they've said right, I'm going to do this percentage of results wrong, then they can do a dial.
10:21.499 --> 10:28.943
[SPEAKER_00]: But I think they can target from, from trackers pretty effectively, because most prompts, frankly,
10:29.683 --> 10:33.285
[SPEAKER_00]: a unique or more or less unique, very rarely were typed in.
10:33.325 --> 10:37.787
[SPEAKER_00]: So if they see the same prompt three times and they didn't think it's likely to be good.
10:41.369 --> 10:43.970
[SPEAKER_00]: They've identified some where they start polluting the results.
10:45.051 --> 10:50.274
[SPEAKER_00]: So I'm absolutely not surprised that that's what they're doing.
10:55.056 --> 11:03.719
[SPEAKER_00]: I think it will affect my industry for sure the prompt tracking industry they're going to be an interesting place and the ranked checkers are going to be in a very interesting place.
11:05.039 --> 11:17.984
[SPEAKER_00]: Because especially the ones that do volume the ones where where they kind of say right here's a search volume of this term you know that means that doing large amounts of searches, huge amounts of searches.
11:18.584 --> 11:20.686
[SPEAKER_00]: which to do any other way is expensive.
11:21.527 --> 11:24.850
[SPEAKER_00]: And I think it's going to damage them a lot.
11:25.891 --> 11:35.019
[SPEAKER_00]: I think I'm quite pleased that wayk started out by using APIs for all of this stuff, instead of for instead of scraping.
11:35.039 --> 11:44.667
[SPEAKER_00]: We do, we did bow to customer pressure with our prompt tracking, because while prompt tracking, we started by tracking via API.
11:45.668 --> 11:56.991
[SPEAKER_00]: The only person that also stuck with that on that theory was Dan Petrovick from Dasion SEO, and based on the two of us have done a lot of things that know how to make me in different directions or every other company.
11:57.111 --> 12:00.032
[SPEAKER_00]: Oh no, Ryan, Ryan's on the case as well.
12:00.532 --> 12:03.373
[SPEAKER_00]: It'd be fair, you know.
12:03.413 --> 12:09.695
[SPEAKER_00]: But we bow to the tracking, but we never gave up the API.
12:09.735 --> 12:13.236
[SPEAKER_00]: So I think we're in a better place than most, but I'm surprised it's taken in this long.
12:14.936 --> 12:15.716
[SPEAKER_00]: No, no, no, you're good.
12:15.896 --> 12:19.578
[SPEAKER_03]: What do you think about Ryan, yeah, go ahead, Ryan, let's say, what do you think about Ryan's solution other?
12:19.818 --> 12:24.380
[SPEAKER_02]: Yeah, I'm amazed at how AI and prompts exacerbated the issue, right?
12:24.421 --> 12:30.523
[SPEAKER_02]: Because we had all these SEO tools for decades and that's how we'll be around this call where all elders.
12:30.904 --> 12:34.565
[SPEAKER_02]: But we've had SEO tools for decades and Google just didn't care.
12:34.625 --> 12:37.006
[SPEAKER_02]: And then AI came around and all the AI tools.
12:37.026 --> 12:40.108
[SPEAKER_02]: I'm not going to name any of them because I use most of them, and partner with most of them.
12:40.608 --> 13:07.224
[SPEAKER_02]: But all the AI tools just crawled at this massive scale to where Google was like whoa whoa whoa whoa we have a problem and I hired people I actually found the engineer they hired on LinkedIn I won't say his name on the call but he gave it them from that up uh the person running it uh I did a little stalking and found him um he's not too hard to find if you know what to look for but it's getting bad I don't know if you guys saw the thing that Reddit did a while ago but Reddit ran an experiment where they put some content out there that was only available
13:09.625 --> 13:10.665
[SPEAKER_02]: And right Google has an API.
13:10.685 --> 13:13.006
[SPEAKER_02]: You can verify Google Bot's IP address as to know it's Google Bot.
13:13.826 --> 13:16.687
[SPEAKER_02]: If you came at this content, anywhere else you got a 404, there was nothing there.
13:16.727 --> 13:18.667
[SPEAKER_02]: If you were a Google Bot IP, you got this content.
13:19.567 --> 13:29.550
[SPEAKER_02]: Within hours of releasing it, every major AI provider out there was citing and quoting that content, meaning they all scraped Google.
13:30.617 --> 13:34.021
[SPEAKER_02]: or use Google Index to find it.
13:34.121 --> 13:35.082
[SPEAKER_00]: I wouldn't be surprised.
13:35.382 --> 13:41.969
[SPEAKER_00]: I reckon that that was a honeypot trap for the core case with Google and Serp API.
13:42.029 --> 13:44.452
[SPEAKER_00]: So for those who don't know the core case,
13:45.713 --> 13:49.616
[SPEAKER_00]: Google, Google sued SirPAPI for scraping.
13:50.317 --> 14:01.646
[SPEAKER_00]: It was thrown out of court by the US courts saying, saying it's not, we have noticed that Google scraped the internet to get this information in the first place, so it's thrown out.
14:01.686 --> 14:12.895
[SPEAKER_00]: But Google have twisted it and changed it round to say, well, the reason that we are having to do this is because we have made deals with people like Reddit,
14:14.316 --> 14:19.339
[SPEAKER_00]: And the clause in their contracts is that we cannot just freely give that information to other people.
14:19.879 --> 14:28.604
[SPEAKER_00]: So, to Ryan's solution, an API does the same thing, but giving it to other people, although, you know, maybe not for free, so they could share it with Reddit, I suppose.
14:29.564 --> 14:37.048
[SPEAKER_00]: But, but I would say that's one of the reasons they're not making it easy for this, this to happen in AI mode.
14:40.070 --> 14:46.915
[SPEAKER_01]: Yeah, it was blown through its own, one of its own core principles to make the world's information freely available to it.
14:48.396 --> 14:50.578
[SPEAKER_02]: Oh, yeah, that's a good point.
14:50.618 --> 14:54.180
[SPEAKER_01]: They did sort of say clearly, this is what they were going to do.
14:54.220 --> 15:04.147
[SPEAKER_01]: Now, in Google's defense, they scraped everything that was ever put on the web, and they have a copy of it, and the status centers around around the world.
15:04.728 --> 15:07.270
[SPEAKER_01]: But when they hit your website, they hit...
15:07.850 --> 15:11.872
[SPEAKER_01]: your website, then you're in the next one, then the next one, the next year we're on, and so on down the chain.
15:12.452 --> 15:18.394
[SPEAKER_01]: When everybody hits Google, they're all hitting Google's data centers, and scraping as much of that information as they possibly can get.
15:18.414 --> 15:25.997
[SPEAKER_01]: And they have to hit repeatedly over and over and over and over again at lightning speed, and it's, they do have a legitimate resource, it should happen.
15:27.028 --> 15:28.530
[SPEAKER_01]: Yeah, I'm not disagreeing with you.
15:28.970 --> 15:29.951
[SPEAKER_00]: No, actually.
15:30.071 --> 15:43.627
[SPEAKER_03]: Yeah, we do know that Conday Nass had to put a season to sister to, I believe it was her flexibility, but or anthropic, I can't remember which, is there getting hit so much every day that they almost deduct their sights?
15:44.612 --> 15:46.894
[SPEAKER_01]: Okay, before we go in, before we go any further.
15:48.416 --> 15:52.460
[SPEAKER_01]: The English voice here, the British accent here is Dickson Jones.
15:52.580 --> 15:59.446
[SPEAKER_01]: Dickson Jones has been involved in creating tools for as long as most people have been practicing SEO.
15:59.466 --> 15:59.847
[SPEAKER_01]: He was,
16:02.008 --> 16:18.777
[SPEAKER_01]: One of the original people at Majestic, the company that basically mapped the internet and created the deepest link pool that I probably did because the deepest link pool exists in today outside of the major search engines.
16:21.078 --> 16:36.573
[SPEAKER_01]: He's the founder of Inlinks, which is one search technology of the year, at least once, probably probably more, was one of the leading entity SEO technologies, and in March of year before last.
16:36.634 --> 16:38.936
[SPEAKER_01]: So, but a year and a half ago released a WAKA.
16:39.516 --> 16:41.258
[SPEAKER_00]: Well, you pronounce your origins.
16:41.278 --> 16:43.681
[SPEAKER_00]: So, thank you for pronouncing it right.
16:44.621 --> 17:03.117
[SPEAKER_01]: What I know is the book you, which is a powerful and extremely cool tool that I've used extensively tracking several different AI's and your performance against your competitors and against yourself and over time in the AI's.
17:15.390 --> 17:31.281
[SPEAKER_01]: Countless other other conferences around the world and known to get known to people who been in the industry for a long time, Ryan Jones, also known to have been around forever, the creator of a billion tools that people use.
17:33.978 --> 17:37.286
[SPEAKER_01]: on without even knowing that there were Ryan's creations.
17:37.326 --> 17:44.742
[SPEAKER_01]: He's a senior vice president of S.U.R.A.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R.R
17:45.939 --> 18:01.504
[SPEAKER_01]: one of the first AI entity matching tools, tracking here, tracking how closely you're matching entities with other entities out on the web and how you're performing an AI around that.
18:02.024 --> 18:07.606
[SPEAKER_01]: So, a long time contributor to pretty much every industry journal and speaker at many industry conferences, and
18:08.866 --> 18:10.787
[SPEAKER_01]: both Joneses.
18:11.207 --> 18:16.049
[SPEAKER_01]: How has Google's reluctance to share data affected the tools that you're making?
18:16.509 --> 18:18.790
[SPEAKER_01]: So I was starting with Dickson and then moving to Ryan.
18:21.011 --> 18:28.015
[SPEAKER_00]: Well, so for me, throughout history, throughout my history, I've tried hard to
18:29.367 --> 18:33.692
[SPEAKER_00]: not rely on Google if I can help it.
18:33.873 --> 18:45.246
[SPEAKER_00]: So with Majestic, Majestic did the same thing the Google did, it goes round every website and then looks like the links on their website and a base robot.txt and
18:46.788 --> 19:01.908
[SPEAKER_00]: And it is that kind of tool, it sees one of the biggest crawlers on the internet, but that meant it could build a technology and a system that didn't rely on ranked checking, because, you know, we always saw that ranked checking.
19:03.290 --> 19:09.654
[SPEAKER_00]: had a dangerous aspect to it, a risky aspect to it.
19:10.854 --> 19:21.901
[SPEAKER_00]: I mean, when we saw started out, of course, we had 20 different search engines, so it was a whole different game, and there was only one in town with a little bit of a kickback from Bing or Dr. Go.
19:22.341 --> 19:28.104
[SPEAKER_00]: Then suddenly became, well, there's one monopoly and they could cut off the source in many different ways.
19:31.526 --> 19:53.877
[SPEAKER_00]: ways in which we're not going to get affected by Google, and that's one of the great things about AI coming out, because all of a sudden now I can build a technology that's got Claude as a competitor, open AI as a competitor, and complexity as a competitor, as well as Gemini, and that gives me a game.
19:54.557 --> 20:04.242
[SPEAKER_00]: because no, you know, there's no Google, there's no cloud that wants to build the tools that are going to tell people about all four of those people.
20:04.322 --> 20:22.892
[SPEAKER_00]: So it gives me an opening, unfortunately, it's given a thousand other people in opening, but the vast majority of building a vibe-coated solutions which will fall at the first hurdle when someone like Google starts blocking the search results.
20:25.412 --> 20:26.113
[SPEAKER_00]: That's a good thing.
20:26.133 --> 20:26.393
[SPEAKER_00]: Yeah.
20:26.934 --> 20:32.901
[SPEAKER_01]: How about you with, uh, with with, with, with, super card house, uh, house, uh, Google's reluctance with data affected you.
20:33.561 --> 20:34.342
[SPEAKER_02]: It's a mixed bag.
20:34.362 --> 20:37.926
[SPEAKER_02]: We've, uh, so super kind of relies on, like Dixon said earlier, we rely on APIs.
20:38.026 --> 20:39.548
[SPEAKER_02]: We don't actually go out and scrape your website.
20:39.568 --> 20:43.372
[SPEAKER_02]: We let other people worry about getting past Google and solve it and we rely on APIs for that.
20:43.393 --> 20:44.994
[SPEAKER_02]: And there's a bunch of players out there in this space.
20:46.156 --> 20:48.038
[SPEAKER_02]: When it comes to LLM data, we don't scrape that.
20:48.058 --> 20:57.367
[SPEAKER_02]: We use the LLM APIs directly, all absorbed into the cost so that we go right to Gemini or Google's API or OpenAI's or Proplexities or Clouds to get the LLM data.
20:58.128 --> 21:00.070
[SPEAKER_02]: I know there's controversy in that in the marketplace.
21:00.811 --> 21:02.633
[SPEAKER_02]: I believe the API data is better.
21:02.913 --> 21:03.894
[SPEAKER_02]: A lot of people want to scrape.
21:04.757 --> 21:05.758
[SPEAKER_02]: API gives you more info.
21:06.118 --> 21:09.360
[SPEAKER_02]: Like for example, Google tells you the query fan out when you hit it through the API.
21:09.480 --> 21:12.281
[SPEAKER_02]: So I think it doesn't tell you when you scrape the AI overview, right?
21:12.341 --> 21:13.442
[SPEAKER_02]: So I like the more data.
21:13.522 --> 21:14.883
[SPEAKER_02]: So I love using the APIs for that.
21:16.324 --> 21:18.245
[SPEAKER_02]: In some spots, we'll actually scrape the AI over YouTube.
21:18.285 --> 21:19.466
[SPEAKER_02]: We'll have a third part to do it.
21:20.006 --> 21:23.748
[SPEAKER_02]: But I'm seeing craziness all over the third part of the results, right?
21:23.848 --> 21:28.851
[SPEAKER_02]: Like for redundancy sake, I'm integrated with four major providers in Surprecon, right?
21:28.891 --> 21:31.793
[SPEAKER_02]: And so if one goes down or something, we can compare them and see the difference.
21:33.075 --> 21:37.872
[SPEAKER_00]: And we've seen the same problems with three providers at least, yeah.
21:38.112 --> 21:38.434
[SPEAKER_02]: Yeah.
21:39.320 --> 21:44.082
[SPEAKER_02]: And I'll run reports two minutes apart and I'll get 80% differences in the search results.
21:44.122 --> 21:50.865
[SPEAKER_02]: Now, that doesn't break my tool because we're analyzing the actual websites of who ranks, not the rankings themselves.
21:51.165 --> 22:01.610
[SPEAKER_02]: And so even if it's not the same rule here, you see on Google, we're still getting a good text and semantic analysis of the relevant copy and content to give you recommendations.
22:01.650 --> 22:04.631
[SPEAKER_02]: So it doesn't really break us if the order is wrong in a search result.
22:06.567 --> 22:08.488
[SPEAKER_02]: I think a lot of tools are going to need to pivot.
22:08.989 --> 22:14.872
[SPEAKER_02]: I'm already anticipating a pivot where maybe I give you, maybe I give a different search engine, right?
22:14.892 --> 22:16.633
[SPEAKER_02]: Because again, I don't care about the order.
22:16.653 --> 22:18.775
[SPEAKER_02]: We're not a ranked tracker ourselves.
22:18.795 --> 22:23.798
[SPEAKER_02]: We're just using ranking to get to the content so we can analyze the content and give you semantic references.
22:24.599 --> 22:28.841
[SPEAKER_02]: Maybe I do a Chrome plugin that lets you analyze any search results you're on on any engine, right?
22:28.861 --> 22:32.203
[SPEAKER_02]: That's something I've started developing just in case it goes that way.
22:32.864 --> 22:34.025
[SPEAKER_02]: But I think we're going to have to find other
22:35.201 --> 22:36.082
[SPEAKER_02]: other routes, right?
22:36.282 --> 22:41.688
[SPEAKER_02]: The tools, the tools that mainly rely on ranked tracking, which I haven't used a pure ranked tracker in years.
22:42.148 --> 22:44.711
[SPEAKER_02]: I've relied on Google Search Console as the search chief for ranked tracking.
22:44.731 --> 22:48.755
[SPEAKER_02]: I just haven't used a stand-alone ranked tracker in forever, right?
22:48.855 --> 22:51.418
[SPEAKER_02]: I use the tools to get the ranking data to other stuff with.
22:51.898 --> 22:55.342
[SPEAKER_02]: But I think the ranked trackers are going to have to pivot.
22:55.362 --> 22:57.965
[SPEAKER_02]: Like Dickson said earlier, they don't have much runaway left.
23:00.608 --> 23:03.151
[SPEAKER_02]: It's going to be interesting in it for the ranked trick, is really.
23:05.133 --> 23:06.935
[SPEAKER_02]: It's going to be a giant cat and mouse game, right?
23:06.975 --> 23:10.459
[SPEAKER_02]: Like, somehow you have to serve results to a user that aren't poisoned.
23:10.579 --> 23:15.865
[SPEAKER_02]: You poison, you know, 2% of users' results and then Google takes a huge hit and you lose market share.
23:16.285 --> 23:18.288
[SPEAKER_02]: Like you can't poison real users' results.
23:18.908 --> 23:20.888
[SPEAKER_02]: So, what are they going to do, right?
23:20.909 --> 23:22.269
[SPEAKER_02]: There's always going to be game of cat and mouse.
23:22.329 --> 23:26.090
[SPEAKER_02]: If real users get real data, then there's going to be a way for tools to get real data.
23:26.130 --> 23:28.090
[SPEAKER_02]: But it's going to be a lot of costs that goes into that.
23:28.270 --> 23:40.493
[SPEAKER_01]: Well, when we talk about even a small percentage, a fraction of a percentage of traffic that might be poisoned at Google, we're talking about an awful lot of traffic, just Google's scale.
23:42.635 --> 23:54.031
[SPEAKER_01]: Can regular users trust data that's coming from Google search results sets or from search results sets compiled by AI tools that are scraping from Google?
23:55.482 --> 24:00.485
[SPEAKER_02]: I wouldn't trust any AI tool that's great Google right now, because you don't know what it's getting back from Google.
24:00.986 --> 24:13.293
[SPEAKER_02]: And that includes everything that name Gemini, to be honest, you know, I don't know, I'm not going to officially claim they're all scraping or not scraping right for legal purposes, but there's been allegations of all of them scraping.
24:13.774 --> 24:19.337
[SPEAKER_02]: And if those allegations are true, then I wouldn't trust anything not from Google that's pulling from Google.
24:20.718 --> 24:23.580
[SPEAKER_03]: And that also leads us to the question of the prompt tracking,
24:25.952 --> 24:27.738
[SPEAKER_03]: So what do your thoughts on prompt drinking?
24:27.958 --> 24:30.607
[SPEAKER_03]: Have my thoughts, but to share the guest, I'll let you give your thoughts.
24:31.658 --> 24:34.100
[SPEAKER_00]: I said, yeah, okay, you guys will respond.
24:34.600 --> 24:36.302
[SPEAKER_02]: Oh, I'll get first.
24:36.662 --> 24:40.585
[SPEAKER_02]: I read a statistics somewhere, and maybe it was somebody at a LLM company that told me.
24:41.066 --> 24:45.810
[SPEAKER_02]: But it's more than half of prompts I've never been prompted before, and they will never be prompted again, right?
24:45.850 --> 24:48.011
[SPEAKER_02]: Dixing head on that for unique.
24:48.612 --> 24:54.116
[SPEAKER_02]: And so on my tools, I have to write Supercom and then FreeSOV.com, which is really free.
24:54.136 --> 24:55.978
[SPEAKER_02]: Just plug in your API keys, write Pay that cost.
24:56.438 --> 24:56.638
[SPEAKER_02]: But,
24:57.499 --> 25:03.002
[SPEAKER_02]: I don't look at the actual prompts so much, we do, everybody looks at the actual prompt, but that's kind of a vanity metric to me.
25:03.682 --> 25:22.930
[SPEAKER_02]: What I more care more about are what are the topics, what are the themes, what are the entities, what is that semantic gap, what is the user intent, all of that stuff around the prompts, and then how do I optimize for a topic or an entity, not so much as specific prompt, because the real user prompt is going to be 52 words long and
25:25.838 --> 25:27.520
[SPEAKER_02]: So that's not what I want to optimize for.
25:27.540 --> 25:32.065
[SPEAKER_02]: So that's my approach, and Dixon, I know with WAKA, you probably have a lot more to add to that.
25:32.085 --> 25:33.307
[SPEAKER_02]: So I'll shut up and let you add to it.
25:34.122 --> 25:40.305
[SPEAKER_00]: No, I mean, you know, that we both kind of see similar and similar on this idea.
25:41.306 --> 25:52.932
[SPEAKER_00]: And yeah, so the example I kind of give to potential customers about trying to do prompt volume is, well, the most popular volume is replies to this email, or a popular prompt might be replies to this email.
25:54.092 --> 25:56.373
[SPEAKER_00]: It's the context in which somebody asked something.
25:57.294 --> 26:01.296
[SPEAKER_00]: And usually, it's after a lot of chats with that,
26:04.017 --> 26:07.561
[SPEAKER_00]: Claude or that open our eye session that they're doing at the time.
26:08.121 --> 26:14.326
[SPEAKER_00]: Now that when it comes to prompt tracking, as Ryan says, causes all sorts of problems and
26:15.497 --> 26:31.729
[SPEAKER_00]: But it also is why there was a drive by the industry to say, well, you should be using a user browser kind of methodology to collect this information because it's more like what the user is typing in.
26:32.330 --> 26:35.072
[SPEAKER_00]: My logic is very much like Ryan's in that, you know.
26:35.832 --> 26:47.159
[SPEAKER_00]: what the user is typing, you can't emulate in any way shape or form because it's not only unique, but it's also unique to that session as well that they've said before and what they say afterwards.
26:47.780 --> 26:55.725
[SPEAKER_00]: So if you're using a browser that's asking a question about aeroplanes one day and horses the next, then you're going to mess everything up.
26:56.165 --> 26:58.446
[SPEAKER_00]: So I was absolutely against
27:00.328 --> 27:03.933
[SPEAKER_00]: product tracking using browser data.
27:04.053 --> 27:09.540
[SPEAKER_00]: And I exactly like you, I use the APIs.
27:09.940 --> 27:12.564
[SPEAKER_00]: And so we did use the APIs for prompt tracking.
27:13.265 --> 27:13.445
[SPEAKER_00]: And
27:15.946 --> 27:20.249
[SPEAKER_00]: Yeah, but then you also have got to do what the users are asking for in the customers as we're asking for it.
27:20.269 --> 27:25.853
[SPEAKER_00]: So we did it reluctantly it actually costs a lot less to scrape.
27:26.173 --> 27:36.000
[SPEAKER_00]: Well, I wonder why and it doesn't cost Well, it costs a lot less to scrape if you're buying buying through a data provider that is doing not a mess.
27:36.480 --> 27:38.501
[SPEAKER_00]: but it's a short term game.
27:39.581 --> 27:44.223
[SPEAKER_00]: I think customers are going to have to get real to say this data is going to cost.
27:44.604 --> 27:47.025
[SPEAKER_00]: They should probably focus back on the API options.
27:50.817 --> 27:53.259
[SPEAKER_00]: And yes, it removes all the usability.
27:53.639 --> 27:58.843
[SPEAKER_00]: What it also should do for you is say, well, actually, I don't want to track a thousand phrases.
27:59.383 --> 28:01.565
[SPEAKER_00]: To be honest with you, 15 does a pretty good job.
28:01.585 --> 28:03.006
[SPEAKER_00]: I've got five products on my site.
28:03.446 --> 28:10.111
[SPEAKER_00]: I'll do three kind of sample phrases for those things, and that will give me a spread of weather.
28:10.431 --> 28:14.975
[SPEAKER_00]: And AI has properly found me in the list of brands that it recommends.
28:15.755 --> 28:17.897
[SPEAKER_00]: And from that, I can work out share of voice and stuff.
28:19.758 --> 28:26.002
[SPEAKER_00]: buy a set by running from a scraper in, you know, and choose Toronto.
28:26.022 --> 28:26.783
[SPEAKER_00]: I'll go for Toronto.
28:29.306 --> 28:32.627
[SPEAKER_02]: I think you had an interesting point where give the customers what they want.
28:32.687 --> 28:40.989
[SPEAKER_02]: And I feel like the entire AI prompt tracking industry has gone this way about, I mean, it's mostly vanity metrics, like look at sentiment, right?
28:41.149 --> 28:49.611
[SPEAKER_02]: Everything is talking about sentiment, because every CEO wants to know, they don't care if they should show up, they don't care that they can't correlate it to sales because no one's doing media mix modeling yet.
28:49.751 --> 28:53.452
[SPEAKER_02]: Despite all of us at agencies saying, you need to be doing media mix modeling, right?
28:53.752 --> 28:55.172
[SPEAKER_02]: Let's create a machine that we're trying to do.
28:55.212 --> 28:56.252
[SPEAKER_02]: We're all trying to do it.
29:00.013 --> 29:18.221
[SPEAKER_02]: But every CEO just wants to know what's the same good things to bad things about my brand is it Recognizing like why don't I show up when it says this and my showing up when it says this right now That's really actionable, but it's the question CEOs want to know right like sentiment is my pep peeve because Every age like all the prompt models the system prompts leak.
29:18.241 --> 29:23.284
[SPEAKER_02]: There's a GitHub with the system prompt for every model out there It's it leaked I somehow clearly did all
29:24.024 --> 29:28.628
[SPEAKER_02]: They're all told to be neutral unless the user specifically asking positive or negative about a brand.
29:28.708 --> 29:30.310
[SPEAKER_02]: They're all told to treat the brand neutral.
29:30.630 --> 29:33.533
[SPEAKER_02]: And so when we measure sentiment, what are we actually measuring?
29:35.007 --> 29:39.730
[SPEAKER_00]: Absolutely, and that's this is what we didn't put sentiment into into the way KT.
29:39.770 --> 29:44.673
[SPEAKER_00]: We have fact checking hallucination tracking when it gets in wrong we want to know and want to fix it.
29:45.154 --> 29:56.481
[SPEAKER_00]: But the sentiment stuff seemed to seem a bit absurd because what we the way we saw it was that if you're going to ask, you know, what's the best brand of, of
29:57.717 --> 30:03.999
[SPEAKER_00]: hairdryer, then it's going to come with a bunch of brands and it's going to say what's good about each brand as best as it can.
30:04.439 --> 30:08.800
[SPEAKER_00]: So we'll err on the side of telling you, it's trying to tell you what's different between the brands.
30:08.820 --> 30:10.701
[SPEAKER_00]: It's not trying to tell you which one's good and bad.
30:12.121 --> 30:14.342
[SPEAKER_00]: So if it's bad, it's not in the list.
30:15.023 --> 30:19.388
[SPEAKER_00]: So I agree, it's sentiment, it seems an odd thing for people to do.
30:19.788 --> 30:28.298
[SPEAKER_00]: The other thing I think is odd about prompt tracking is that, for me, it's like having the management accounts, the month after you've done the work.
30:29.091 --> 30:35.099
[SPEAKER_00]: In other words, the SEO doesn't tell the SEO what they've got to do at all.
30:35.640 --> 30:40.286
[SPEAKER_00]: It just says to the board, did the SEO good job.
30:40.426 --> 30:44.091
[SPEAKER_00]: So as SEOs, that's not where we should be looking at all.
30:44.491 --> 30:46.894
[SPEAKER_00]: We should be looking at things like,
30:47.855 --> 31:10.540
[SPEAKER_00]: what Ryan came out with today to give a shout out on the analysis of a paper or paper was it that the complexity they're out with, which pretty much gives some huge clues for SEOs as to how
31:15.072 --> 31:17.873
[SPEAKER_02]: Yeah, just to dive in on that in 15 seconds for everybody.
31:18.313 --> 31:24.695
[SPEAKER_02]: Proplexity put out a brag, like a research paper blog post about how they increase the speed of forplexity, which boring, right?
31:25.035 --> 31:29.476
[SPEAKER_02]: But buried in that article, they actually described how the algorithm works.
31:29.577 --> 31:34.178
[SPEAKER_02]: They've talked about how they do retrieval fine candidates to cite for citations and mentions.
31:34.718 --> 31:36.418
[SPEAKER_02]: Right, and they gave us all the details.
31:36.438 --> 31:37.659
[SPEAKER_02]: Like, here's the metrics we use.
31:37.719 --> 31:38.999
[SPEAKER_02]: Here's how we use these metrics.
31:39.039 --> 31:41.300
[SPEAKER_02]: This is exactly how we take a candidate to site.
31:41.860 --> 31:47.501
[SPEAKER_02]: And all the PR places ran about proplexities faster and like DSEO's missed this one.
31:47.581 --> 31:51.462
[SPEAKER_02]: So if you go to the SurpreyCon blog, I did a whole deep dive.
31:51.542 --> 31:54.002
[SPEAKER_02]: It is going to take you about 20 some minutes to read this.
31:54.822 --> 31:56.603
[SPEAKER_02]: But there's takeaways and checklists and everything.
31:57.323 --> 32:05.870
[SPEAKER_02]: I dove deep into the research paper in their blog post and broke it down and plain English for, here's what they do and here's what you need to do if you want to write in perplexity.
32:05.990 --> 32:11.455
[SPEAKER_02]: And I know what people are saying, who uses perplexity, but all the LLMs work the same, right?
32:12.175 --> 32:14.237
[SPEAKER_02]: Retrieval and rag is a science.
32:14.537 --> 32:17.620
[SPEAKER_02]: There's best practices for doing it and there's best ways of doing it.
32:17.680 --> 32:23.965
[SPEAKER_02]: So while they may not do it exactly the same, you can bet every other LLLM does it in very similar ways.
32:24.841 --> 32:28.882
[SPEAKER_03]: So there's a, I'm sorry, it's just not your own person.
32:29.162 --> 32:30.763
[SPEAKER_03]: They just show how I met your mother.
32:30.863 --> 32:34.904
[SPEAKER_03]: I don't know if Dickson sort of it, but was popular in the US on reruns all the time.
32:35.385 --> 32:39.966
[SPEAKER_03]: And there's an episode where there's a girl who's like really clueless and they call, they just go, oh honey.
32:40.666 --> 32:51.590
[SPEAKER_03]: And I just feel like they have no idea the spam that is coming for them in general anyway, but now that they're releasing this kind of information, like what do you think get around like by Christmas?
32:52.795 --> 32:55.837
[SPEAKER_02]: Yeah, I mean, sadly though, the spam has to be useful.
32:55.877 --> 32:57.497
[SPEAKER_02]: Like if you look at how they work, right?
32:57.538 --> 33:01.560
[SPEAKER_02]: Like the spam has to be very useful in relevant.
33:01.760 --> 33:06.422
[SPEAKER_02]: So like it's, I mean, yeah, am I gonna try this spam at 100% I've already started.
33:06.742 --> 33:10.844
[SPEAKER_02]: But, um, but, but, but, but yeah, it's, yeah.
33:13.125 --> 33:18.187
[SPEAKER_03]: I mean, the PPC of search, I'm sorry, I didn't mean to step on you there, Nixon.
33:18.467 --> 33:23.788
[SPEAKER_03]: From a typical porn and casino so survives since 1996, so they're going to continue to survive.
33:23.828 --> 33:25.049
[SPEAKER_03]: They're going to go ahead, Nixon.
33:26.229 --> 33:41.214
[SPEAKER_00]: No, it does tell quite nicely to you, I talked that the guy before me on Brighton last week, like I call Tom Wells, he lives in Kazakhstan of all places, but he did some interesting analysis.
33:41.854 --> 33:54.718
[SPEAKER_00]: between, instead of testing, you know, whether you come back in results, he tested compare brand A to brand B, which one do things better, basically, for different models and for different generations of models.
33:55.078 --> 34:00.139
[SPEAKER_00]: And then started to collate information that came back.
34:00.660 --> 34:07.202
[SPEAKER_00]: What was interesting that Dove tells with one of the some of the things that was on Ryan's post is that he found
34:08.762 --> 34:10.824
[SPEAKER_00]: Soptucle, he was testing the SaaS businesses.
34:10.984 --> 34:20.574
[SPEAKER_00]: He said, soptuccompliance, they gave a massive boost to, if your business was soptucle-plank, and the other one wasn't that was a big one.
34:21.415 --> 34:24.418
[SPEAKER_00]: If you had 4.5 out of five ratings,
34:26.920 --> 34:34.923
[SPEAKER_00]: and 20,000 reviews, is that better or worse than 4.9 out of 5 ratings and only 20 reviews?
34:35.223 --> 34:37.784
[SPEAKER_00]: And some engines went one way, some engines went another.
34:38.204 --> 34:51.508
[SPEAKER_00]: But on all of them, there seemed to be limits to how well you could go on any one kind of metric and it kind of stopped and you talked about that in yours, I think, Ryan, where certain things
34:51.908 --> 34:56.690
[SPEAKER_00]: I let it and they wouldn't give you more bonus points just because you are more of something or other.
34:58.511 --> 34:58.831
[SPEAKER_00]: 100%.
34:58.871 --> 35:03.694
[SPEAKER_02]: I want to talk about the technical details.
35:03.714 --> 35:06.635
[SPEAKER_02]: I want to talk about the technical details, the reciprocal rank fusion.
35:06.655 --> 35:08.056
[SPEAKER_02]: But we're not going to do that on this post.
35:10.057 --> 35:14.419
[SPEAKER_01]: Well, we're in the neighborhood that we may as well dive into the complexity.
35:14.779 --> 35:17.120
[SPEAKER_01]: And there, but it's the photon algorithm.
35:17.500 --> 35:17.680
[SPEAKER_01]: Yeah.
35:19.830 --> 35:25.493
[SPEAKER_01]: You're saying that if complexity is doing it, it's likely that other LLMs are doing something similar, correct?
35:26.193 --> 35:27.994
[SPEAKER_02]: Yes, because this is a science, right?
35:28.915 --> 35:30.696
[SPEAKER_02]: There are best ways math, it's math.
35:30.876 --> 35:33.137
[SPEAKER_02]: There are best mathematical ways to do stuff.
35:33.437 --> 35:37.720
[SPEAKER_02]: And honestly, it works very, very similar to the way a search engine works.
35:37.920 --> 35:43.603
[SPEAKER_02]: I mean, they're using BM25, which is the first scoring metric every search engine uses.
35:44.163 --> 35:44.363
[SPEAKER_02]: right?
35:44.443 --> 35:46.064
[SPEAKER_02]: They're waiting different fields separately.
35:46.484 --> 35:48.105
[SPEAKER_02]: They're doing passage indexing, right?
35:48.145 --> 35:52.526
[SPEAKER_02]: Google's done passage indexing since almost 10 years now, 20, you know, right?
35:53.707 --> 35:55.968
[SPEAKER_02]: They're basically building a search engine.
35:55.988 --> 35:58.949
[SPEAKER_02]: They're doing everything that Google does that we're seeing, right?
35:58.969 --> 36:01.830
[SPEAKER_02]: They're breaking their doing factors semantic relevance, right?
36:03.210 --> 36:09.513
[SPEAKER_02]: You know, and they're scoring, I guess the one takeaway for SEO is they're scoring your page based on your top matching passage.
36:10.053 --> 36:32.587
[SPEAKER_02]: Right, your entire page score is the best passage on your page to that user's query and how that is is very multiple scores BM25 which means basically it's like keyword density but padding doesn't help right it's semantic relevance it's saying the same words in the same orders right if your query coverage isn't the isn't what people are looking for.
36:33.187 --> 37:00.227
[SPEAKER_02]: If a user is looking for a specific number or term and you don't have that specific number of term, you're kind of boneed and then you know it's got to be a clear-concinct passage and not all over the web page so just relevant isn't enough you got to have that passage so there's a million takeaways in what they're doing but when I first read it I said holy crap this is a search engine like they're literally doing everything I know about from every book I've ever read about information or travel on so it was really really interesting to me to dive in
37:01.142 --> 37:07.243
[SPEAKER_03]: My property is, I'm sorry, Jim, the property is Google's ranked search and some of their citation.
37:07.764 --> 37:08.744
[SPEAKER_03]: Work where the others don't.
37:08.784 --> 37:11.745
[SPEAKER_03]: There is no ranking signal on the other types of those ones, correct?
37:13.305 --> 37:14.525
[SPEAKER_02]: Um, unknown.
37:16.045 --> 37:17.926
[SPEAKER_03]: I just can't that's why they're always so different, right?
37:17.946 --> 37:22.347
[SPEAKER_03]: Because Google was supposed to add them to their AI products, but haven't seen it happen yet.
37:23.714 --> 37:29.800
[SPEAKER_01]: Well, what will says it uses ranking placements to ground the answers that it's giving.
37:29.980 --> 37:37.327
[SPEAKER_01]: It might not be drawing the answers from those top 10, but it's using that top 10 to figure out if the answers it's generating are accurate.
37:38.658 --> 37:39.939
[SPEAKER_00]: Well, I don't think that's true.
37:40.019 --> 37:46.005
[SPEAKER_00]: I think it's, so the query fan out, kind of knocks out slight sideways.
37:46.045 --> 37:54.512
[SPEAKER_00]: So with the Gemini API, you've got a toggle on the Gemini API and you can decide whether to have a grounded,
37:55.593 --> 37:56.915
[SPEAKER_00]: response or not.
37:56.935 --> 38:13.737
[SPEAKER_00]: So you can actually ask Gemini and we do an article with them without, you know, with the search with the search, look up, tell me this, without a search, a look up, tell me this, and comparing the difference, which is a very interesting kind of comparison, suddenly enough, sometimes it gets better without looking up the internet.
38:14.137 --> 38:19.205
[SPEAKER_00]: but as Ryan says, Gemini then gives you a query fan out.
38:19.245 --> 38:22.430
[SPEAKER_00]: So what they're doing is changing the queries that they're looking up.
38:22.530 --> 38:29.240
[SPEAKER_00]: So if you say something, it's going to do a little bit of thinking, get rid of your tie pose because you're bound to have made one, especially if you're me.
38:30.000 --> 38:55.705
[SPEAKER_00]: uh... it changes around a little bit i'll find out what the connection is said therefore understanding the user in ten then come up with a you know ten twelve however many other queries then it looks all of those up so now it's doing twelve lookups but none of them are the one that you wanted and then it's retrieving those results and analyzing those and let's write those differently that i don't think they're looking up the ten blue lines i think they're looking up a hundred different blue lines
38:56.786 --> 39:02.792
[SPEAKER_02]: 100% and they're starting with your template for a summary and then, you know, generating the original Google patent.
39:02.932 --> 39:08.237
[SPEAKER_02]: I covered this in my talk at SEO week and my pub count and state of search keynotes this year.
39:09.138 --> 39:14.523
[SPEAKER_02]: And the original Google patent on AI overview is very interesting because it would do the query fan out.
39:14.543 --> 39:15.744
[SPEAKER_02]: It would generate a response.
39:16.545 --> 39:20.653
[SPEAKER_02]: But then it wouldn't actually automatically cite something that it pulled back in the query fan out.
39:20.853 --> 39:27.946
[SPEAKER_02]: It would pull out snippets from its response and then go find a website that match that's snippet generated exactly and cite that website.
39:28.807 --> 39:29.627
[SPEAKER_02]: It's back into it.
39:30.807 --> 39:31.788
[SPEAKER_02]: That was the original patent.
39:31.848 --> 39:33.348
[SPEAKER_02]: I don't know that they're doing that anymore.
39:33.528 --> 39:43.751
[SPEAKER_02]: They were at the time that I presented this on stage many, many months ago, but that was interesting to me because it couldn't, at the time, it couldn't remember where it pulled that from when it generated its answer.
39:44.231 --> 39:48.392
[SPEAKER_02]: So it had to go back door into it and go like, Garg, here's what I made up and spit out.
39:48.472 --> 39:50.492
[SPEAKER_02]: Go find me a website that matches what I spit out.
39:50.812 --> 39:57.854
[SPEAKER_00]: And if there's no website that I'm pretty sure that was, they didn't have a lot of choice because the machine learning doesn't keep the URLs
39:58.774 --> 39:59.614
[SPEAKER_00]: How do you boil an egg?
39:59.674 --> 40:03.636
[SPEAKER_00]: It learned from a million different places and it's got the data.
40:03.736 --> 40:14.241
[SPEAKER_00]: So if you look at the Gemini stuff in training data mode, then it doesn't give citations in the API very often.
40:14.261 --> 40:16.562
[SPEAKER_00]: It'll maybe give you the website that you're talking.
40:16.582 --> 40:24.027
[SPEAKER_00]: about but generally it won't give citations and whereas in the granted mode it comes back with loads of citations.
40:24.147 --> 40:43.221
[SPEAKER_00]: So you're absolutely right the has to go and find it and one of the reasons it has to find it is because as SEO's one of our biggest bug bears would have been if AIR reviews hadn't got citations because then we would have come out then with you know hell have no fury like us SEO's school and gun that's
40:44.182 --> 40:44.682
[UNKNOWN]: That's true.
40:45.422 --> 40:47.063
[SPEAKER_02]: Sure, there's a legal implication too, right?
40:47.083 --> 40:53.104
[SPEAKER_02]: Like if Google just gave us the answer and didn't say where it got it from now, all of a sudden they're liable, but now they can be like, oh, no, it's from this website.
40:53.124 --> 40:53.764
[SPEAKER_02]: See, here's a link.
40:53.784 --> 40:54.544
[SPEAKER_02]: We fold it from there.
40:54.564 --> 40:54.784
[SPEAKER_02]: Yeah.
40:54.804 --> 40:55.564
[SPEAKER_02]: Not our problem.
40:55.584 --> 40:57.245
[SPEAKER_02]: We told you to put glue in your pizza.
40:58.085 --> 40:58.405
[SPEAKER_02]: Remember that?
40:58.425 --> 40:59.425
[SPEAKER_03]: Yeah.
40:59.865 --> 41:07.387
[SPEAKER_03]: And also originally not keeping URLs where they took the content from less and their ability to be proved as copyright infringers in court.
41:09.261 --> 41:22.234
[SPEAKER_00]: Yeah, well, I mean, the things they don't have to because they've already got indexed so they can call on the fly where it's every, you know, the open AIs and this will don't They're building one, they're building one, yeah, of course, yeah, yeah, absolutely
41:24.237 --> 41:32.426
[SPEAKER_00]: Okay, so I hope I'm hoping that one of them is going to buy majestic for an awful lot of money and I can, you know, go and return on the little little iron.
41:32.786 --> 41:36.790
[SPEAKER_02]: So we can we can we can we lump super can in there too.
41:36.830 --> 41:37.872
[SPEAKER_02]: Like I'll take millions.
41:38.332 --> 41:38.953
[SPEAKER_02]: I'm okay with that.
41:40.695 --> 41:43.938
[SPEAKER_01]: Well, let's be clear here when he says millions.
41:43.998 --> 41:45.259
[SPEAKER_01]: He means tens of millions.
41:46.940 --> 41:49.742
[SPEAKER_01]: You want to do any sex?
41:49.762 --> 41:50.402
[SPEAKER_02]: I'll take that.
41:51.523 --> 41:53.964
[SPEAKER_01]: Well, that's what the search engines need to know about.
41:54.264 --> 41:57.846
[SPEAKER_01]: That's what the engines need to know about you.
41:57.886 --> 41:58.506
[SPEAKER_01]: But what do I say?
41:58.546 --> 42:06.951
[SPEAKER_01]: I always need to know about them based on what you, on coming up with the properties, search, ranking algorithms.
42:07.471 --> 42:10.512
[SPEAKER_01]: How does somebody come up with a general
42:16.117 --> 42:18.258
[SPEAKER_02]: Yeah, it's kind of the same way Google does it, right?
42:19.299 --> 42:22.340
[SPEAKER_02]: You know, they essentially the same way, right?
42:22.360 --> 42:28.224
[SPEAKER_02]: They come up with their search results from either one of their partners, or as people alleged scraping Google.
42:28.804 --> 42:30.605
[SPEAKER_02]: Again, people alleged that, I don't know, like that.
42:31.906 --> 42:32.206
[SPEAKER_02]: But...
42:32.946 --> 42:35.850
[SPEAKER_02]: They basically break it down into several different way to ones, right?
42:35.870 --> 42:37.532
[SPEAKER_02]: They retrieve a candidate list of sites.
42:38.052 --> 42:44.800
[SPEAKER_02]: They do their initial ranking, which is essentially just BM25, or as SDS think of it as term frequency, even though it's way more complicated than that.
42:46.142 --> 42:50.066
[SPEAKER_02]: They get that second stage ranking then up from that list, where they do semantic relevance.
42:50.086 --> 42:52.769
[SPEAKER_02]: This is where we're talking about vectors and entities and things like that.
42:53.410 --> 43:15.990
[SPEAKER_02]: to narrow it down, and then once they've narrowed that down, they pull in whatever the authority scores are, for a search engine, that's links and clicks, for a proplexity or open AI, no idea what their authority score is, I'm guessing it has to do with their internal knowledge graph, since they don't have click data or link data, so I'm guessing it's internal knowledge graph based, like how often is it mentioned or show up across the web and their knowledge graph, that's my guess.
43:16.990 --> 43:28.758
[SPEAKER_02]: And then, then they generate their summary based on that short list, you know, that stage four short list of answers, they generate their summary and sites the pages that survive into that fourth list.
43:28.918 --> 43:34.662
[SPEAKER_02]: Essentially, if I had to break it down in the super plain English without all the math text stuff, that's probably the best I can do.
43:35.747 --> 43:57.641
[SPEAKER_01]: Okay, this is the, I have a hundred other questions I want to ask both of you, but we do that, we got to keep moving forward, open-ended questions, it's a terribly open-ended question, but Dixon and then Ryan, get a total of search markers, I mean, the changes that are happening in search results in the way that, uh,
43:58.722 --> 44:10.512
[SPEAKER_01]: people that consumers are relating to, to search right now, if anything you had to tell like search marketers that they need to pivot towards, think about, or start working on, in relation to the tools that you make.
44:11.713 --> 44:13.314
[SPEAKER_01]: Oh, yes, I've got to think.
44:13.915 --> 44:21.842
[SPEAKER_00]: So, my thing that I think search marketers haven't got their ass into gear on is MCPs and skills.
44:22.702 --> 44:23.943
[SPEAKER_00]: And I'm not talking about
44:24.804 --> 44:26.985
[SPEAKER_00]: us using MCPs and skills.
44:27.825 --> 44:52.595
[SPEAKER_00]: I'm talking about us building MCPs and skills and so I see that ecosystem as very much like Apple one day had to mobile phones and then they showed the Apple store and everybody started building apps that you download in the Apple store all of a sudden if your app was on the phone you had sticky a sticky relationship with that app with that iPhone
44:53.355 --> 44:55.816
[SPEAKER_00]: and I think that skills are the same sort of thing.
44:56.236 --> 45:15.942
[SPEAKER_00]: If you can get your business skills or MCP onto a user's version of Claude and they're using it day-to-day, then they're going to be much stickier as a customer of yours and also you're going to increase their effectiveness on Claude as you look at something decent.
45:16.462 --> 45:26.368
[SPEAKER_00]: So I think that marketers in general should be thinking about MCPs as a distribution channel, as a channel to market.
45:26.448 --> 45:26.768
[SPEAKER_01]: Right.
45:27.028 --> 45:27.869
[SPEAKER_01]: Same question.
45:28.815 --> 45:30.036
[SPEAKER_02]: I agree with Dixon.
45:30.116 --> 45:37.583
[SPEAKER_02]: There's been a whole lot of what Sheldon and Big Bang were called Hookham out there on stuff that helps LLMs rank, right?
45:37.863 --> 45:45.469
[SPEAKER_02]: Whether it's LLMs.Tex, it no one's really using, or some other platforms and protocols I won't name or chain people with, but they're out there, right?
45:45.509 --> 45:50.173
[SPEAKER_02]: There's just a big study about something someone made up that said work, but no one's using it, right?
45:51.054 --> 45:51.715
[SPEAKER_02]: But they're telling us.
45:52.596 --> 45:55.080
[SPEAKER_02]: This is how to interact with us at MCP, right?
45:55.140 --> 45:58.025
[SPEAKER_02]: So if we're like, no, no, no, what about if I create a catch.txt?
45:58.045 --> 45:58.566
[SPEAKER_02]: What about that?
45:58.606 --> 46:00.049
[SPEAKER_02]: We'll use that, right?
46:00.269 --> 46:02.072
[SPEAKER_02]: It's, uh, we're kind of crazy.
46:02.433 --> 46:02.754
[SPEAKER_02]: I wish.
46:04.128 --> 46:09.513
[SPEAKER_02]: My biggest pet peeve is how many SEOs and marketers out there have no understanding how a search engine or LLM works at the technical level.
46:09.934 --> 46:20.185
[SPEAKER_02]: I think that's becoming more required to do the job now, the days of following the checklist somebody else wrote, they're still effective, but you're going to always be lagging behind.
46:20.786 --> 46:23.469
[SPEAKER_02]: So I think having a knowledge of how the technology works,
46:24.810 --> 46:39.660
[SPEAKER_02]: But I think my advice would be, you know, pivot, pivot fast and own it, the GEO space has seen too many PR companies and social media companies come in and own stuff that the SEO is kind of dropped the ball on should have been owning so.
46:40.560 --> 46:42.742
[SPEAKER_02]: understand pivot, pivot fast.
46:43.423 --> 46:44.905
[SPEAKER_02]: A search might not be a destination.
46:44.945 --> 46:46.686
[SPEAKER_02]: It might not be 10 blue results in the future.
46:46.706 --> 46:49.809
[SPEAKER_02]: It might not be a person searching, but something's going to be searching.
46:49.830 --> 46:54.775
[SPEAKER_02]: And as long as there's an answer, somebody gets in whatever format, we're going to try to influence it.
46:55.055 --> 46:58.698
[SPEAKER_02]: And so, you know, stay on top of it to keep trying to influence it.
46:59.399 --> 47:07.960
[SPEAKER_01]: Okay, well, Dixon Jones from InNinks and our week, and Ryan Jones, I see you guys present of S.H.O.R.A.R.A.
47:07.980 --> 47:10.241
[SPEAKER_01]: to first gentlemen, thank you both for joining us today.
47:10.941 --> 47:11.601
[SPEAKER_01]: Thanks for having me.
47:11.621 --> 47:12.221
[SPEAKER_03]: Have a good year, guys.
47:12.481 --> 47:13.941
[SPEAKER_03]: Yeah, I'll have to have you back on.
47:16.082 --> 47:16.422
[SPEAKER_03]: Okay.
47:17.002 --> 47:17.622
[SPEAKER_01]: I love having me.
47:17.662 --> 47:19.022
[SPEAKER_01]: I love having those guys on.
47:19.202 --> 47:21.683
[SPEAKER_01]: It, like you put no notes.
47:21.703 --> 47:22.863
[SPEAKER_01]: They were actually so interested.
47:22.883 --> 47:24.023
[SPEAKER_01]: You went over time.
47:24.703 --> 47:28.524
[SPEAKER_01]: I would apologize to the audience for that, but I don't think it's actually necessary, eh?
47:30.018 --> 47:36.484
[SPEAKER_03]: No, no, no, no, and I did adjust the grid the day a bit so that we could accommodate because I knew it would go along.
47:37.444 --> 47:42.469
[SPEAKER_01]: Yeah, so now we've got to get back into like actual, you know, searched.
47:42.489 --> 47:51.937
[SPEAKER_01]: I mean, easy reporting the search the search news was was fun and uplifting and, you know, it was like the feeling you get after an interview like that.
47:52.277 --> 47:55.820
[SPEAKER_01]: You know, your head is just so full of interesting information.
47:59.492 --> 48:07.556
[SPEAKER_01]: Well now we get to talk about like the born Ukraine, uh, uh, uh, starlink and, um, you know, stuff like that.
48:07.817 --> 48:14.060
[SPEAKER_03]: So, well, one positive is the one story from your best today is, is not a, was a good thing.
48:14.080 --> 48:20.924
[SPEAKER_01]: So, um, possibly a good thing, although where it's being done is is more than suspect.
48:20.944 --> 48:21.384
[SPEAKER_03]: Yeah.
48:21.704 --> 48:22.024
[SPEAKER_03]: Yeah.
48:22.044 --> 48:22.204
[SPEAKER_03]: Okay.
48:22.224 --> 48:23.065
[SPEAKER_03]: It's never a good thing.
48:23.984 --> 48:26.206
[SPEAKER_01]: We're looking into the abyss again.
48:26.226 --> 48:28.268
[SPEAKER_01]: This is a becoming a weekly segment.
48:28.308 --> 48:38.739
[SPEAKER_01]: It's existed a weekly segment since the Trump regime started bad the affecting the economy, the internet, the environment that we work in.
48:40.501 --> 48:43.223
[SPEAKER_01]: Since tech started being a harboredger
48:48.774 --> 48:58.543
[SPEAKER_01]: control and, well, fascism as much as it's been a harbinger of great change in opportunity for for for the rest of us.
48:58.603 --> 49:02.007
[SPEAKER_01]: So we have the segment looking into the Abyss and this is way deep.
49:02.888 --> 49:08.653
[SPEAKER_01]: The other day, Star Lake was was completely cut off
49:12.048 --> 49:12.969
[SPEAKER_03]: the agreement.
49:13.049 --> 49:13.970
[SPEAKER_03]: Start for instance.
49:14.331 --> 49:22.259
[SPEAKER_01]: Start field the the military side of Starlink, which is apparently run by the president not by Starlink itself.
49:23.401 --> 49:23.881
[SPEAKER_01]: I digress.
49:24.502 --> 49:24.682
[SPEAKER_01]: Um.
49:25.663 --> 49:33.292
[SPEAKER_03]: Well, it's true there's a poor distinction because the musk is claiming that he had nothing to do with it because the Starfield, but also Starlink was also down so they're both done.
49:33.929 --> 49:44.543
[SPEAKER_01]: Okay, so anyway, after a tremendous series of Ukrainian drone strikes on Russian data centers,
49:47.155 --> 50:13.344
[SPEAKER_01]: which closed shutdown, I think three or four major data centers, major end-ex data centers, and had a subsequent effect of limiting far-right bot activity in Western social media, especially in the United States around the U.S. midterms in Canada, around Alberta separation, and in Europe, around
50:15.465 --> 50:17.966
[SPEAKER_01]: the effects of the Russian war on Ukraine?
50:18.926 --> 50:19.186
[SPEAKER_01]: Yes.
50:20.006 --> 50:23.108
[SPEAKER_03]: Just to clarify, Miannik's ever just were hit.
50:24.408 --> 50:26.249
[SPEAKER_03]: But a lot of people equate that to Google.
50:26.309 --> 50:31.791
[SPEAKER_03]: So some of the articles that I read for this today were, I go, it wasn't a big deal because it's just Google.
50:32.331 --> 50:33.673
[SPEAKER_03]: for Russia, but it's not that.
50:33.733 --> 50:41.543
[SPEAKER_03]: Yandex is critically involved in the primary ecosystem and Russia for mail, cloud storage maps, ride-hailing, GPT.
50:41.583 --> 50:44.286
[SPEAKER_03]: So unlike here, we have a lot of different companies to do something.
50:44.767 --> 50:47.390
[SPEAKER_03]: Yandex is deeply involved in the Russian infrastructure.
50:47.450 --> 50:49.653
[SPEAKER_03]: So when they say they hit the Yandex,
50:50.375 --> 51:02.776
[SPEAKER_03]: status center and the bot stop calling I saw some or some writers say oh that just that would have affected the bots but yes it would have because of how yandex is involved in the primary ecosystem of the web there.
51:03.096 --> 51:03.918
[SPEAKER_03]: So I'm just going to add that.
51:05.458 --> 51:17.105
[SPEAKER_01]: In North America, and in Western countries, our economies have engendered a great deal more competition than, say, Russia's has over the last 15 to 20 years.
51:17.866 --> 51:24.750
[SPEAKER_01]: We, NDEX has a massive, like, NDEX is a big part of Russia's digital footprint.
51:25.571 --> 51:32.615
[SPEAKER_01]: So taking down the three of their major data centers did have a major impact on
51:34.036 --> 51:41.562
[SPEAKER_01]: external activity from Russia, including killing a number of botnets that have infected social media.
51:41.582 --> 51:51.730
[SPEAKER_01]: Now, when I say killing, killing is a weird word on the internet because things could be killed and be resurrected just as quickly on another set of servers in another country somewhere else.
51:51.790 --> 51:53.451
[SPEAKER_01]: And that is happening.
51:53.511 --> 52:00.996
[SPEAKER_01]: So the right wing bots are filtering back onto the, onto the internet, but their absence was notable for a bit.
52:04.490 --> 52:28.324
[SPEAKER_01]: the internet was either cut off or limited for military applications that were moving faster than 160 miles an hour, like a drone on its way to shut down a data center that has to potential to affect the outcome of the US midterms, which is why I think the American government got
52:33.300 --> 52:43.445
[SPEAKER_03]: Right, so the operational impact was that on hubs in Sussalo, Caluga and Vladimir, they were temporary crippled, super computer data clusters.
52:44.745 --> 52:48.007
[SPEAKER_03]: And so those digital services were dependent on that.
52:48.327 --> 52:51.088
[SPEAKER_03]: Their researchers who don't believe that it actually had much of an effect.
52:51.468 --> 53:02.231
[SPEAKER_03]: And it would not have fully had an effect, but there might be areas, both might be true, there might be areas of the internet that he's dropped off because those happen to be in those super clusters.
53:02.731 --> 53:10.513
[SPEAKER_03]: And then others it didn't because super clusters exist in other places to take to fully take down Russia's centers.
53:10.573 --> 53:13.493
[SPEAKER_03]: I believe you'd have to 26 data centers, including other countries.
53:14.194 --> 53:17.694
[SPEAKER_03]: So yeah, so you did didn't fully take doubt, but they did enough that I think
53:21.275 --> 53:36.661
[SPEAKER_03]: But I'm saying researchers are saying no, and I'm just saying I think it's possible both our true researchers or researchers are much Usually, we see a much broader aspects, and there are probably certain unique issues like Twitter bots seem to disappear overnight and like
53:37.341 --> 53:50.592
[SPEAKER_03]: Reddit bots seem to disappear and they may have been held on those super clusters where other 24 super clusters were still alive and active so they did you got to remember there's not just official state players at Playing here.
53:50.612 --> 53:52.754
[SPEAKER_01]: There's a number of private independent.
53:52.934 --> 53:55.616
[SPEAKER_01]: They're called military bloggers in Russia.
53:55.676 --> 54:02.221
[SPEAKER_01]: Yeah, and it's like it's the Russian secret service that's
54:04.198 --> 54:08.019
[SPEAKER_01]: responsible for all of the bots, although they are responsible for a good deal of them.
54:09.380 --> 54:13.961
[SPEAKER_01]: It's a number of independent operators, a number of people who don't have nothing to do with stick security.
54:14.001 --> 54:22.543
[SPEAKER_01]: They're just making a back-off of trolling, and they're the ones who would get hit really hard by this because they're more likely to send children around the around the cheap solution.
54:23.104 --> 54:31.426
[SPEAKER_03]: And also, I forgot, I feel terrible because he was such a horrific man who rushed, if who Putin had killed in the plane, it was like his right-hand guy.
54:31.928 --> 54:41.113
[SPEAKER_03]: One of the original things he did was develop the internet research agency to create influence with social media bots and things in that nature back in the day.
54:41.733 --> 54:45.375
[SPEAKER_03]: So back I 10 years ago or 12 years ago, so he was a private.
54:45.876 --> 54:50.439
[SPEAKER_03]: not private, like Russia private, not our private, but he still was independent of food.
54:50.459 --> 54:52.261
[SPEAKER_01]: You're talking about the former mercenary leader.
54:52.361 --> 54:53.561
[SPEAKER_01]: I've forgotten his name as well.
54:53.601 --> 55:02.167
[SPEAKER_03]: We tried to beat the Kuigensta against, yeah, but originally he was a big developer of the ecosystem that did that's exactly what you're saying.
55:02.227 --> 55:07.491
[SPEAKER_03]: He would have been affected by something like this, but maybe Russian state, Asian state attacks were not of him.
55:08.612 --> 55:09.633
[SPEAKER_01]: Okay.
55:11.432 --> 55:13.654
[SPEAKER_01]: Bots got, Bots did get knocked.
55:14.074 --> 55:17.076
[SPEAKER_01]: I'd check them over the European Parliament.
55:17.136 --> 55:20.079
[SPEAKER_01]: Thomas, is it off here?
55:20.499 --> 55:34.790
[SPEAKER_01]: He put a post up on X, which said, if this is confirmed, and it's telling Russian data centers get hit, and online activity around the AFD, that basically the far, far right in Germany.
55:35.631 --> 55:37.712
[SPEAKER_01]: Activity around the AFD suddenly plummets.
55:38.693 --> 55:39.233
[SPEAKER_01]: It happened.
55:39.313 --> 55:40.154
[SPEAKER_01]: It happened again.
55:40.174 --> 55:40.814
[SPEAKER_01]: We can trace it.
55:41.194 --> 55:44.215
[SPEAKER_01]: The activity around Alberta separatists plundered for a bit of day.
55:45.955 --> 55:53.098
[SPEAKER_01]: Activity around right wing candidates in the American primaries declined for a bit of day.
55:53.158 --> 55:54.938
[SPEAKER_01]: But again, it's filtering back up again.
55:55.598 --> 56:04.641
[SPEAKER_03]: So we did see some big influence we were going to attack all the time, right bot check and then we put all the topics that normally get bots and they didn't get bots.
56:04.781 --> 56:06.542
[SPEAKER_03]: I thought, you know, I put my comment there
56:07.782 --> 56:20.270
[SPEAKER_01]: and uh... but you're right in a day later than and sent on my own account on my tic-tac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac-ac
56:25.353 --> 56:30.538
[SPEAKER_01]: Access trying to add more elements to try to eliminate body counts.
56:31.259 --> 56:37.645
[SPEAKER_01]: It has a new liveness check option that will provide another way of testing suspected counts.
56:39.467 --> 56:49.256
[SPEAKER_01]: This one, I think, is about user activity, motion checks, just to prove that
56:51.727 --> 56:57.491
[SPEAKER_01]: That means your is human that means the user has to like be available to be seen like turn the camera on and wave at or something.
56:58.111 --> 57:12.901
[SPEAKER_03]: Well, from social media today, when they detect activity that's just the account may be run by a bot, it sends a loudness check weird word to that profile, which requests that it usually there's been a selfie video or a face ID or touch ID to prove their human.
57:14.002 --> 57:18.485
[SPEAKER_03]: They also be able to do a hand motion check, which will crush, you have to wave it in for your hand.
57:18.802 --> 57:30.245
[SPEAKER_03]: But we covered a story last year from four or four media, the China had one deep take technology where you can raise the hand and it wouldn't matter, so I don't know if it'll affect nation state accounts that are there to do the thing.
57:30.726 --> 57:33.566
[SPEAKER_03]: But I'm only going to do, yeah.
57:34.427 --> 57:38.348
[SPEAKER_03]: It's also going to have an age checking process powered by GROC.
57:39.139 --> 57:40.279
[SPEAKER_03]: to involve age restrictions.
57:40.300 --> 57:43.041
[SPEAKER_03]: So I don't, I don't believe this is coming from us.
57:43.141 --> 57:44.001
[SPEAKER_03]: Probably someone internal.
57:44.021 --> 57:48.363
[SPEAKER_03]: I know people that had worked there and tried to cut down on the bot activity and they wouldn't let them.
57:48.383 --> 57:54.526
[SPEAKER_03]: This is a endorsy owned in, because of the engagement they needed for their funding and stuff.
57:54.586 --> 57:55.186
[SPEAKER_03]: But we'll see.
57:55.386 --> 57:56.847
[SPEAKER_03]: I mean, it is hurting the platform.
57:56.867 --> 58:02.209
[SPEAKER_03]: A lot of people have left the platform just because they're sick and tired of the fake accounts attack and ballot act.
58:02.249 --> 58:02.349
[SPEAKER_03]: So
58:02.970 --> 58:05.991
[SPEAKER_01]: for what it's worth of sorry it's on it's on the x-platform.
58:06.091 --> 58:15.714
[SPEAKER_01]: So I don't I don't believe it's in their interest to get rid of the bots because I don't think they care about making money.
58:16.114 --> 58:21.076
[SPEAKER_01]: I think they care about influencing the outcome of American politics and that's why they exist.
58:21.216 --> 58:23.757
[SPEAKER_01]: Not even American politics and that's why they exist.
58:24.077 --> 58:24.397
[SPEAKER_03]: I agree.
58:24.777 --> 58:24.957
[SPEAKER_03]: Yeah.
58:25.667 --> 58:30.355
[SPEAKER_01]: Um, and it would be really nice if American corporations stayed out of the rest of our politics.
58:30.436 --> 58:34.422
[SPEAKER_01]: Like you want to mess in your own politics and turn your own country into a stopia.
58:35.396 --> 58:38.157
[SPEAKER_01]: We advise against it, but go for it at your country.
58:38.677 --> 58:39.397
[SPEAKER_01]: Stay out of ours.
58:40.597 --> 58:41.537
[SPEAKER_01]: Said the rest of the world.
58:42.398 --> 58:43.538
[SPEAKER_01]: I'm serious.
58:43.618 --> 58:45.798
[SPEAKER_01]: I'm tired of American tech-to-crats.
58:46.379 --> 59:00.382
[SPEAKER_01]: I've had tech, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh
59:05.423 --> 59:13.119
[SPEAKER_01]: the banks that are supporting the AI industry, the investment houses are beginning to get nervous.
59:14.181 --> 59:15.984
[SPEAKER_01]: People want their money back.
59:20.588 --> 59:39.534
[SPEAKER_03]: But we talked about this a year ago and I was actually involved with a friend of ours Renee in startup culture here in Las Vegas so I learned a lot and if by year five you are not producing eight to nine times their investment in revenue that they can take back to get to really never something they just shipped on their investment.
59:40.142 --> 59:51.365
[SPEAKER_03]: So AI is different just because it's so heavily invested in and it's like 25% of our economy in the United States and all of these things, but they are starting to get nervous because we're going into your five.
01:00:05.502 --> 01:00:12.445
[SPEAKER_03]: So every year, the projections are it has to make a trillion more than last year to break even was $768 billion.
01:00:12.465 --> 01:00:14.846
[SPEAKER_03]: And then it was $1.6 trillion this year.
01:00:14.866 --> 01:00:18.188
[SPEAKER_03]: Like, one or two years from now, it's supposed to be $3 trillion to break even.
01:00:18.668 --> 01:00:23.990
[SPEAKER_03]: And then it's going to be a revenue stream that is going to produce them $3 trillion a year or pay these investors or money back.
01:00:24.271 --> 01:00:31.694
[SPEAKER_03]: Especially at 8 to 10, 8 to 9 times valuation, unless the IPOs go and IPOs go well, but we're seeing like with character,
01:00:34.503 --> 01:00:37.145
[SPEAKER_01]: It character, well, it's a character I in a few minutes.
01:00:37.185 --> 01:00:38.565
[SPEAKER_01]: That's a completely different fish.
01:00:39.446 --> 01:00:46.270
[SPEAKER_01]: Character I was something that was built off of other people's models or bastardize other people's models, other people's research.
01:00:46.330 --> 01:00:50.012
[SPEAKER_01]: But yeah, that was a very different thing.
01:00:50.513 --> 01:01:01.399
[SPEAKER_01]: But it is the first of the larger and well-known brands and AI to have to declare bankruptcy because the revenue model just is...
01:01:03.258 --> 01:01:04.019
[SPEAKER_01]: Not even not there.
01:01:04.039 --> 01:01:14.685
[SPEAKER_01]: I mean, it's so skewed and absurd and skewed against ever-making money that is hard to believe that they called it a revenue model to begin with.
01:01:15.006 --> 01:01:18.628
[SPEAKER_01]: That's my problem because I don't want to call it a revenue model because it's the wrong words.
01:01:20.349 --> 01:01:21.390
[SPEAKER_03]: That's what you're talking about.
01:01:21.470 --> 01:01:22.270
[SPEAKER_01]: Our tumbling now.
01:01:22.370 --> 01:01:24.251
[SPEAKER_01]: I mean, you want to really indicate her.
01:01:24.752 --> 01:01:27.893
[SPEAKER_01]: It's the shovels people will put into the ground to dig the gold.
01:01:28.133 --> 01:01:29.054
[SPEAKER_01]: That's your indicator.
01:01:29.114 --> 01:01:32.756
[SPEAKER_01]: Our people buying shovels, if they're buying shovels, they want to dig gold.
01:01:33.256 --> 01:01:38.239
[SPEAKER_01]: But if chipsawks are tumbling, then investors are thinking people don't want to buy shovels so much.
01:01:38.799 --> 01:01:42.902
[SPEAKER_03]: Well, this is Shamsi and investors.com with CNBCF to cover this in the story.
01:01:43.382 --> 01:01:47.745
[SPEAKER_03]: But interestingly enough, the title on some of Conductor and Manufacturing Company, TSM,
01:01:49.137 --> 01:02:12.225
[SPEAKER_03]: We recorded great profits, but it wasn't enough to keep these stocks from sliding this week because people are starting to ask, I don't know why investors have it asked us the entire time and because I'm sure they knew the problem, but they're not making revenue and there's no revenue there is no revenue stream available, they keep we'll talk about something in a minute like another product someone's putting out with AI.
01:02:12.919 --> 01:02:28.685
[SPEAKER_03]: But $20 a month from uni is not going to make $3 trillion a break even so they have to do military and surveillance technology, but even that has a whole risk like risk limit and also it's issues with those two, so where are they going to get $3 trillion in two years?
01:02:29.206 --> 01:02:29.646
[SPEAKER_01]: They're not.
01:02:30.366 --> 01:02:31.086
[SPEAKER_03]: Right, exactly.
01:02:31.146 --> 01:02:32.887
[SPEAKER_03]: And the investors are working in their money back.
01:02:33.027 --> 01:02:33.987
[SPEAKER_03]: They're not getting their money back.
01:02:34.047 --> 01:02:35.908
[SPEAKER_03]: And I don't know why they're just starting to realize it.
01:02:35.928 --> 01:02:40.150
[SPEAKER_03]: But at these year five, and this is usually their risk limit is five years of investment.
01:02:40.792 --> 01:02:50.959
[SPEAKER_01]: Well, okay, so character AI, which is a, I don't think it's here, is not open AI.
01:02:51.479 --> 01:03:05.108
[SPEAKER_01]: Character AI is not chatGBT, it's not complexity, it's not one of the major LLMs, it does use an LLM to create
01:03:10.532 --> 01:03:17.098
[SPEAKER_01]: It was one of the largest of the shell-like holiday girlfriend experience.
01:03:17.118 --> 01:03:17.778
[SPEAKER_01]: So pain, aye?
01:03:17.978 --> 01:03:19.079
[SPEAKER_03]: Well, let's just go for it.
01:03:19.119 --> 01:03:20.180
[SPEAKER_03]: Talk to the panion, everyone.
01:03:20.200 --> 01:03:20.781
[SPEAKER_01]: Talk to the panion.
01:03:20.801 --> 01:03:24.424
[SPEAKER_03]: Talk if I did other things, we said, like, let's type it up, so.
01:03:24.951 --> 01:03:45.441
[SPEAKER_01]: They even got involved with, say, Disney, Disney wanted to get involved with them and offered them $20 million to license some intellectual property and then had to offer a season to assist when people started using their corporate characters without permission.
01:03:53.837 --> 01:03:55.899
[SPEAKER_01]: It's just not making money.
01:03:56.819 --> 01:03:59.782
[SPEAKER_03]: This is why I'm Mark and Jason's company.
01:04:00.102 --> 01:04:01.763
[SPEAKER_03]: It's his company A16Z.
01:04:01.783 --> 01:04:05.246
[SPEAKER_03]: It put in 150 million and around in 2023.
01:04:05.787 --> 01:04:08.349
[SPEAKER_03]: As Jim said, there's lots of problems with this tech.
01:04:08.689 --> 01:04:11.191
[SPEAKER_03]: But it has the same basic though issue we just talked about.
01:04:11.900 --> 01:04:13.341
[SPEAKER_03]: It couldn't produce enough revenue.
01:04:13.421 --> 01:04:14.662
[SPEAKER_03]: It is not making any money.
01:04:15.222 --> 01:04:18.324
[SPEAKER_03]: And the difference is at this point, because it's causing five, it's cut.
01:04:18.745 --> 01:04:23.968
[SPEAKER_03]: It's, there's been a 56 million and compute and just six months or eight months alone this year.
01:04:24.889 --> 01:04:25.189
[SPEAKER_03]: Right?
01:04:25.269 --> 01:04:26.670
[SPEAKER_03]: So about nine million a month.
01:04:26.870 --> 01:04:27.891
[SPEAKER_03]: We're burning through cash.
01:04:27.951 --> 01:04:29.032
[SPEAKER_03]: It's just a small scale.
01:04:29.782 --> 01:04:33.126
[SPEAKER_03]: But the same problems that open the eyes having anthropic, if it doesn't go live, P.O.O.
01:04:33.146 --> 01:04:35.088
[SPEAKER_03]: isn't sure it'll stay alive.
01:04:35.309 --> 01:04:45.861
[SPEAKER_03]: It's because the cost and the revenue, and this one was companion AI, which I thought was really going to be the thing to take off, but it hasn't, and they've had a lot of issues with self-harm.
01:04:45.961 --> 01:04:48.845
[SPEAKER_03]: There have been suicides based on these characters, and there was a time when,
01:04:49.387 --> 01:05:00.945
[SPEAKER_03]: Disney characters are doing bad things and having sex and so I don't think it's going to miss that much but it is a representative on a smaller scale of the issues of the larger companies as well.
01:05:01.767 --> 01:05:07.051
[SPEAKER_01]: Well, for the longest time, I've been saying I love, I honestly, I love the revolution.
01:05:07.111 --> 01:05:15.198
[SPEAKER_01]: I think it's one of the most phenomenal times in intellectual human history, but the way we're using AI is completely and absolutely wrong.
01:05:16.579 --> 01:05:21.863
[SPEAKER_01]: Using it to think for us and set of using it to augment the ways we think.
01:05:22.764 --> 01:05:27.448
[SPEAKER_01]: Using it, as they vehicle to acquire tailor-made porn,
01:05:28.420 --> 01:05:34.422
[SPEAKER_01]: rather than using it to the vehicle to explore the human genome.
01:05:34.802 --> 01:05:36.403
[SPEAKER_03]: Well, here's my one question, everybody.
01:05:36.423 --> 01:05:37.483
[SPEAKER_03]: It's just a true historical.
01:05:37.503 --> 01:05:41.205
[SPEAKER_03]: We don't have to discuss it, but they did the math problem, right?
01:05:41.225 --> 01:05:47.247
[SPEAKER_03]: Because it was like 90% and they put 10,000 bots on it, and they supposedly solved it, otherwise we found out not that it didn't really solve it.
01:05:48.127 --> 01:05:49.848
[SPEAKER_03]: But I haven't seen them come up with a cancer cure.
01:05:50.685 --> 01:06:14.333
[SPEAKER_01]: or you have actually know you have you've seen masks there have been interesting you've seen massive increases in um here's from major diseases all of the chaplain because AI sped up not on the we have absolutely hundred percent absolutely and those are great those are amazing what are you using of AI and research and yes if if if if
01:06:15.293 --> 01:06:17.717
[SPEAKER_01]: You don't need to believe me or not, but it has happened.
01:06:18.398 --> 01:06:27.271
[SPEAKER_01]: But that wasn't done by the person spending $20 a month in their living room, or by the corporation trying to replace all of their accountants with AI.
01:06:27.331 --> 01:06:28.592
[SPEAKER_01]: That's not the same thing.
01:06:28.612 --> 01:06:30.796
[SPEAKER_03]: Anyway, if I could finish it all the same.
01:06:32.055 --> 01:06:44.662
[SPEAKER_03]: Um, so the cancer and those kinds of researchers have been some breakthroughs, but they spent 10,000 bots on, um, doing the, um, doing the math problem.
01:06:45.002 --> 01:06:45.343
[SPEAKER_02]: Mm-hmm.
01:06:45.423 --> 01:06:45.683
[SPEAKER_03]: Right.
01:06:45.843 --> 01:06:51.266
[SPEAKER_03]: So I was trying to say was they're not just spending that money on the things that it would be good for.
01:06:51.726 --> 01:06:52.947
[SPEAKER_03]: So things are not profitable.
01:06:54.014 --> 01:06:55.236
[SPEAKER_03]: and then near-term.
01:06:55.857 --> 01:06:58.039
[SPEAKER_03]: But there's nothing right now that's profitable either.
01:06:58.060 --> 01:07:02.165
[SPEAKER_01]: That's not only for highs, it's just what new would you ask about your mathematicians.
01:07:02.646 --> 01:07:05.169
[SPEAKER_01]: Unfortunately, being a mathematician ain't profitable anymore, either.
01:07:09.968 --> 01:07:24.137
[SPEAKER_01]: because of our misuse of the other of these elements now people won't get into that field and we won't see a lot more advancement but we'll be able to figure out exactly where we are now to the end of the degree.
01:07:25.858 --> 01:07:26.798
[SPEAKER_01]: That's why it's been crazy.
01:07:31.321 --> 01:07:32.442
[SPEAKER_01]: And even even though
01:07:34.798 --> 01:07:47.169
[SPEAKER_01]: Open AI and anthropic and other LLMs can come up with incredibly elegant, not necessarily tested, but incredibly elegant solutions to math problems.
01:07:48.070 --> 01:07:53.835
[SPEAKER_01]: 10, 15, 20 years from now will there be well-trained mathematicians around to understand them?
01:07:55.026 --> 01:08:00.691
[SPEAKER_03]: Yeah, it's very true because it doesn't have it, although, like we said, they haven't, they can't prove, they don't have a proof.
01:08:00.891 --> 01:08:05.575
[SPEAKER_03]: And if you have a proof in math at the level, then it's not considered a solution.
01:08:06.316 --> 01:08:07.436
[SPEAKER_03]: There's nobody can replicate it.
01:08:08.237 --> 01:08:21.068
[SPEAKER_03]: So we'll see if it has the impact that we think it could have, but right now, yes, it is going to display people from joining as mathematicians who love math who just solve math problems with the eyeballs can just come behind them and solve it, you know, instead.
01:08:22.238 --> 01:08:29.824
[SPEAKER_03]: So moving on, we don't have to go in depth to some of the people it's there.
01:08:30.245 --> 01:08:34.328
[SPEAKER_03]: Meta published a big counter data center concern post.
01:08:35.008 --> 01:08:42.414
[SPEAKER_03]: On the data centers, there's over 150 that have been hold back and canceled because of public outreach about the data centers.
01:08:42.935 --> 01:08:44.236
[SPEAKER_03]: So now there's a big push.
01:08:45.488 --> 01:08:59.659
[SPEAKER_03]: There's cards, there's ads, there's a super PAC, all to try to make everyone think that it's just fine and nothing that everyone's been told is true and it's just because China's telling us to do these things so meta is not joined the fray by adding this note.
01:09:00.099 --> 01:09:09.166
[SPEAKER_01]: Okay, just in case people are are interested meta says it's come up with ways it's coming up with ways to make data centers.
01:09:12.258 --> 01:09:17.283
[SPEAKER_01]: less impactful, there are ways to do it.
01:09:17.303 --> 01:09:20.806
[SPEAKER_01]: There are ways to make that a century smaller and less impactful.
01:09:21.787 --> 01:09:26.511
[SPEAKER_01]: It's probably quite expensive and again, at the rapid rate of build-out.
01:09:35.030 --> 01:09:40.935
[SPEAKER_01]: We're having a massive debate in Canada of it, whether should we should allow Dennis Anderson or not, but we're also having a massive debate about data sovereignty.
01:09:41.215 --> 01:09:46.139
[SPEAKER_01]: We want to keep our data here, instead of shipping it and storing it in the United States.
01:09:47.260 --> 01:09:57.648
[SPEAKER_01]: You can't have one without the other, but the impact of data centers is still tremendous on an e-local population.
01:09:59.143 --> 01:10:06.570
[SPEAKER_03]: Well, we do have data centers that can be very responsible, switch data centers in Las Vegas.
01:10:06.950 --> 01:10:10.654
[SPEAKER_03]: I guess, down right outside one and you won't hear a thing, because I don't do the diesel engines.
01:10:11.735 --> 01:10:12.976
[SPEAKER_03]: They have their own water supply.
01:10:13.036 --> 01:10:16.820
[SPEAKER_03]: They provide their own power, the only thing that uses our lines to transport it.
01:10:16.840 --> 01:10:18.041
[SPEAKER_03]: So there is responsible data center.
01:10:18.061 --> 01:10:20.003
[SPEAKER_03]: You can't be a perfect data center, but they're responsible.
01:10:20.944 --> 01:10:21.945
[SPEAKER_03]: The problems these are not.
01:10:22.025 --> 01:10:22.165
[SPEAKER_03]: And,
01:10:22.699 --> 01:10:32.969
[SPEAKER_03]: So people understand, everything we ran prior to AI on data centers, every single thing in the world, every single thing required to growth rate of about one to 2% a year of data centers.
01:10:33.009 --> 01:10:44.460
[SPEAKER_03]: Hyperscaling AI throwing all these, this one, I think it's in Ireland, or maybe it's UK, it's 7, 758 acres, and it's a Facebook data center.
01:10:45.421 --> 01:10:46.803
[SPEAKER_01]: It's a lot of acreage.
01:10:47.211 --> 01:10:49.933
[SPEAKER_03]: Yeah, so that these are strictly for the AI.
01:10:50.133 --> 01:10:54.055
[SPEAKER_03]: And so I think one thing countries could do real simply.
01:10:54.095 --> 01:10:57.077
[SPEAKER_03]: It's just put a cap on how many data centers are built a year.
01:10:57.097 --> 01:11:02.701
[SPEAKER_03]: As I said, one to two percent growth is what the standard was prior to the AI data centers.
01:11:04.622 --> 01:11:04.842
[SPEAKER_01]: OK.
01:11:06.959 --> 01:11:16.325
[SPEAKER_01]: I've got to move forward, but Meta's, you know, talking about Meta's got a PR PR campaign on, says it can make bigger, better, faster, cleaner.
01:11:16.725 --> 01:11:17.045
[SPEAKER_01]: We'll see.
01:11:17.585 --> 01:11:21.588
[SPEAKER_01]: It's also launching a next-stage agetic artificial intelligence.
01:11:22.308 --> 01:11:24.989
[SPEAKER_01]: I'll be able to call it that, or is it supposed to be so?
01:11:25.009 --> 01:11:25.589
[SPEAKER_03]: No, it's not a call.
01:11:26.130 --> 01:11:26.710
[SPEAKER_01]: I don't know.
01:11:26.750 --> 01:11:29.071
[SPEAKER_01]: This is a meta project, a meta did bend the knee.
01:11:29.731 --> 01:11:33.393
[SPEAKER_01]: I think this should be in a gigantic super intelligence push.
01:11:33.713 --> 01:11:34.293
[SPEAKER_01]: Muse wonder.
01:11:35.754 --> 01:11:46.698
[SPEAKER_01]: It's a bigger, better, faster, stronger, and it's part of the Muse wonder series with Muse wonder, and the next episode.
01:11:46.818 --> 01:11:47.579
[SPEAKER_01]: Wonderful.
01:11:48.315 --> 01:11:52.297
[SPEAKER_03]: Otherwise, it's not spelled wonder, a few else, it's wonder space.
01:11:52.917 --> 01:11:53.737
[SPEAKER_01]: Oh, yeah.
01:11:54.037 --> 01:11:55.678
[SPEAKER_01]: Okay, I'm done reading this story.
01:11:56.278 --> 01:12:09.903
[SPEAKER_03]: Well, I do think we should do in my people that this tech is super surveillance in a design and acute little toy that makes profile pages of all your friends and neighbors and family, and anybody who are connected to you and everything that they do connected to you.
01:12:10.043 --> 01:12:14.225
[SPEAKER_03]: And then when I encourage you to talk to it to make relationship decisions and things like that.
01:12:14.708 --> 01:12:14.948
[SPEAKER_01]: Indeed.
01:12:15.829 --> 01:12:19.211
[SPEAKER_01]: But before we go on, we've parted on that last week, and it's meta.
01:12:19.651 --> 01:12:20.772
[SPEAKER_01]: It's Facebook.
01:12:21.272 --> 01:12:24.594
[SPEAKER_01]: It's hovering all of your data and using it any way it sees fit.
01:12:27.136 --> 01:12:28.437
[SPEAKER_01]: And you're in your data.
01:12:29.757 --> 01:12:35.441
[SPEAKER_01]: If you're in the Facebook Instagram or what's up universe, part of your data is your friend's data.
01:12:35.521 --> 01:12:38.043
[SPEAKER_01]: You would agree to that when you would agree to be part of that universe.
01:12:38.083 --> 01:12:40.264
[SPEAKER_01]: You probably didn't know you would agree to it, but you did.
01:12:42.889 --> 01:12:54.617
[SPEAKER_01]: they are now allowed to use that information anyway they see fit which is going to be really uncool in a lot of the circumstances, but like me, you said cool, so there you go.
01:12:55.518 --> 01:13:00.341
[SPEAKER_01]: Not your question, but you, everybody, I mean, I've all of us dead, because we're using the environment.
01:13:00.962 --> 01:13:03.884
[SPEAKER_03]: Now, loaded immediately like them in the first week, I think it was.
01:13:04.598 --> 01:13:10.242
[SPEAKER_03]: But one of the promises of this AI Muse is that you'll be able to go shopping with it.
01:13:10.422 --> 01:13:20.488
[SPEAKER_03]: Well, we already know Amazon blocks are completely well now JP Morgan Chase is blocking all outside agents from accessing customer accounts because it's security issues.
01:13:20.508 --> 01:13:21.508
[SPEAKER_03]: Yeah.
01:13:21.528 --> 01:13:31.314
[SPEAKER_03]: So you'll never, you will not be able to use it with any any AI not open as financial agent or news agents and I'm sure they're not going to be the first banks to do it.
01:13:31.354 --> 01:13:32.355
[SPEAKER_03]: I'm sure there's something more to come.
01:13:33.184 --> 01:13:40.468
[SPEAKER_01]: You know, who else has banned a number of agentic AI agents from its system and doesn't want to let them in?
01:13:41.268 --> 01:13:45.831
[SPEAKER_01]: Amazon has, although Amazon did just launch it to new personal AI device.
01:13:46.291 --> 01:13:46.791
[SPEAKER_01]: Yeah.
01:13:46.831 --> 01:13:52.314
[SPEAKER_01]: Which of course we'll listen to your conversation, take notes, and figure out what to
01:13:53.715 --> 01:14:11.766
[SPEAKER_01]: recommend to you based on those conversations and notes, um, and there's also announced the bee eclipse, a credit card size device that magnetically attaches to the back of a phone, it goes for 1999, and um, it's one of their wearables going with the bee frontier, which is a
01:14:18.438 --> 01:14:18.738
[SPEAKER_01]: Here.
01:14:18.758 --> 01:14:20.819
[SPEAKER_03]: We're thinking, no, go ahead, sir.
01:14:21.080 --> 01:14:21.640
[SPEAKER_03]: Yeah.
01:14:21.700 --> 01:14:25.342
[SPEAKER_01]: It's kind of like, it's kind of like wearing a digital watch that does a lot more.
01:14:26.382 --> 01:14:28.704
[SPEAKER_03]: That's not like wearing a digital watch for digital watch.
01:14:29.244 --> 01:14:32.686
[SPEAKER_03]: It's not, it's not taking your conversations and turning them into emails.
01:14:33.186 --> 01:14:34.027
[SPEAKER_01]: Dick Tracy's dead.
01:14:35.207 --> 01:14:35.668
[SPEAKER_03]: This is true.
01:14:35.708 --> 01:14:37.048
[SPEAKER_03]: I still don't have to trace you.
01:14:37.068 --> 01:14:39.650
[SPEAKER_03]: No, I realized that was a bad move on my part.
01:14:40.210 --> 01:14:40.850
[SPEAKER_03]: Here's the thing.
01:14:41.030 --> 01:14:47.214
[SPEAKER_03]: It says that Amazon can't access your data and they're deleted after they are transcribed.
01:14:47.943 --> 01:14:53.086
[SPEAKER_03]: But tiny little device has a chip large enough to transcribe data.
01:14:54.087 --> 01:15:06.334
[SPEAKER_01]: Well, Amazon also can connect B to existing apps like Gmail, Google Counter Microsoft, that look and other Google Microsoft products that do record your data.
01:15:07.795 --> 01:15:13.859
[SPEAKER_01]: So Amazon's keeping your data then losing it, like, community losing it half an hour later when the session ends.
01:15:14.559 --> 01:15:15.780
[SPEAKER_01]: For some reason, I don't believe that.
01:15:16.628 --> 01:15:22.252
[SPEAKER_03]: I didn't either because when they're first AI, what is their little home device called Alexa?
01:15:22.673 --> 01:15:23.674
[SPEAKER_03]: And that's where it came out.
01:15:23.874 --> 01:15:24.955
[SPEAKER_03]: I kind of made the same claims.
01:15:25.375 --> 01:15:26.516
[SPEAKER_03]: And then research was found out.
01:15:26.916 --> 01:15:32.420
[SPEAKER_03]: Oh my gosh, it's calling back to Amazon like every two minutes, giving the information.
01:15:32.480 --> 01:15:36.944
[SPEAKER_01]: So, oh, we, that's Sean and I mess with our Alexa devices.
01:15:37.825 --> 01:15:40.527
[SPEAKER_01]: Regularly, just to see what happens when we do.
01:15:41.552 --> 01:15:56.217
[SPEAKER_03]: Well, it's funny because there have claimed things that aren't too like you went to their stores and it was all AI being read everything and you walk out and it automatically pays it for the people on cameras actually putting it in a checkout box and putting your totals and that's why I took a few minutes for it to charge.
01:15:56.758 --> 01:15:59.359
[SPEAKER_03]: So anyway, if you're going to use this device, I wouldn't this is.
01:16:11.892 --> 01:16:26.004
[SPEAKER_01]: Okay, folks who are using ChatGPT for advertising, for media has a pretty much a how-to in how ChatGPT targets ads to you based on prompts, phone number, email, et cetera.
01:16:26.744 --> 01:16:34.351
[SPEAKER_01]: So if you want to know how ChatGPT is serving up ads, what their cues are,
01:16:37.634 --> 01:16:47.803
[SPEAKER_01]: This is a good place to start, and then again, to show you how your customers might be getting your advertising right in front of them.
01:16:48.524 --> 01:16:51.726
[SPEAKER_03]: And also, I don't know that Google, I don't know if Facebook allows us any more.
01:16:51.746 --> 01:16:53.208
[SPEAKER_03]: I haven't done Facebook ads in over a year.
01:16:54.608 --> 01:16:59.290
[SPEAKER_03]: Facebook, when you first started, you could upload email justice and it would target those people specifically.
01:16:59.890 --> 01:17:01.511
[SPEAKER_03]: So it's doing the same thing here.
01:17:02.131 --> 01:17:05.713
[SPEAKER_03]: And I saw online people were pretty like a little bit outraged and I was like, oh, I get that.
01:17:06.433 --> 01:17:12.556
[SPEAKER_03]: But like a friend of our funny story that a friend of ours in the industry, he had a play.
01:17:12.576 --> 01:17:13.957
[SPEAKER_03]: It was going on vacation and didn't tell him.
01:17:15.057 --> 01:17:16.758
[SPEAKER_03]: So he created an ad yet of 25.
01:17:16.858 --> 01:17:19.359
[SPEAKER_03]: He just addresses it or say when he faked the others.
01:17:19.859 --> 01:17:22.220
[SPEAKER_03]: And then he put the guy's email and he kept running ads about,
01:17:25.082 --> 01:17:30.424
[SPEAKER_03]: You might, you should probably tell your boss if you're going on vacation, which is the guy came and told me I had no idea that had been done.
01:17:30.744 --> 01:17:36.425
[SPEAKER_03]: But this can be targeted based on email addresses, which a lot of people have used to on new stories.
01:17:36.926 --> 01:17:42.007
[SPEAKER_03]: They get the addresses of people that are in journalists and they make sure they see the story.
01:17:42.167 --> 01:17:45.228
[SPEAKER_03]: So, to just be aware that that's happening on Trotship BT.
01:17:46.849 --> 01:18:10.513
[SPEAKER_01]: Okay, opening I chat should be T ads account manager representatives are out there real opening eyes and playing people to reach out to advertisers and help the manager chat GPT ads are called chat GPT ads account managers and yeah do you remember when Google was out there doing this look Google ads rep would call you up and
01:18:16.987 --> 01:18:18.568
[SPEAKER_03]: Yeah, didn't they don't this other day?
01:18:18.588 --> 01:18:21.249
[SPEAKER_01]: They did, I don't know if he's still there, but I'm sure they did.
01:18:21.449 --> 01:18:22.389
[SPEAKER_03]: They did.
01:18:22.429 --> 01:18:22.730
[SPEAKER_03]: Yeah.
01:18:23.570 --> 01:18:24.050
[SPEAKER_03]: Yeah.
01:18:24.230 --> 01:18:29.272
[SPEAKER_03]: And yeah, the other thing is they're not usually there to help you that you should, they're to help Google make more money.
01:18:29.292 --> 01:18:31.593
[SPEAKER_03]: I said, I don't know how to use it.
01:18:31.633 --> 01:18:40.957
[SPEAKER_03]: It is interesting though, that since the initial buy-in to this program for ads was $100,000 a month and then $50,000 a month and then $25,000 a month that they didn't have an account manager when they started.
01:18:45.867 --> 01:18:51.410
[SPEAKER_03]: So let's say 100,000 a month that my ads aren't doing anything, you know, I'm not going to spend 100,000 next month.
01:18:52.591 --> 01:18:56.334
[SPEAKER_01]: So yeah, it's now too good time to get the account manager to call up, eh?
01:18:57.054 --> 01:18:57.774
[SPEAKER_03]: Yeah, exactly.
01:18:58.035 --> 01:18:58.575
[SPEAKER_03]: It's so funny.
01:18:59.545 --> 01:19:02.287
[SPEAKER_01]: So SEOs can't have anything nice.
01:19:02.327 --> 01:19:19.257
[SPEAKER_01]: We talked in our long interview with Ryan Jones and Dixon Jones about how Google is making it harder for analytics and reporting companies to scrape their data to report back to the industry.
01:19:19.277 --> 01:19:25.601
[SPEAKER_01]: To remember when the Google and said that they were going to develop a protocol,
01:19:27.962 --> 01:19:32.867
[SPEAKER_01]: a UTM track should be used so that you basically have an idea where the traffic was coming from.
01:19:32.887 --> 01:19:38.052
[SPEAKER_01]: You can attach it to a to to to to chat to B.T.
01:19:38.092 --> 01:19:43.398
[SPEAKER_01]: or Gemini or whatever and then Google said, look chat the opening eye and then the thru operator going along with us.
01:19:43.878 --> 01:19:45.460
[SPEAKER_01]: So we're all going to have UTM tracking.
01:19:46.864 --> 01:19:53.788
[SPEAKER_01]: They talk too soon because appears opening eye is not going along with them and have removed UPM.
01:19:54.309 --> 01:19:59.712
[SPEAKER_01]: UTM tracking parameters from Lates and ChatGPT, which means you ain't tracing them no more.
01:20:00.733 --> 01:20:08.318
[SPEAKER_03]: You know what, though, this to me is just kind of down on their part because if you don't know it came from ChatGPT, right?
01:20:08.838 --> 01:20:14.602
[SPEAKER_03]: It's already like one 360 of the topic of Google, but if you don't know it came from ChatGPT,
01:20:17.417 --> 01:20:18.498
[SPEAKER_03]: about chat to B.T.
01:20:19.460 --> 01:20:34.137
[SPEAKER_03]: I have no idea how many of this is the time we already know it's one, it's three percent generally speaking, and they have one day's worth of Google traffic, like how do you get your boss and say any more money for LLM optimization since chat to B.T.
01:20:34.177 --> 01:20:35.179
[SPEAKER_03]: is 90 percent of it.
01:20:36.899 --> 01:20:53.627
[SPEAKER_01]: Um, according to the, uh, study that's originally and ran, um, touch PT was, uh, accounting for about 5% of visits and, um, there was a demonstrable drop in visit they could attribute to chat GPT.
01:20:54.808 --> 01:21:03.292
[SPEAKER_01]: Um, but at the same time, there was a demonstrable uplift in, uh, basically unattributable traffic or direct traffic.
01:21:04.817 --> 01:21:06.600
[SPEAKER_01]: Um, so touch your B.T.
01:21:06.720 --> 01:21:14.471
[SPEAKER_01]: and the people are adopting you to drop it and it's an important environment, but it's impossible to How to say that.
01:21:14.611 --> 01:21:18.697
[SPEAKER_01]: So with confidence because you can't measure with confidence.
01:21:18.757 --> 01:21:18.997
[SPEAKER_01]: I mean,
01:21:19.878 --> 01:21:36.946
[SPEAKER_01]: I can tell you, Christina, I can tell you I'm a really good writer, I have the I have the entity spread covered, I'm pretty sure that what I'm writing for you is going to be relational with like a number of other authoritative pages out there and with, you know, stuff that we've mentioned in other other things that we've put on the same site.
01:21:38.487 --> 01:21:45.430
[SPEAKER_01]: And I can tell you all that I can, I can probably see the moon in the stars, but the honest truth is Christina, I cannot provide you accurate measurements for how it's gone.
01:21:47.290 --> 01:21:48.391
[SPEAKER_03]: No, no, I can't.
01:21:48.711 --> 01:21:50.912
[SPEAKER_01]: And I think that's because I'm selling you a product.
01:21:52.333 --> 01:21:57.815
[SPEAKER_03]: Yeah, and if you want people to care about opening it, I'd have to go to prove that people are using it.
01:21:58.936 --> 01:22:05.799
[SPEAKER_03]: And since we know originally it had people directly using it, which is why it's a little quick to.
01:22:06.520 --> 01:22:11.662
[SPEAKER_03]: But now also people, it's only part of the journey now, more and more as a study at Multi-out Tribution.
01:22:12.503 --> 01:22:12.723
[SPEAKER_03]: And so
01:22:13.950 --> 01:22:15.972
[SPEAKER_03]: You know, it's a good visibility to hold on and wrong.
01:22:15.992 --> 01:22:26.322
[SPEAKER_03]: You want to be visible in the air search, but to get funding for budget for doing the work to do the L.M.
01:22:26.382 --> 01:22:31.687
[SPEAKER_03]: presence if you can't track it, much, much harder to get people to care about it in an enterprise, I hope.
01:22:33.241 --> 01:22:44.395
[SPEAKER_01]: Yeah, it's, I honestly don't know why they're taking the UTM's out, they're taking it making it really hard to track.
01:22:46.798 --> 01:22:58.993
[SPEAKER_01]: and Silverary Evidence says that more and more and more uses are adopting information out of an AI overview or some LLM generative experience.
01:22:59.594 --> 01:23:01.016
[SPEAKER_01]: But again, something to prove it.
01:23:01.096 --> 01:23:02.878
[SPEAKER_01]: It's nothing to have to tobacco it up with.
01:23:03.038 --> 01:23:03.319
[SPEAKER_01]: Yeah.
01:23:04.783 --> 01:23:09.587
[SPEAKER_01]: The engines can say this to us all they want, but they're not giving us the receipts.
01:23:11.308 --> 01:23:20.275
[SPEAKER_01]: And then we have to give those receipts to our clients, who really want really, really, really wants something definite to spend money on.
01:23:20.295 --> 01:23:23.278
[SPEAKER_01]: And they can accept that, this is the best attempt service.
01:23:23.658 --> 01:23:27.481
[SPEAKER_01]: They've been doing advertising for decades in many cases.
01:23:28.142 --> 01:23:32.946
[SPEAKER_01]: But like, to not even have something as flaky as Dilsen ratings.
01:23:33.718 --> 01:23:36.302
[SPEAKER_01]: I mean, seriously, this is a digital age, eh?
01:23:37.384 --> 01:23:39.286
[SPEAKER_03]: Yeah, I'll tell you, Gree.
01:23:39.507 --> 01:23:42.171
[SPEAKER_03]: But let's get the next one to make that our last one, if you want.
01:23:42.571 --> 01:23:44.334
[SPEAKER_01]: Oh, sure, yeah, I got, yeah, indeed.
01:23:44.354 --> 01:23:47.118
[SPEAKER_01]: We got so much, so much to report, but there was such a great interview, okay?
01:23:47.138 --> 01:23:47.298
[SPEAKER_01]: So,
01:23:48.270 --> 01:24:12.337
[SPEAKER_01]: at the beginning of the uh... do you remember what everybody was about to get fired year of a to lose your job and we're all gonna lose your job just trade the a i first that you're fired well that was an age when uh... compute was limitless and uh... apparently i was going to be able to do anything one of the worst companies for this was the company extension they
01:24:14.373 --> 01:24:35.598
[SPEAKER_01]: Um, replace so many people in their organization with AI, and then they told the rest to use AI's to, um, make up for the, for the, for the missing employees, and some of their employees were spending as much as 1000 bucks a month on cloud, um, there was a threat or more, um, to, again, make up for, for, for, for lost colleagues.
01:24:36.338 --> 01:24:41.659
[SPEAKER_01]: Well, now, Accenture has lowered the boom on this, so telling employees they can now only spend
01:24:42.767 --> 01:24:47.291
[SPEAKER_01]: from 1,000 a month, 29 a month, on Claude.
01:24:48.632 --> 01:24:54.677
[SPEAKER_03]: Yeah, there were spending 1,000 and down to $29, I mean, I mean, how do you even... What does that pet brought about?
01:24:55.778 --> 01:25:00.241
[SPEAKER_03]: Yeah, I like great is, how much do you do for $29 a month?
01:25:01.042 --> 01:25:01.983
[SPEAKER_03]: The cow limited, are you?
01:25:02.043 --> 01:25:07.007
[SPEAKER_03]: I mean, forcing employees to use AI was not smart move in the first place.
01:25:07.391 --> 01:25:15.274
[SPEAKER_03]: Well, she forced them to use it and rely on it and then say, by the way, you get $29 a month, what is that like, half a day's use just a week?
01:25:15.674 --> 01:25:31.160
[SPEAKER_01]: I haven't had a date, there are 250 a week, isn't, I mean, seriously, with the load that Claude is carrying for a lot of developers, 250 a week is not like, a lot of money, a lot of money is not a reasonable at all.
01:25:37.257 --> 01:25:38.518
[SPEAKER_03]: And they're not the only ones.
01:25:38.698 --> 01:25:59.867
[SPEAKER_03]: This is a lowest effort, but a lot of companies have cut out a lot of the usage internally, like we did one story from four or four, if you remember, where the predominant use of AI was for power points in email and not for development or engineers, so they cut it out from all those people in the company, because they're spending hundreds of thousands of dollars on.
01:26:00.854 --> 01:26:01.975
[SPEAKER_03]: You know, do you own PowerPoint?
01:26:02.135 --> 01:26:05.336
[SPEAKER_03]: We work with technology so expensive.
01:26:05.356 --> 01:26:06.577
[SPEAKER_03]: We haven't even seen the cost yet.
01:26:07.037 --> 01:26:08.198
[SPEAKER_01]: It's a work for Accenture.
01:26:08.458 --> 01:26:09.678
[SPEAKER_01]: You're doing a lot of PowerPoints.
01:26:10.379 --> 01:26:17.262
[SPEAKER_03]: Yeah, and it's often an open AI keep talking about raising their prices and they start to and then they pull back because it gets us where so much blowback.
01:26:17.783 --> 01:26:23.726
[SPEAKER_03]: But at some point in the near future, they're going to have to just to cover the burn and they can't even cover the burn with that.
01:26:23.766 --> 01:26:26.147
[SPEAKER_03]: But right now, they're just taking money, throwing it in a
01:26:32.027 --> 01:26:36.209
[SPEAKER_01]: I get the sense that you have some time constraints happening in your world.
01:26:36.469 --> 01:26:38.250
[SPEAKER_01]: This is a long, long show.
01:26:38.330 --> 01:26:41.471
[SPEAKER_01]: It's a whole bunch of stuff that we still have on the grid.
01:26:43.472 --> 01:26:46.473
[SPEAKER_01]: We got AI generated music being labeled at YouTube.
01:26:48.134 --> 01:26:51.255
[SPEAKER_01]: AI search chat pot also being rolled out at YouTube.
01:26:52.456 --> 01:26:54.817
[SPEAKER_01]: Plain language algorithm controls at Facebook.
01:26:55.999 --> 01:26:58.609
[SPEAKER_01]: what app-viding guest links for non-users?
01:26:59.051 --> 01:27:00.617
[SPEAKER_01]: That's actually kind of cool.
01:27:03.293 --> 01:27:05.975
[SPEAKER_01]: We've had a story of a Pinterest in the grid here.
01:27:06.215 --> 01:27:07.095
[SPEAKER_03]: Wow!
01:27:07.336 --> 01:27:08.416
[SPEAKER_03]: I like this one.
01:27:08.756 --> 01:27:10.518
[SPEAKER_03]: The other ones we probably don't need to go on that one.
01:27:10.618 --> 01:27:12.099
[SPEAKER_03]: They're pretty self-explanatory.
01:27:12.139 --> 01:27:14.080
[SPEAKER_03]: But this is an interesting one.
01:27:14.280 --> 01:27:17.642
[SPEAKER_03]: Pinterest is asking brands to send people back to their lives.
01:27:17.882 --> 01:27:19.183
[SPEAKER_03]: This is a social media today.
01:27:19.663 --> 01:27:26.307
[SPEAKER_03]: So the companies released a marketing playbook outlining how to inspire offline activity instead of encouraging infinite scrolling.
01:27:26.427 --> 01:27:30.370
[SPEAKER_03]: So Pinterest is continuing its anti-social media promotional push.
01:27:30.770 --> 01:27:37.453
[SPEAKER_03]: with a new report that highlights how younger users are increasingly want to use online connections to fuel offline experiences.
01:27:37.934 --> 01:27:41.135
[SPEAKER_03]: The playbook also explores how Pinterest plants play a role in that.
01:27:41.896 --> 01:27:45.077
[SPEAKER_03]: According to them, 40% of generations Z users said they would.
01:27:45.617 --> 01:27:47.398
[SPEAKER_03]: It was social media and never been invented.
01:27:47.858 --> 01:27:51.760
[SPEAKER_03]: And 63% of Gen Z adults want to unplug a more intentional way.
01:27:52.361 --> 01:27:54.442
[SPEAKER_03]: So they're going to send
01:27:55.809 --> 01:28:01.676
[SPEAKER_03]: help devise the platform to get people to get offline, which is very interesting because what they're saying is very true.
01:28:01.996 --> 01:28:10.687
[SPEAKER_03]: We know that Gen Z has been rejecting AI tech, they've been rejecting apps, they've had the least amount of the least likely they're adding up to their phone.
01:28:11.968 --> 01:28:14.872
[SPEAKER_03]: So they're kind of an analogous cool on Gen Z kids.
01:28:15.702 --> 01:28:28.789
[SPEAKER_03]: So solid Pinterest is actually doing the, one of the few we've heard of that's actually tapping into that trend and figuring out how they can use that trend for their brand, which is probably gonna serve them very well going into the future.
01:28:29.910 --> 01:28:31.531
[SPEAKER_01]: Uh, that's hilarious.
01:28:31.611 --> 01:28:36.253
[SPEAKER_01]: I'm curious what Pinterest Market share is in relation to some of their competitors.
01:28:36.833 --> 01:28:41.316
[SPEAKER_01]: And if they're really just trying to even the grow dope by getting people to stop using their competitors,
01:28:42.417 --> 01:29:05.141
[SPEAKER_03]: Well, they're seeing searches like down phone as increased to 150%, and a log as static has increased to 160%, thanks to doing instead of scrolling, 365% of tick and searches for offline inspiration, travel sticker collection ideas, and the dazzle headphones, 1253% increase in the search.
01:29:05.581 --> 01:29:10.262
[SPEAKER_01]: You know, the number one selling phone that's not an Apple brand is an OQFlet phone.
01:29:10.995 --> 01:29:11.275
[SPEAKER_03]: Yeah.
01:29:11.755 --> 01:29:13.576
[SPEAKER_03]: Which by the way, Google also uses for testing.
01:29:14.776 --> 01:29:15.756
[SPEAKER_03]: You tell me you're lazy up.
01:29:16.276 --> 01:29:16.817
[SPEAKER_03]: Yes.
01:29:16.857 --> 01:29:18.037
[SPEAKER_03]: I mean, I like this idea.
01:29:18.157 --> 01:29:23.198
[SPEAKER_03]: And it says, Pinterest said, let's lastly set them free.
01:29:23.218 --> 01:29:27.180
[SPEAKER_03]: This is where advertisers need to be brave once people have found what they're looking for, let them go.
01:29:27.200 --> 01:29:30.200
[SPEAKER_03]: Don't trap people in the experience, let's do brand-new the job.
01:29:30.561 --> 01:29:33.241
[SPEAKER_03]: And then send them back into a life for having found what they're looking for.
01:29:33.681 --> 01:29:35.642
[SPEAKER_03]: Oh my gosh, smart marketing.
01:29:36.834 --> 01:29:46.101
[SPEAKER_03]: right, tapping in, giving them what they want and telling people and their brand users to like not try to trap them online because they don't want to be trapped online anymore.
01:29:47.309 --> 01:29:48.610
[SPEAKER_03]: I think it's super smart.
01:29:48.910 --> 01:29:49.831
[SPEAKER_03]: We'll see if you can do it.
01:29:49.851 --> 01:29:58.196
[SPEAKER_01]: You know what in the weirdest way I can relate to that, it's the 15th of October and it's still relatively warm outside before winter sets in.
01:29:58.876 --> 01:30:00.957
[SPEAKER_01]: I can understand it, but not where you'd be trapped online.
01:30:02.078 --> 01:30:05.980
[SPEAKER_01]: And you know, I get the feeling they get that, that you need to get out of the studio pretty quickly.
01:30:06.060 --> 01:30:08.222
[SPEAKER_01]: So we're going to have to wrap up here pretty fast.
01:30:08.422 --> 01:30:10.183
[SPEAKER_03]: I don't know if you get out on time.
01:30:11.964 --> 01:30:13.865
[SPEAKER_01]: You just suggest, okay.
01:30:14.986 --> 01:30:15.246
[SPEAKER_03]: I did.
01:30:15.886 --> 01:30:20.551
[SPEAKER_03]: Yeah, you did, you just said that we were, um, I said we skipped the one story and put it at the end.
01:30:28.360 --> 01:30:37.490
[SPEAKER_03]: Anyway, so the next story is about, um, uh, we got threats is going to start allowing people to see where their kids blocked people.
01:30:38.892 --> 01:30:42.601
[SPEAKER_03]: in their user activity to help reduce harm for kids.
01:30:43.022 --> 01:30:43.382
[SPEAKER_01]: Indeed.
01:30:43.563 --> 01:30:45.147
[SPEAKER_01]: And we drove all the Google stories.
01:30:45.528 --> 01:30:47.472
[SPEAKER_03]: And I Google is the next one.
01:30:48.849 --> 01:30:52.431
[SPEAKER_03]: and Google discovered a transform in the hyper personalized concierge.
01:30:52.732 --> 01:30:58.535
[SPEAKER_01]: That's actually something we should talk about a lot because Google's trying to move away from the 10 blue links.
01:30:58.615 --> 01:31:00.957
[SPEAKER_01]: It's almost certainly going to have.
01:31:01.097 --> 01:31:03.879
[SPEAKER_01]: Uh, remember how Google is asking you to recommend pages.
01:31:03.999 --> 01:31:05.020
[SPEAKER_01]: Be my favorite page.
01:31:05.120 --> 01:31:06.080
[SPEAKER_01]: Be my recommended page.
01:31:06.100 --> 01:31:12.865
[SPEAKER_01]: That's our stuff so that I could know what to tell recommend for people when they first share the cell phones on and discover.
01:31:14.431 --> 01:31:23.234
[SPEAKER_01]: I think this is a yet another sign that they want to move away from people actually conducting searches.
01:31:23.575 --> 01:31:33.018
[SPEAKER_01]: They just want to be the answer and stand in between you and life experience, engine.
01:31:34.839 --> 01:31:42.302
[SPEAKER_03]: Is it Google as a new job post, and you can, uh, principal UX search to discover, holy, the efforts to transform Google to discover into a proactive hyper-personalized
01:31:43.973 --> 01:31:47.315
[SPEAKER_03]: I don't even use Google the cover, so I don't know if it'll work or not, but.
01:31:48.055 --> 01:31:51.237
[SPEAKER_01]: What do you turn yourself on on when you you're very first, uh, substrate.
01:31:52.157 --> 01:31:55.959
[SPEAKER_01]: Don't you don't you have a number of, uh, uh, things recommended by Google?
01:31:56.620 --> 01:31:59.501
[SPEAKER_03]: No, uh, uh, uh, your Google browser.
01:31:59.781 --> 01:32:00.882
[SPEAKER_01]: Yeah.
01:32:00.902 --> 01:32:02.963
[SPEAKER_01]: Yeah, I use the Chrome browser on an Android device.
01:32:03.363 --> 01:32:06.225
[SPEAKER_01]: Um, discovery is a very big part of that experience.
01:32:06.265 --> 01:32:06.685
[SPEAKER_01]: You've got to
01:32:11.352 --> 01:32:23.245
[SPEAKER_01]: Okay, Google's got information out there on how they rank PDFs, what that means for SEO, and also for generative engineering optimization.
01:32:23.265 --> 01:32:30.654
[SPEAKER_03]: Yeah, so listen to the part of this, everyone will want to go to Search Central,
01:32:34.318 --> 01:32:45.747
[SPEAKER_03]: to go look it to listen to this, but basically PDFs are binary code and so Google has to turn it into HTML and then they explain how that works, how links work, they do work, but how they work.
01:32:46.208 --> 01:32:58.658
[SPEAKER_03]: And basically once it turns into HTML, the PDFs are read like any other page, but there's a bunch of information here about PDF ranking with lots of quotes from Roger Monty on search engine journal, so it's a little bit too much fresh to get into here.
01:33:00.588 --> 01:33:11.738
[SPEAKER_01]: Um, by the way, this almost goes with the discover story, um, but it's not about the same different thing, but I got a feeling they're actually kind of about the same thing.
01:33:12.298 --> 01:33:16.362
[SPEAKER_01]: Google testing AI mode in left sidebar on main search results.
01:33:16.402 --> 01:33:23.668
[SPEAKER_01]: So, if you did a search of it only normally produce the 10 boot links, wouldn't produce a number of in solar re-google, um, search
01:33:26.604 --> 01:33:28.305
[SPEAKER_01]: Well, you're not going to just heat up the links anymore.
01:33:28.345 --> 01:33:31.988
[SPEAKER_01]: Google's throwing an AI view off on the left side of the results.
01:33:32.289 --> 01:33:35.151
[SPEAKER_01]: So you'll have both sort of acting in parallel with each other.
01:33:36.352 --> 01:33:39.995
[SPEAKER_03]: Yeah, so it's got a big menu on the left hand side.
01:33:40.195 --> 01:33:42.877
[SPEAKER_03]: Personally, I don't want anything to take any more space where it's in my screen.
01:33:43.638 --> 01:33:45.740
[SPEAKER_03]: Also, I can turn it off, but some people might find it useful.
01:33:46.632 --> 01:33:50.174
[SPEAKER_01]: I, again, I think this is something SEO should be watching really closely.
01:33:50.234 --> 01:33:58.938
[SPEAKER_01]: The more Google introduces AI results into results set, set traditionally didn't have them, or performatively wouldn't produce them.
01:33:59.479 --> 01:34:02.700
[SPEAKER_01]: It's a signal from Google that we're going to see a lot more AI generated,
01:34:08.976 --> 01:34:20.061
[SPEAKER_01]: I'm still going with my theory Christine that Google's long-term intention is to stop being a search engine and stop and to start being a lifestyle in a fulfillment engine if you will.
01:34:20.081 --> 01:34:21.002
[SPEAKER_03]: I think it's going to be both.
01:34:21.122 --> 01:34:22.443
[SPEAKER_03]: I'm not there to get anybody to search.
01:34:23.682 --> 01:34:35.648
[SPEAKER_01]: Well, I think people searches will be unfulfilled in that they'll just get you to where they think you want to go and that'll be good enough for the vast majority and they're going to get a little slice off of everything.
01:34:36.849 --> 01:34:40.531
[SPEAKER_01]: And they're going to kill a golden goose along the way.
01:34:43.152 --> 01:34:46.314
[SPEAKER_03]: Yeah, um, maybe maybe not by the way, that's my speculation.
01:34:46.734 --> 01:34:48.375
[SPEAKER_03]: Well, my speculation is they're trying.
01:34:48.917 --> 01:34:56.959
[SPEAKER_03]: But Google search results have gotten worse and worse, and people will move to Bing, not LLM, when they've gotten, you know, one to delete.
01:34:58.100 --> 01:35:00.160
[SPEAKER_03]: And so, I don't know, I agree.
01:35:00.260 --> 01:35:02.181
[SPEAKER_03]: I think it's, I just think it's a bad move.
01:35:02.241 --> 01:35:04.722
[SPEAKER_03]: I think they've been trying to do this since we started in search.
01:35:04.742 --> 01:35:09.483
[SPEAKER_03]: When Eric Schmidt said, we want to answer questions before you even know what you need to answer.
01:35:09.503 --> 01:35:10.803
[SPEAKER_03]: He said that somewhere on 2008 or 2009.
01:35:13.728 --> 01:35:15.829
[SPEAKER_03]: And I'm not buying helpful.
01:35:16.269 --> 01:35:19.351
[SPEAKER_03]: I want to just get a result as I need a piece of information.
01:35:19.391 --> 01:35:21.112
[SPEAKER_03]: I'm not finding all this stuff helpful.
01:35:21.132 --> 01:35:22.953
[SPEAKER_03]: I'm finding that it's a big pain in my butt.
01:35:23.273 --> 01:35:24.654
[SPEAKER_03]: So we'll see how it goes.
01:35:26.355 --> 01:35:30.977
[SPEAKER_03]: Google puts also, but it's just getting worse and worse to the eye over visual getting worse, too.
01:35:30.997 --> 01:35:34.759
[SPEAKER_03]: Google puts inline links higher and more AI search responses.
01:35:34.859 --> 01:35:35.700
[SPEAKER_03]: That's good to see.
01:35:36.360 --> 01:35:37.281
[SPEAKER_03]: Also, net is Google.
01:35:37.421 --> 01:35:39.402
[SPEAKER_03]: And Gemini, if I am using it,
01:35:40.382 --> 01:35:46.546
[SPEAKER_03]: And AMO'd will stop sending me information like a summary or an answer and just give me links.
01:35:47.947 --> 01:35:49.508
[SPEAKER_03]: I go to speak like we're just going to do links here.
01:35:49.528 --> 01:35:50.869
[SPEAKER_03]: I'm not going to answer anymore.
01:35:50.889 --> 01:35:53.970
[SPEAKER_03]: I have to push it to get you more answers after you've been involved with it for a little bit.
01:35:54.291 --> 01:35:55.932
[SPEAKER_03]: Which is a good way to spend less money.
01:35:57.993 --> 01:35:59.854
[SPEAKER_01]: Okay.
01:36:00.134 --> 01:36:06.198
[SPEAKER_01]: Google is going to be showing pharmaceutical and drug ads in AI mode and AI overviews.
01:36:08.734 --> 01:36:22.605
[SPEAKER_01]: Google has a list of dangerous products that it tries not to cross into traditional or now AI search results or allow advertising paid add and traditional or now in AI search results.
01:36:23.386 --> 01:36:30.872
[SPEAKER_01]: Pharmaceuticals and drugs have been on and off that dangerous list throughout the years.
01:36:32.053 --> 01:36:33.595
[SPEAKER_01]: It's apparently off of this now.
01:36:34.808 --> 01:36:40.229
[SPEAKER_03]: Well, I'd stand if you're like Google's getting desperate for cash because they're doing guns now.
01:36:40.509 --> 01:36:41.609
[SPEAKER_03]: There's a test for guns.
01:36:42.230 --> 01:36:44.190
[SPEAKER_03]: If it lacks talent, they do adult.
01:36:45.050 --> 01:36:47.251
[SPEAKER_03]: Now they're relaxing from particles.
01:36:48.691 --> 01:36:51.091
[SPEAKER_03]: So it's like they need money.
01:36:51.111 --> 01:36:59.493
[SPEAKER_03]: Now how much I need money, I don't know, but they're obviously burning through money too because every church now costs a thousand dollars for an age to under some research.
01:36:59.513 --> 01:37:04.014
[SPEAKER_03]: So yeah, loosening all this up means access to big money.
01:37:04.932 --> 01:37:15.878
[SPEAKER_03]: these are things that these are topics that generate a lot of money, but, um, and I will say it just feels like a whole degradation of Google to me as a one-time Google user and SEO.
01:37:16.158 --> 01:37:16.379
[SPEAKER_03]: So.
01:37:17.879 --> 01:37:18.220
[SPEAKER_01]: Okay.
01:37:18.260 --> 01:37:21.181
[SPEAKER_01]: Well, you will add to maybe interest in more money.
01:37:21.241 --> 01:37:23.503
[SPEAKER_01]: It's updated as sexual content policy.
01:37:23.563 --> 01:37:25.043
[SPEAKER_01]: It's given more examples.
01:37:25.444 --> 01:37:27.965
[SPEAKER_01]: And surrounded those examples with paid at with
01:37:31.896 --> 01:37:37.878
[SPEAKER_01]: Google Ads is updated in the sexual contact policies to provide additional examples of clear information to advertisers.
01:37:40.159 --> 01:37:45.200
[SPEAKER_01]: The main section that was updated is called the sexually suggestive element and theme section.
01:37:46.461 --> 01:37:47.741
[SPEAKER_01]: Check out SEO Roundtable.
01:37:47.761 --> 01:37:53.123
[SPEAKER_01]: It gives a breakdown of all the changes to it.
01:37:55.083 --> 01:37:58.985
[SPEAKER_03]: With topics such as sexualized and animal objects,
01:38:00.381 --> 01:38:07.984
[SPEAKER_03]: being all present assess, body movements poses in expressions, non-graphic sex and in imagery, non-graphic sex and language.
01:38:08.004 --> 01:38:11.645
[SPEAKER_03]: So all that is now detailed with examples and the new documentation.
01:38:12.146 --> 01:38:16.907
[SPEAKER_01]: Do they have a subsection specifically relating to Sydney, Sweeney, advertisements?
01:38:17.968 --> 01:38:18.948
[SPEAKER_03]: I don't know who that is.
01:38:18.968 --> 01:38:19.648
[SPEAKER_01]: I don't see that.
01:38:22.640 --> 01:38:29.543
[SPEAKER_01]: Um, okay, Google poor image quality won't stop products from showing in search or at or in ads.
01:38:30.363 --> 01:38:34.524
[SPEAKER_01]: Uh, apparently you can upload potato picks and Google will still show them.
01:38:35.385 --> 01:38:40.566
[SPEAKER_01]: But I bet you it will absolutely impact the number of clicks you actually get.
01:38:40.586 --> 01:38:41.887
[SPEAKER_03]: Definitely.
01:38:42.107 --> 01:38:42.487
[SPEAKER_03]: Yeah.
01:38:42.647 --> 01:38:46.569
[SPEAKER_03]: But I think was there a time then they weren't likely to show a low quality image.
01:38:46.589 --> 01:38:47.609
[SPEAKER_03]: I just didn't remember that.
01:38:48.596 --> 01:38:50.277
[SPEAKER_03]: But they're telling you now, it's not a problem.
01:38:50.758 --> 01:38:51.739
[SPEAKER_03]: Oh, your phone is.
01:38:52.019 --> 01:38:54.781
[SPEAKER_01]: I've got one of my phone's was just confirming what I said.
01:38:54.801 --> 01:39:02.368
[SPEAKER_01]: It will affect the number and quality of clicks you get, but it will not affect the number of impressions you get.
01:39:02.388 --> 01:39:09.434
[SPEAKER_03]: This is if people don't know Jimson and it's the chat at least one time or so, even though he usually has it under a pillow.
01:39:10.014 --> 01:39:15.739
[SPEAKER_01]: No, this time I've got it on the sofa, on the other side of my, on the other, at the, at the, at the, the other wall of my office.
01:39:15.839 --> 01:39:16.420
[SPEAKER_01]: I mean, like.
01:39:18.973 --> 01:39:28.740
[SPEAKER_01]: You just, anyway, our friendly Gemini was very interested in that answer probably just cost me another dollar.
01:39:30.822 --> 01:39:30.982
[SPEAKER_03]: Yep.
01:39:31.062 --> 01:39:31.682
[SPEAKER_03]: And then, too.
01:39:33.123 --> 01:39:34.464
[SPEAKER_01]: And that cost them $5.
01:39:34.504 --> 01:39:36.646
[SPEAKER_01]: That's the problem with the entire economy, isn't it?
01:39:37.418 --> 01:39:38.779
[SPEAKER_03]: I also, this one's a nice one.
01:39:38.819 --> 01:39:44.542
[SPEAKER_03]: Google Local has added dietary attributes, gluten-free dairy-free, hollow and kosher.
01:39:44.962 --> 01:39:54.547
[SPEAKER_03]: I am gluten-free, and often I have to pull up that PDF menu they have on here, website and try to go through the list and no one clearly marks things in most cases.
01:39:54.628 --> 01:39:56.048
[SPEAKER_03]: So it'd be really nice.
01:39:56.108 --> 01:39:59.330
[SPEAKER_03]: It's just a new attribute you can add to your listing.
01:39:59.430 --> 01:40:02.192
[SPEAKER_03]: It's a special area when you're in the back end.
01:40:03.439 --> 01:40:07.401
[SPEAKER_01]: I live in a city that that builds itself as the most multicultural city on earth.
01:40:07.521 --> 01:40:14.323
[SPEAKER_01]: So knowing where hello and kosher food is is really, really important because chances are in any group that you're with.
01:40:14.683 --> 01:40:15.804
[SPEAKER_01]: It's important to somebody.
01:40:16.644 --> 01:40:17.524
[SPEAKER_01]: Yeah, that's great.
01:40:17.584 --> 01:40:18.205
[SPEAKER_01]: This is wonderful.
01:40:18.805 --> 01:40:19.045
[SPEAKER_03]: Yeah.
01:40:20.025 --> 01:40:20.606
[SPEAKER_01]: We're happy.
01:40:20.866 --> 01:40:23.367
[SPEAKER_01]: Really quickly where we're off the close to the end here.
01:40:26.569 --> 01:40:31.032
[SPEAKER_01]: Google Search Console Performance Report is insanely extremely delayed.
01:40:31.052 --> 01:40:36.355
[SPEAKER_01]: We're looking at like three days at this point, 64 hours and counting.
01:40:37.216 --> 01:40:41.559
[SPEAKER_01]: We're probably in the close to 70 hours at that time of reading this report.
01:40:42.460 --> 01:40:45.562
[SPEAKER_01]: And, um, I'm not taking it yet.
01:40:46.262 --> 01:40:52.767
[SPEAKER_03]: Yeah, because we launched a site, we launched a site and it's like, I'm going to look at do you see, you know, like, oh, it hasn't updated in four days.
01:40:52.867 --> 01:40:53.487
[SPEAKER_03]: I was last week.
01:40:53.747 --> 01:40:54.708
[SPEAKER_03]: And I was like, yeah.
01:40:55.238 --> 01:41:23.928
[SPEAKER_01]: Well, you can kind of extrapolate through through Google Analytics, just looking at traffic that's coming specifically from Google Organic, but there's so much information that search consequences you on the health and vitality of your website that extrapolating through Google Analytics just keeps hearing anxiety down, doesn't actually give you actionable information.
01:41:25.612 --> 01:41:27.433
[SPEAKER_03]: Yeah, exactly.
01:41:27.673 --> 01:41:31.455
[SPEAKER_01]: Google Testing a new search bar in the Google app.
01:41:31.495 --> 01:41:35.757
[SPEAKER_01]: In the Google Mobile app, there's a new redesigned search bar.
01:41:36.478 --> 01:41:37.478
[SPEAKER_01]: It says search.
01:41:37.738 --> 01:41:39.879
[SPEAKER_01]: Instead of search, it says ask Google.
01:41:41.000 --> 01:41:41.740
[SPEAKER_03]: Answer engine.
01:41:42.181 --> 01:41:42.981
[SPEAKER_01]: Yeah, I think.
01:41:43.001 --> 01:41:44.021
[SPEAKER_01]: But that's kind of telling.
01:41:44.682 --> 01:41:44.942
[SPEAKER_03]: Yeah.
01:41:45.567 --> 01:41:46.568
[SPEAKER_01]: Okay, so there you go.
01:41:46.728 --> 01:41:47.909
[SPEAKER_01]: It's changed around.
01:41:48.470 --> 01:41:58.518
[SPEAKER_01]: It's also got new buttons, buttons for uploads for nanobanana and for AI mode.
01:41:58.538 --> 01:41:59.219
[SPEAKER_01]: 90% chance.
01:41:59.239 --> 01:42:02.762
[SPEAKER_01]: Okay, according to search engine journal.
01:42:04.574 --> 01:42:15.259
[SPEAKER_01]: Um, the, uh, according to prediction model signals, there's a 96.8 chance of a Pandas style update within a month.
01:42:15.779 --> 01:42:17.860
[SPEAKER_03]: These are, sorry, Jim, they recommended reading.
01:42:18.060 --> 01:42:20.741
[SPEAKER_03]: Um, the red bar and the thing is off.
01:42:20.761 --> 01:42:21.941
[SPEAKER_03]: This is just recommended reading.
01:42:22.481 --> 01:42:23.442
[SPEAKER_01]: Okay, so what is that?
01:42:23.542 --> 01:42:27.584
[SPEAKER_01]: We're going to talk about 98 96.8 chance of a Pandas style update.
01:42:27.604 --> 01:42:28.984
[SPEAKER_01]: Okay, so that's hard stopping.
01:42:29.784 --> 01:42:30.625
[SPEAKER_03]: Well, I mean,
01:42:31.501 --> 01:42:33.442
[SPEAKER_03]: I've been talking about this myself for a while.
01:42:33.482 --> 01:42:39.646
[SPEAKER_03]: I think that they're going to go ahead and put out some sort of AI directed algorithm, John liked the posts when I put it out there.
01:42:40.319 --> 01:43:06.079
[SPEAKER_03]: not that that necessarily means anything but Dan Taylor put out a whole article about about it and he also talks a lot about Panda which most if you're new to the industry probably know nothing about you know it was the grandfather of all our you know content quality algorithms so I thought it would be good for people to read that it's in search engine journal and it's just a 90% chance of Panda stop date within a month he says a month
01:43:10.424 --> 01:43:28.223
[SPEAKER_01]: I really don't like when, uh, giving a, there's a 10% 10, 15, 20% but then Taylor knows the shit like I, I've like 10, 10 is writing, um, that's I shared it because I wouldn't normally share if it's a 90% because who knows what percentage of this, but his, the article itself is really founded on.
01:43:28.721 --> 01:43:33.562
[SPEAKER_03]: Like, how I panned a word and how it rolled out and then explaining why it mirrors a lot of what's going on right now.
01:43:33.862 --> 01:43:42.403
[SPEAKER_01]: So check that when I would search engine journal and another one, what Google AI mode follow up questions reveal about your content over at a search engine land.
01:43:43.364 --> 01:43:49.785
[SPEAKER_03]: Yeah, this is basically when you put a question in AI mode and it's like, hey, did you think about this and did you think about that?
01:43:50.365 --> 01:43:56.686
[SPEAKER_03]: This is the this is an article, it's pretty well researched about where that's coming from.
01:43:58.103 --> 01:44:04.506
[SPEAKER_01]: Yeah, this was a piece that was published on October the 9th, written by Claudia Tumina.
01:44:05.167 --> 01:44:10.169
[SPEAKER_01]: Again, over at Search Engine Land, well, Google AIMO'd follow up questions reveal about your content.
01:44:11.170 --> 01:44:14.151
[SPEAKER_01]: Last one, I required reading lists this month.
01:44:15.272 --> 01:44:18.854
[SPEAKER_01]: Oh, but, you know, there's one more thing, Ryan's for flexibility thing.
01:44:18.874 --> 01:44:22.616
[SPEAKER_01]: We'll get to the exact TikTok SEO, how to optimize your search,
01:44:23.822 --> 01:44:28.205
[SPEAKER_01]: How do I optimize for search and for the for you page?
01:44:29.306 --> 01:44:31.828
[SPEAKER_03]: Yeah, so I'm ticked to get real traffic.
01:44:31.848 --> 01:44:35.991
[SPEAKER_03]: You have to hit the for you page, which is where random people get your content.
01:44:36.011 --> 01:44:41.555
[SPEAKER_03]: There's not a lot of people that write about, by the way, when they call this the Microsoft Association, that's not really ticked.
01:44:41.936 --> 01:44:42.756
[SPEAKER_03]: But that's what they call it.
01:44:43.196 --> 01:44:51.863
[SPEAKER_03]: So, but this is one that gives you suggestions on how to get your content into their algorithm and also into Google's algorithm where it reports on social media.
01:44:52.952 --> 01:45:13.105
[SPEAKER_01]: quick warning on on this article it this one is written by search engine land itself which means it's written by search engine lands a. I can't attribute this to any author so I'm assuming it's written by the array if you don't tell me who's written it it's your machine writing it.
01:45:31.808 --> 01:45:39.333
[SPEAKER_03]: And it's a talk and so it does help you at least better understand how you get more views because TikTok is infamous for putting you in a 2 to 500-year-old jail stuff.
01:45:39.814 --> 01:45:49.400
[SPEAKER_01]: Okay, lastly, and this is the serious, most required reading, most series of them all, and the other ones are still important, reading important, but this one was recommended.
01:45:49.420 --> 01:45:59.628
[SPEAKER_01]: By Ryan Jones, supercon.com slash blog slash how to rank in perplexity photon.
01:46:00.647 --> 01:46:25.934
[SPEAKER_01]: how to rank in complexity what photon reveals and how supercon helps by Ryan Jones check it out at the supercon blog a breakdown on complexities algorithm by one of the more creative SEOs in the industry yes it's really good and he has to have a new free tool um... that will let you look at your share of voices free so v dot com what is it dot com i can remember
01:46:26.980 --> 01:46:38.454
[SPEAKER_03]: Um, so they check that out and then I start pre-confer looking at like these at course on relevance of your entities on a page, which are really important for parsing for elements and, uh, you might have new tool coming up, but we can't talk about it yet.
01:46:38.494 --> 01:46:42.518
[SPEAKER_01]: So indeed, you know, I'd love to that because that's actually going to be great.
01:46:42.558 --> 01:46:43.419
[SPEAKER_03]: Yeah, it's a really good one.
01:46:43.460 --> 01:46:44.561
[SPEAKER_03]: Yeah, yes, really good.
01:46:45.220 --> 01:46:47.843
[SPEAKER_01]: Okay, folks, we've gone around the clock a couple of times.
01:46:47.883 --> 01:46:49.945
[SPEAKER_01]: That was really fun show, really fun interview.
01:46:49.965 --> 01:46:53.669
[SPEAKER_01]: There was some stuff that we didn't get to talk about that.
01:46:53.929 --> 01:46:55.731
[SPEAKER_01]: Unfortunately, it's still there on the grid.
01:46:56.432 --> 01:47:05.122
[SPEAKER_01]: We're going to try to get back to that circle back to it next week, but chances are the world ain't going to let us, because it's a whole bunch of stuff that's going to be coming up between now and then until then.
01:47:07.063 --> 01:47:13.669
[SPEAKER_01]: Be nice to each other, be patient, be empathetic, listen really closely because that's where the search question is set.
01:47:14.690 --> 01:47:23.217
[SPEAKER_01]: And rank well, be well, and on behalf of Creasing and Shackinger from Cites Without Walls, this is Jim Hedger of Dich always media.
01:47:23.918 --> 01:47:24.939
[SPEAKER_01]: We'll talk to you next week.
01:47:25.719 --> 01:47:26.480
[SPEAKER_03]: Bye everybody.
01:47:38.347 --> 01:47:39.073
[UNKNOWN]: Thank you.