A Potential Death Blow To AI Search - WAN Show June 12, 2026

[00:00] WHAT IS UP EVERYONE AND WELCOME to the W

[00:04] show. We've got a great show lined up

[00:06] for you guys today. Google has been

[00:08] found liable for false information

[00:12] provided by AI overview. This is in

[00:16] Germany, so it might not necessarily be

[00:18] helpful to everyone, but it is a step

[00:21] towards holding big tech accountable for

[00:25] the false information that it propagates

[00:28] with its um, you know, let's call them

[00:31] sophisticated and yet foundationally

[00:35] flawed AI technology. In other news,

[00:38] speaking of flawed technology, Donut

[00:41] Labs's solidate battery has been exposed

[00:45] as a regular lithium ion thanks to some

[00:49] incredible work from one of our our

[00:52] fellow YouTubers who I Oh man, I hate it

[00:56] when people do this. Uh, so whose name

[00:59] I'm actually going to use, unlike the

[01:01] way that people usually cite YouTubers,

[01:03] Zeroth, fantastic video we're going to

[01:05] be talking about a little bit later. No,

[01:07] don't you hate that, Luke? When the

[01:08] headline of a traditional media article

[01:10] is YouTuber,

[01:12] >> yeah,

[01:13] >> YouTuber has a name. Uh, what else we

[01:16] got this week?

[01:18] >> You can just tell the Instagram

[01:20] algorithm what you want now, which is

[01:22] which is actually kind of interesting.

[01:25] And this is a this is a sounds really

[01:28] fantastic on the surface and is kind of

[01:31] but is also terrifying. Uh a Florida

[01:35] hospital is using Palunteer

[01:37] to catch sepsis earlier and it's saved

[01:40] 886 lives so far.

[01:44] So is this one of those cases where like

[01:47] even a broken clock is right twice a day

[01:49] Palunteer did something that we're

[01:52] supportive of?

[01:54] Uh well

[01:56] let's get more into it.

[02:19] The show is brought to you today by ODU

[02:21] Op manager Nexus, Squarespace, and

[02:24] Server Parts Deals alongside Dbrand, our

[02:27] rap partner. Razer, our laptop partner,

[02:29] and also Razer, our oops, that logo is

[02:32] supposed to be visible, I think, our

[02:34] chair partner. I'll be getting to this a

[02:35] little bit later. This is super cool,

[02:37] but but it will come later. All right,

[02:39] why don't we jump right into our

[02:41] headline topic, which is that Google has

[02:44] been found liable for false information

[02:47] provided by AI overview. In a potential

[02:50] landmark case, the regional court of

[02:52] Munich has ruled that Google, if they're

[02:55] going to, you know, have their product,

[02:57] spit out an output based on all of the

[03:00] data that they ingested, some of it

[03:03] legally, some of it, it seems, uh, we're

[03:05] still questioning the legality of it.

[03:08] They're accountable for what it says,

[03:10] which, you know,

[03:12] maybe is a good thing. The same way that

[03:14] newspapers were accountable for what

[03:16] they printed and you know radio shows

[03:19] are accountable for what they say. Well,

[03:21] Google's AI is also accountable. Yeah.

[03:24] PE, if you're a if you're a mass media

[03:26] outlet and you reach literally millions

[03:28] of people, you should have some

[03:30] accountability for having there be a

[03:33] factual basis to the things that are

[03:34] being propagated.

[03:36] >> Sure. I just don't know how often that

[03:37] actually like applies. It it applied to

[03:40] um uh what's his name? Um,

[03:47] >> you're not going to bring up Alex. You

[03:48] are okay. That's where you're going with

[03:50] it. Uh, well, yeah. I mean, it relies on

[03:52] somebody to actually hold you

[03:55] accountable to it in many cases, but

[03:57] there is there is a basis for but there

[04:00] is an expectation that if you are

[04:03] spreading information that you are

[04:06] responsible for the accuracy of it. The

[04:08] case was initially brought by two German

[04:11] publishing companies after Google's AI

[04:14] overview presented information about

[04:15] them, combined with that of shady

[04:19] companies and online scams, and then

[04:22] drew information that didn't appear in

[04:24] any of the linked sources,

[04:28] which I mean, Luke, have you ever

[04:29] experienced that with Google's AI

[04:31] overview where it goes blah blah blah

[04:33] this thing citation and then you click

[04:35] it?

[04:38] completely not in there. Like

[04:40] >> citations by AIS are are are pretty

[04:42] sketchy in general, but um this this

[04:45] does line up. This was pointed out by

[04:47] Flowplane chat as well, but I I think I

[04:49] lost the message, but this does line up

[04:51] with other things. There's uh customer

[04:53] support chat bots where what they say

[04:56] has been held as like fact. So the

[05:00] customer support chatbot offers you

[05:01] something and uh the the company's held

[05:05] to to follow through with it. Um, did I

[05:08] did I just see it? I thought I just saw

[05:10] it. Darn it. Somebody in chat pointed

[05:11] out I think it was Air Canada. Yeah, Pa

[05:13] Paige in chat said Air Canada was

[05:15] already held liable for something

[05:16] incorrect their support chatbot said.

[05:18] And I think there was other companies as

[05:20] well, not just Air Canada. So,

[05:23] >> Google,

[05:23] >> yeah, it kind of follows other things.

[05:25] Google pulled the classic defense of,

[05:27] "Well, users can fact check for

[05:30] themselves, and they do plan to appeal

[05:32] the ruling with a spokesperson saying,

[05:34] "This case focuses on specific and

[05:36] narrow errors, not the foundational way

[05:39] that AI overviews display web content."

[05:43] Um,

[05:45] right. But

[05:47] these specific and narrow errors are

[05:51] inherent to the way that foundationally

[05:55] AI overviews seem to work, at least at

[05:59] this stage in the game. Like these

[06:00] hallucinations are not

[06:03] new. They're not an unknown. Um, and to

[06:09] Google's credit, they do say in AI

[06:12] overviews that, you know, AI can make

[06:14] errors and that you need to check it or

[06:16] whatever, but like

[06:19] I I think there's a

[06:24] I think there's still a if we all know,

[06:28] right? It's kind of like the it's like

[06:30] what went down with end user license

[06:32] agreements, right? like you had these

[06:34] these long theoretically legally binding

[06:37] documents.

[06:39] Everybody knew on both sides, the

[06:41] companies writing these end user license

[06:43] agreements, the people signing them,

[06:45] everybody knew that nobody was reading

[06:47] them. So when push came to shove, it was

[06:50] ultimately found that if if someone was

[06:54] not actually wellinformed and was was

[06:57] not actually aware, then this stuff

[07:00] wasn't going to be binding. it didn't it

[07:02] didn't just absolve you of of breaking a

[07:05] law the fact that you wrote into your

[07:06] end user license agreement that you were

[07:08] like allowed to do it. So, in the same

[07:10] way, if we know that people are trusting

[07:12] this stuff and that they're not noticing

[07:14] these these errors and they're not

[07:16] actually doing the due diligence, then

[07:18] there, in my opinion, there's a certain

[07:20] responsibility on the companies that are

[07:22] putting this information. I'm going to

[07:24] put that in giant air quotes, putting

[07:26] this information out there to ensure

[07:28] that it is factual. And if it's not

[07:30] factual, who's accountable? Who's

[07:33] responsible?

[07:35] And yeah, Google's going to say the

[07:37] user, but like I think Google knows

[07:41] probably better than just about anyone

[07:43] else that the user is just not going to

[07:46] do that work. Like the entire the entire

[07:50] premise of this feature is that we're

[07:53] just going to surface the information in

[07:55] an even more convenient manner because

[07:57] we know the user doesn't want to click

[07:59] through and read a bunch of stuff.

[08:01] That's the whole point.

[08:03] >> Yeah. Yeah. So, we have a few discussion

[08:06] questions here. Um, first of all,

[08:10] should the AI pushers be allowed to use

[08:12] the public as their beta testers while

[08:14] they work out the kinks here? Is there a

[08:16] level I I'm going to clarify this or I'm

[08:18] going to add a layer to this one. Is

[08:20] there a level of disclaimer that you

[08:23] would consider

[08:25] acceptable so that it could just spew

[08:27] misinformation like this? Like, how beta

[08:30] would it have to look and feel?

[08:34] Um, I'm trying to see what it looks like

[08:35] right now, just out of curiosity. Sure.

[08:38] Um, it doesn't look like there's

[08:42] any flags whatsoever. So, above the

[08:45] fold, um, I I just Googled testing and

[08:50] it gave me an AI overview of the word

[08:52] testing. Um, and before clicking show

[08:56] more, there's there's nothing. It it has

[08:58] the little Gemini icon. It says AI

[09:01] overview. There's the the quick summary

[09:03] with its uh citations,

[09:06] some more information, more links on the

[09:08] right hand side, but there's no

[09:09] indication that this isn't ready to go.

[09:12] >> Yeah, I've got I did the same thing just

[09:14] to to make it a little bit easier for

[09:15] the people. So, it's got a couple

[09:17] citations here that makes it look very

[09:20] credible.

[09:21] >> I click show more.

[09:22] >> Once you click show more, there's fine

[09:24] print at the very bottom that says AI

[09:27] can make mistakes, so double check

[09:29] responses.

[09:29] >> Holy crap. I I actually do do I have

[09:32] that? Am I missing it, guys? Am I

[09:34] missing something here?

[09:36] >> It's on mine. It's right above where the

[09:39] chat input is where you would type just

[09:42] to the left of the like share, like, and

[09:44] dislike buttons.

[09:45] >> Uh-huh.

[09:47] Am I Am I Guys, help me out here. I feel

[09:50] like I'm gaslighting myself because I

[09:53] was pretty sure there was a thing. I do

[09:56] not see it.

[09:56] >> Whoa. I don't see it on yours. What the

[09:59] heck?

[10:00] >> Mine has it.

[10:02] What?

[10:04] What the heck?

[10:07] >> That's super weird. It's It's on my

[10:10] side.

[10:11] >> Okay, doesn't always show.

[10:13] >> People are asking if I have an ad

[10:14] blocker. Uh, I mean, it shouldn't do

[10:17] that, but like, what extensions do I

[10:19] have?

[10:20] >> I got nothing here, guys. Here's my

[10:21] extensions.

[10:23] What What else we got here, guys? What

[10:26] else we got?

[10:28] Do you want to open a a private browsing

[10:30] tab and do the same thing?

[10:31] >> I would like to do that very much.

[10:37] Okay.

[10:40] Blah blah. There it is.

[10:43] >> Yeah.

[10:44] >> Whoa, whoa, whoa, whoa. Okay. So, that's

[10:45] interesting. It also moved this

[10:47] interface over to the right.

[10:50] Yeah, it's probably just reactive, but

[10:52] yeah,

[10:54] >> it didn't need

[10:57] Yeah, but it's It went all the way from

[10:58] left aligned to right aligned, though.

[11:00] Oh, no. It's not quite left aligned. Oh,

[11:02] you're right. Interesting.

[11:05] Is it even quite right aligned, though?

[11:07] You know what? No, no, I think

[11:08] >> Yeah. No, it's just buffered off on both

[11:10] of them.

[11:10] >> Interesting.

[11:12] So, it just it just didn't for some

[11:16] reason sometimes. I mean man.

[11:20] Okay, so back to my question then. What

[11:23] is the degree of disclaimer that would

[11:26] be acceptable to you? Obviously what I

[11:29] encountered just there with no

[11:32] disclaimer and this is this is Chrome.

[11:34] This is like the chromest chrome ass

[11:39] Chrome of Chrome browser. I don't even

[11:40] have any extensions or anything like

[11:42] there is no excuse because I'm Googled

[11:46] all the way through the pipeline. So

[11:47] clearly that's not acceptable at all.

[11:50] And honestly, I don't even think the

[11:52] other version that I got in the

[11:53] incognito window, I mean, I think you

[11:55] and I could probably both agree that's

[11:56] not enough if it's going to be making

[11:59] like egregious errors like this.

[12:01] >> Yeah, fine print that you have to click

[12:04] show more to even see. Like it's

[12:06] literally not like you you can get

[12:09] results

[12:11] that are conclusive results and not see

[12:15] the disclaimer because you didn't click

[12:16] show more. You need to click another

[12:19] thing in order to get the disclaimer is

[12:20] like obviously not enough. I think it

[12:22] would need a pretty major badge. Um I

[12:25] don't look at the AI overview too often

[12:27] so I wasn't really ready for this but I

[12:29] was expecting it to next to where it

[12:31] shows the Gemini logo and AI overview. I

[12:34] was expecting a big fat beta testing

[12:37] batch. Um,

[12:41] >> right. So, you want it like right here

[12:44] like AI overview

[12:46] beta uh you know results.

[12:49] >> Yeah.

[12:50] >> May contain uh misin factual errors and

[12:54] uh incorrect information. Uh check all

[12:56] cited links in order to uh verify before

[13:00] believing anything it says. That would

[13:01] be would that be enough for you?

[13:05] Uh, I mean, I don't like the feature at

[13:07] all, but yeah, I think that would be

[13:08] enough for me. Um, someone in chat, uh,

[13:13] Mini Catalyst said, "Dumb user here. I

[13:15] didn't know it was in beta." I don't

[13:18] think that it is.

[13:21] >> Um,

[13:21] >> I don't think it's in I haven't seen I

[13:24] don't see any indication on the feature

[13:26] that it's in beta.

[13:28] Yeah, I just said beta uh because

[13:34] I don't know

[13:37] that's the type of thing that I'm used

[13:39] to flagging, I guess. Uh right, speaking

[13:41] for them. Um, so yeah, like uh I guess

[13:45] don't flag it with beta, but

[13:48] some much more in-your-face thing saying

[13:51] that this information

[13:53] um is man, I I'd like it to even say

[13:56] likely incorrect because I catch the the

[13:59] very few times that I've really used the

[14:02] AI overview on Google, it's it's wrong

[14:05] like

[14:07] >> a lot of the time. I don't know what

[14:08] model they're running, but it seems not

[14:10] sophisticated. It's It's fast. Like it

[14:13] it really does seem like they're running

[14:15] a pretty fast, pretty light,

[14:17] >> pretty

[14:19] not thorough

[14:21] model. And the funny thing about it is I

[14:24] forget what I was looking at earlier

[14:25] today. Um, but I was I was sitting in I

[14:28] was sitting in script review with Elijah

[14:31] and

[14:33] for

[14:34] and maybe this is maybe this is crazy,

[14:37] but we uh Oh, yeah. I remember what it

[14:39] was. Um, we were looking for comments

[14:42] like you know how sometimes in our

[14:43] videos we'll do like a collage of

[14:46] similar comments like last time on tech

[14:49] house you guys asked you know why we

[14:52] scavenged the copper the way we did and

[14:54] we'll go d and we'll put a bunch of

[14:56] comments about that. So, there's a

[14:58] really cool new feature in the YouTube

[15:00] dashboard where you can search your

[15:02] comments using AI search for anything

[15:05] that instead of just searching for like

[15:07] a keyword, you can look for anything

[15:08] about this topic or anything that was

[15:12] critical of Lionus' hairdo or whatever.

[15:14] You can just you can search for for

[15:16] sentiment now, which honestly

[15:20] pretty cool and useful

[15:24] uh feature for for LLMs. Anyway, um, so

[15:27] I searched for something specific. I

[15:29] don't remember exactly what it was, and

[15:31] it crapped out a list of all the

[15:33] comments, and I was like, "Oh, that's

[15:35] super great." And okay, here, Elijah,

[15:37] when you're going through and you're

[15:38] doing the guidance for this video, here,

[15:40] let me just copy this link to the studio

[15:42] dashboard to the search result, and I

[15:44] will throw that into the script um like

[15:47] like guidance column.

[15:50] And then I was like, wait, is that going

[15:51] to work,

[15:53] right? like is it I know it'll bring up

[15:55] the dash. I know I can link someone who

[15:57] also has access to the dashboard and I

[15:59] can even link them to the specific page,

[16:01] but is it going to is it going to link

[16:02] that search query?

[16:05] And so I clicked it to check it

[16:08] and it did.

[16:10] But I was like, hey, wait a second.

[16:13] I think it generated that result again.

[16:16] It didn't cache it at all.

[16:20] which honestly for something like the

[16:22] YouTube dashboard wouldn't surprise me

[16:24] that much. But for something like a

[16:28] super commonly searched term like this,

[16:32] like do did you get exactly the same

[16:34] result as me or did you get an LLM

[16:37] generated output fresh for you? Do you

[16:41] want to have a quick look?

[16:45] >> Um, yeah. So, mine I think I closed that

[16:48] tab already. Okay. I mean, you could

[16:50] just search again. I'm I'll read mine

[16:51] and you tell me when it stops matching.

[16:53] >> Testing is the systematic process of

[16:56] evaluating a system.

[16:57] >> Already off.

[16:58] >> Okay. So, here's a question for me and

[17:03] and you know I I am not a systems

[17:05] architect. I am not a deep expert on

[17:09] LLMs or anything of the sort.

[17:12] But what I do know is that right now we

[17:15] have big tech absolutely desperate to

[17:19] build more data centers in order to

[17:21] support their AI endeavors. Um we have

[17:26] just like a like a massive volume of use

[17:30] of products like search and we have this

[17:33] technology that we've had for you know

[17:36] all of computing um caching that

[17:41] it seems to me might be useful. So why

[17:44] would they run like a crappier model

[17:47] that to your point, yeah, I've noticed

[17:48] that AI overviews makes a ton of

[17:50] mistakes, like more than other better

[17:54] LLMs makes. So in order to manage their

[17:57] resource usage, wouldn't you think they

[17:58] would just like

[18:01] cache this or like cache common queries,

[18:05] update them once a day, once every

[18:07] couple days? like the the the definition

[18:09] of testing

[18:11] >> is not going to change between now and

[18:13] tomorrow.

[18:15] >> Yeah. And and I find it interesting

[18:17] because before if I would Google

[18:20] something like a a specific word and I

[18:23] would do this very often.

[18:24] >> Yeah.

[18:25] >> It would it would surface, you know, a

[18:28] definition. Um,

[18:31] sometimes I will I I've done this in the

[18:34] past and you know, maybe they don't want

[18:35] me to do this, but I've done this in the

[18:36] past where sometimes I'll Google

[18:37] something for a spell check.

[18:39] >> Yeah.

[18:39] >> Uh, because the Google search will be

[18:41] like, "Ah, you spelled this word wrong."

[18:42] Right. Um, now it'll think I'm like

[18:45] trying to talk to it

[18:48] and I run into some weird some weird

[18:50] scenarios. Um,

[18:54] yeah. I I like Google is genuinely

[18:58] significantly less useful for me and

[19:00] I've been using it way less.

[19:03] >> This is interesting.

[19:04] >> AO overview became a thing.

[19:05] >> This is interesting. They're doing at

[19:07] least a little bit of caching. Chat

[19:08] pointed out that my two search results,

[19:11] one in my regular browser and one in my

[19:14] incognito tab, did return identical

[19:17] results.

[19:19] H

[19:23] Okay.

[19:24] >> Interesting. Uh, try a different browser

[19:27] entirely.

[19:28] >> Uh, okay. All right. I'll open it.

[19:30] >> All right. Microsoft, are you are you a

[19:32] Microsoft representative?

[19:35] >> Are you trying to

[19:37] >> No. Well, I mean, you could use Firefox.

[19:40] >> I don't have it on here.

[19:42] >> You could use Chrome to get Firefox.

[19:45] >> I did get the same result.

[19:49] interesting

[19:50] >> because I know there was this thing

[19:52] recently, I don't know if we talked

[19:53] about it on my show or not, but uh

[19:55] Google Chrome is reserving like a bunch

[19:57] of space on your on your drive uh for AI

[20:00] stuff and I don't remember if that in

[20:03] included things like results and

[20:04] whatnot.

[20:05] >> Okay.

[20:06] >> Um so a different browser getting the

[20:08] same one is is interesting. Someone in

[20:11] chat Oh, I just saw it. Uh, Kuro Setsuna

[20:16] 29 said, "I wonder if it's region cache.

[20:18] I'm in a different region."

[20:19] >> Yeah,

[20:19] >> that would be interesting.

[20:20] >> Yeah, that could be it.

[20:22] >> Um, cuz mine goes, "Testing is the act

[20:24] of subjecting a system person or process

[20:27] to experiments." Uh, which is different

[20:29] than yours. Yeah, it's close but

[20:32] different right?

[20:34] >> Yeah. Yeah. A little bit little bit

[20:37] different.

[20:38] >> And I I refreshed the page and got the

[20:40] same thing. Let Let me try in a

[20:42] different browser.

[20:43] >> Okay. All right, this has become kind of

[20:44] an interesting rabbit hole. Sorry, guys.

[20:46] I know you were expecting Wan show, but

[20:49] uh

[20:50] >> I mean this is W show.

[20:51] >> This is kind of Wan show.

[20:53] >> Yeah.

[20:55] >> Testing.

[21:01] Oh,

[21:07] in okay

[21:10] in Firefox

[21:12] >> paste

[21:13] >> on this device. Um,

[21:19] I'm not logged in.

[21:21] >> Mhm.

[21:23] >> So, it's figuring out my geo location

[21:27] and making everything

[21:31] Uh, Mandrin, I'm guessing.

[21:34] >> Oh, that makes sense. Actually,

[21:37] >> it did not give me an AI overview at

[21:39] all.

[21:40] >> Okay, then. Yeah, that checks out.

[21:43] >> English

[21:44] trying to do URL things

[21:48] testing.

[21:49] >> Avon Fox says it's probably quicker to

[21:51] regenerate than to pull from cache.

[21:52] Yeah, but that's exactly what I was kind

[21:54] of wondering is like are they better off

[21:56] using a better model? Um, and then using

[22:01] caching versus a super light model that

[22:04] like makes a ton of mistakes. Um,

[22:06] especially when it's something that, you

[22:08] know, is a is a pretty static uh outcome

[22:11] like a like something that is not going

[22:13] to to change from day to day like the

[22:17] definition of a common word for

[22:18] instance.

[22:24] One second here. Sorry.

[22:27] There we go. Uh, can I change this?

[22:31] >> It's also interesting to me that it um

[22:35] it cites Cambridge dictionary

[22:39] >> as its first one uh with

[22:43] club.mministry of testing, whatever the

[22:45] devil that is, and also some YouTube

[22:46] short. Um, and then when I actually go

[22:49] there, oh, there's an ad. Oh, wow.

[22:52] That's a lot of LT store ads. Oh my

[22:54] goodness.

[22:58] Uh anyway, nowhere on the Cambridge site

[23:04] is like exactly that definition. And

[23:07] you'd think it would be it would be

[23:09] simpler and less errorprone to just give

[23:12] me

[23:13] >> That's a lot of LT store ads.

[23:14] >> That's a lot of LT store ads. Hey,

[23:16] >> holy crap.

[23:17] >> All the LT.

[23:19] >> How much money was just spent spamming

[23:21] your browser with that?

[23:24] >> I shudder to think. you know, it might

[23:26] have been pretty effective because it

[23:27] ended up on W show though. Um, okay. So,

[23:30] mine mine did not match.

[23:33] Um, but I'm wondering if So, so I went

[23:36] to specifically a like Canadian

[23:40] Google search thing. Um, I didn't

[23:43] actually realize this was even a thing

[23:45] at all. I I searched up Google Canada

[23:49] and uh it gave me like

[23:55] I don't know if you go to Google right

[23:57] now, does it show a like red and white

[24:00] Google logo with the Canadian Soccer

[24:03] Association logo and stuff flicking by?

[24:07] >> Uh

[24:08] sorry, this Yeah,

[24:11] >> let me see.

[24:13] I gotta wait for this thing to catch up.

[24:16] Yeah. Okay. Okay. So, I got that

[24:18] basically um and I searched on there.

[24:21] >> Yeah.

[24:22] >> And now I got testing is the act of

[24:25] evaluating, measuring or examining the

[24:27] capability, safety or characteristics of

[24:29] something and then blah blah blah blah

[24:30] blah blah. Is that the same as yours?

[24:33] >> Um I don't

[24:35] >> because that is very different than my

[24:36] previous one.

[24:39] >> Not quite. systematic process of

[24:42] evaluating a system, product or idea to

[24:43] determine its functionality, performance

[24:45] or safety. Yeah. So, it just it shuffles

[24:47] things around a little bit. And when you

[24:49] shuffle things a little bit, well, then

[24:52] >> you introduce the potential for errors.

[24:54] And it's just seems to be a feature

[24:57] rather than a bug. And that's just kind

[24:59] of what we're stuck with. And somebody's

[25:02] somebody's got to be accountable for it.

[25:04] And you know, Google's basically saying,

[25:06] "Hey, well, it should be the user." But

[25:08] I mean, we just with this the simplest

[25:10] of tests, the searching the word testing

[25:14] found that it will sort of move things

[25:16] around. It'll site sources that don't

[25:19] actually have exactly that definition.

[25:22] I just I and it's and it's hard because

[25:25] it's really tempting cuz it's right

[25:28] there. Like I find myself searching for

[25:30] something and going,

[25:32] "Right, right, right, right, right. I

[25:34] can't trust that. Don't look at that.

[25:36] look under that. Go under the fold and

[25:39] start clicking into things because it's

[25:41] just it's just so convenient.

[25:45] Why don't we jump into our next big

[25:46] topic? You want to pick one?

[25:49] >> Sure. Um let's let's tackle this because

[25:52] I'm interested in how this is going to

[25:54] go. Um

[25:58] a Florida hospital is using Palunteer to

[26:00] catch sepsis earlier. It saved 886 lives

[26:04] so far. Um

[26:07] yeah, Tampa General Hospital in Florida

[26:10] has used Palunteer built has used a

[26:12] Palunteer built system called the sepsis

[26:15] hub.

[26:16] >> Okay, that's a scary name.

[26:18] >> It is to save an estimated 886 lives

[26:21] since 2022

[26:23] having sepsis related deaths. The

[26:26] software pulls realtime data from

[26:27] electronic health records, lab results,

[26:30] clinician notes, and bedside monitors

[26:32] across roughly a thousand patients at

[26:34] once, watching for subtle early signals

[26:37] of sepsis that can get lost on a busy

[26:39] floor. Then alerts a rapid response team

[26:42] so flagged patients get antibiotics

[26:45] within an hour. It's also cut sepsis

[26:47] patient hospital stays by 30%. which is

[26:51] uh going to benefit all other problems

[26:54] basically. Uh context why this matters,

[26:57] sepsis kills around 350,000

[26:59] Americans a year, making it the leading

[27:02] cause of death in US hospitals. Tampa

[27:04] General isn't alone in its results

[27:06] either. A separate AI model called

[27:08] Composer tested by UC San Diego across

[27:11] two ERS showed a 17% drop in sepsis

[27:14] deaths after going live. Uh the

[27:17] discussion question is if an AI system

[27:19] could demonstraably have sepsis deaths

[27:22] deaths at one hospital, what's actually

[27:24] stopping every other hospital from

[27:26] deploying the same kind of tool? Um

[27:29] >> I think the thing stopping it is

[27:32] >> posture but also rigor.

[27:34] >> Um like they have to verify that this

[27:38] thing actually does work first. Um, I

[27:40] mean, now that it's working, I think

[27:43] that will probably help it to become

[27:45] something that would be

[27:46] >> you need to make sure it's not doing

[27:47] anything else potentially bad as well.

[27:50] Um,

[27:51] but it it is probably pretty good. And

[27:54] obviously, you can't exactly b 886

[27:57] lives. And I think uh the families of

[28:01] the people saved would probably even be

[28:05] annoyed that I even brought this up. But

[28:07] one thing that flagged me immediately

[28:09] when reading this is that Palunteer is

[28:11] receiving the health records, lab

[28:13] results, clinician notes, and all the

[28:15] details from the bedside monitors of

[28:17] every single patient at this entire

[28:19] hospital. Um,

[28:21] >> well, roughly a thousand patients. I

[28:23] don't know if it's every single

[28:24] >> I think that's everyone at the hospital.

[28:25] I think it's everyone at the hospital

[28:26] that has those things. Um, I could be

[28:30] wrong, but I I I I think that's the case

[28:33] because their their note here, um, the

[28:36] other thing that I I thought was

[28:37] interesting, um, is it's it's not just

[28:40] patients with sepsis

[28:42] >> because they're trying to do early

[28:44] warning detections for patients getting

[28:47] sepsis.

[28:48] >> Um, so I I think they're just if you're

[28:51] if you're that wired up, um,

[28:54] >> I think I think your initial assessment

[28:56] was right. Uh, by the way, this is

[28:58] hilarious. AI overview. Tampa General

[29:01] Hospital is licensed for 1354 total beds

[29:05] systemwide across a blah blah blah blah

[29:07] blah uh blah blah blah blah. About Tampa

[29:11] General Hospital. Tampa General is

[29:13] licensed for 982 beds

[29:16] right under it.

[29:18] >> Beautiful. Fantastic. I I I honestly I I

[29:21] I find mistakes in the AI overview like

[29:23] almost every single time I use it. It's

[29:25] I I hate it. It has genuinely made

[29:27] Google significantly worse for me. Um I

[29:31] I have seen myself using other tools

[29:34] because Google is worse for me now than

[29:37] it was in like I don't know

[29:42] 2015. Like genuinely 11 years ago. Like

[29:45] we've we've gone so far back that we're

[29:47] over a decade behind. Um, and you know,

[29:51] maybe maybe it'll eventually be better

[29:53] or something, but it's it's worse than

[29:55] useless right now. Um, but anyways, back

[29:59] back to the topic.

[30:01] Looking at Palanteer's playbook, um,

[30:05] they want to own all of the data in the

[30:07] world. They've they've basically

[30:08] publicly said that they don't want

[30:10] people to have any form of freedom and

[30:12] that people won't commit crimes anymore

[30:15] when there is genuinely no such thing as

[30:17] privacy because uh any any government in

[30:20] the world would instantaneously know if

[30:22] anyone did anything out of line. Um like

[30:25] they they are

[30:28] not good. Um, and I think it's going to

[30:31] be very very difficult for people to

[30:34] look at this thing that is making a

[30:41] massive positive impact. Um, 886 lives,

[30:46] having the deaths and reducing hospital

[30:49] stays for a common and major issue by

[30:53] 30% are not things that anyone can b at.

[30:57] Those are very serious numbers.

[31:00] >> Those are exciting things. I mean,

[31:01] anytime you talk

[31:02] >> difficult to deny,

[31:03] >> even to an AI skeptic, you know, one of

[31:06] the first things they'll say is like,

[31:08] >> "Yeah, when they when they announced

[31:11] this stuff, they promised it was going

[31:12] to revolutionize medicine. If that's

[31:14] what it was doing, then I would be fine

[31:16] with it." Like pretty much everyone I

[31:19] talk to who's like anti- AI is like okay

[31:23] but if it was this then it would be

[31:27] fine.

[31:28] It's just that unfortunately you've got

[31:31] companies like Palanteer that are doing

[31:34] that stuff where we're saving human

[31:37] lives which seems like a pretty useful

[31:39] and noble endeavor for AI but also using

[31:43] it as a means to collect that data to

[31:45] pursue

[31:47] a lot less comforting future outcomes.

[31:52] Um,

[31:56] >> yeah.

[31:57] >> So, I think that could be one of the

[31:59] things back to our discussion question,

[32:01] preventing other hospitals from

[32:03] deploying this same technology.

[32:06] >> I also think like you wouldn't really

[32:08] have a choice. Um, a lot of times

[32:10] depending on how bad your condition is,

[32:11] you just literally have to go to the

[32:13] closest hospital. Um, I've heard some

[32:17] weird stuff about the American medical

[32:19] system where like you're out of your

[32:22] insurance capture zone or something, so

[32:25] you have to get

[32:28] driven further something. I don't know.

[32:31] I'm a Canadian. Our healthcare system

[32:33] does not work that way. Um, but like

[32:37] this stuff's just going to be running.

[32:38] Like I I I doubt you're, you know,

[32:45] There's there's a lot of

[32:48] systems built around health care and end

[32:50] of life where like the funeral business

[32:53] is incredibly lucrative because

[32:56] when someone passes

[32:58] >> you don't want to see

[33:01] >> yeah people are very unlikely to say no

[33:03] >> y

[33:04] >> to upsells. Uh so the industry makes a

[33:07] lot of money. Uh the healthcare industry

[33:09] is very similar. it. There is a ton of

[33:13] money that goes into especially near end

[33:15] of life because if someone uh closely

[33:18] related to you is suffering, you are

[33:20] very unlikely to say no to um upgrades

[33:24] in in care.

[33:25] >> Um this is one of those situations where

[33:27] it's like are you okay with the health

[33:31] records, lab results, clinician notes,

[33:33] bedside monitor access, and probably

[33:35] tons of other things being mass data

[33:38] harvested from your loved one. Oh, also

[33:40] they might not die if you click this

[33:43] checkbox.

[33:46] A lot of people are going to click the

[33:47] check box. Yep.

[33:48] >> I I would wager pretty much everyone. Um

[33:52] but then that also does lead to

[33:56] the, you know, the super dystopian

[33:58] futures that specifically Palanteer has

[34:01] been saying they want to create. Ivonne

[34:03] keeps asking me what I want for my

[34:05] funeral, and I'm like,

[34:07] >> I'm dead.

[34:10] Do whatever makes you happy. I couldn't

[34:11] possibly care.

[34:13] >> Yeah,

[34:15] >> monetize it. I don't know.

[34:18] Charge admission.

[34:20] I'm sure there's at least someone who'd

[34:21] want to see me dead. They'll show up.

[34:26] Fair farewell. Farewell, inferior tech

[34:28] liners.

[34:30] >> Church admission. Oh my god. Um,

[34:34] >> I mean, you've heard you've heard me say

[34:35] that if you guys catch it on camera, you

[34:37] have to you have to monetize it, right?

[34:40] >> I have. I was actually I was talking to

[34:42] someone about that surprisingly

[34:44] recently, and I I uh

[34:49] Yeah,

[34:50] I I was like, you know, he said that

[34:52] before. I don't know if he still feels

[34:54] that way or not. So, I guess you do.

[34:56] >> Do it. I mean, how like we got

[34:58] everyone's got to get their severance

[34:59] and everything. if for whatever reason

[35:01] we can't keep going without me, right?

[35:02] You got to you got to milk it. Milk it

[35:04] as good as you can, as long as you can.

[35:06] And then when I'm when it's over, it's

[35:09] over. And it was it was a wonderful ride

[35:11] while it lasted,

[35:13] man. Okay. Sorry. Uh Katsune Lane in

[35:17] full chat said, "Luke's dedication to

[35:19] Lionus is a W show delivered live from

[35:21] beside his coffin." That would be so

[35:25] depressing

[35:26] because it would be so long.

[35:29] >> Open casket. I hope

[35:31] >> I'd have to sit there. I'd have to sit

[35:33] there like open casket for like 4 hours.

[35:36] Oh my god.

[35:40] Oh,

[35:40] >> here. Practice.

[35:46] >> I can't even. Hold on one second. You

[35:47] got to

[35:51] Oh,

[35:53] >> how long could we keep the streak alive?

[35:54] I mean, would you would you would you

[35:56] bring would you bring my ashes and put

[35:58] them put them next to you? You and you

[36:00] and Dan can just do the Wan show and

[36:02] just put put me on the shelf.

[36:05] >> I Yeah, I think

[36:06] >> from the N show.

[36:11] >> It's just the sponsor spots. the sponsor

[36:13] spots can just go to like a earn camera.

[36:19] >> We could uh we could make the lid go up

[36:21] and down like the Canadians on South

[36:22] Park.

[36:23] >> If anybody sends a if anybody sends a

[36:25] comment that's a direct question for me,

[36:27] you could just like you could just cut

[36:29] to my camera and

[36:32] then just we'll leave it there for a bit

[36:34] and then cut back and keep going.

[36:36] >> Awkward silence.

[36:39] The one thing, the one place I draw the

[36:42] line though is no LLM resurrection of

[36:45] me. I am one of the people on earth who

[36:49] would be relatively easy to to LLM in

[36:54] honestly a pretty accurate fashion,

[36:57] >> but I don't consent to that. I don't

[37:00] approve of that. Um, so so I'll just on

[37:03] the on the urn camera, I'll just have to

[37:06] like flap the lid of the urn to pretend

[37:08] you're talking and just

[37:09] >> Yeah.

[37:09] >> try to

[37:10] >> You have to puppeteer your voice.

[37:12] >> You can even do the voice. Okay.

[37:15] >> Yeah. Yeah. Yeah.

[37:16] >> I'm fine with that. I'm fine with that.

[37:18] >> I don't know who outside of Dbrand would

[37:20] accept those ad spots.

[37:23] >> I mean Dbrand would do it. If Dbrand's

[37:25] still around,

[37:26] >> they they would do it.

[37:29] Yeah,

[37:30] >> I can find out.

[37:32] >> Also, laughing about

[37:34] flapping your urn. I'm I'm in a rather

[37:37] public place. Uh just so viewers of the

[37:40] show know, I'm at um Lee Hua Hong. I

[37:44] I've I've featured this place a couple

[37:47] times. There was one time somewhere

[37:49] around 10 years ago in the desk and

[37:51] internet cafes video that I I featured

[37:53] this place. And then um there was also

[37:59] a short a little while ago, but I needed

[38:01] a good internet connection.

[38:03] >> Mhm.

[38:03] >> So I was like uh we were having some

[38:05] issues with hotels and their internet

[38:07] connections, so we just decided to do

[38:10] the show from here. Um and people can

[38:14] definitely hear me.

[38:15] >> Nice. Hold on. Hold on. Hold on. My

[38:19] rep's typing. There's a robot. There's a

[38:21] robot typing.

[38:23] I said you would sponsor wrapping my

[38:25] earn, right? Yeah, I'll schedule it,

[38:28] too. Classic.

[38:32] >> Yeah, not not surprised.

[38:35] >> Yeah, that checks out. I mean, they've

[38:37] tried to kill me enough times, sending

[38:39] me packages full of broken glass and

[38:41] whatnot.

[38:42] >> I got a comment.

[38:43] >> Maybe that's why they're just annoyed

[38:44] they haven't been able to do that type

[38:46] of an ad spot yet. They've been waiting

[38:48] for so long.

[38:51] >> Oh man.

[38:52] All right, we should probably jump into

[38:54] our next topic here. Uh, th this is on

[38:59] the surface kind of cool. You can

[39:02] apparently just tell your Instagram

[39:04] algorithm what you want. Now, Instagram

[39:06] is expanding its your algorithm feature

[39:09] to the main feed, letting you directly

[39:12] tell the app what you want to see. It

[39:14] shows you a list of topics Instagram

[39:16] thinks you're interested in based on

[39:18] your activity. And you can add topics

[39:20] that you want or remove ones that you

[39:22] don't. Changes will carry across real

[39:25] explore and now the main feed since it's

[39:28] all one system. It started in the US and

[39:30] is rolling out globally in English.

[39:33] Instagram head Adam Missouri framed this

[39:35] as fixing a one-sided relationship. The

[39:39] system learns from what you tap and

[39:40] watch, but you've never been able to

[39:41] just tell it what you want. He credits

[39:44] large language models for making it

[39:46] possible since older ranking systems ran

[39:48] on data that no human could read, while

[39:50] LLMs can now sort content into plain

[39:53] language topics. For now, it is limited

[39:55] to topics, but Instagram says that

[39:58] controls for specific people, moods, and

[40:02] content types are coming.

[40:09] I mean, I think it's a good thing.

[40:11] >> It It seems like what it always like

[40:14] should have had the ability to do. I I

[40:16] recognize the I guess the the challenges

[40:20] around like metadata and tagging and and

[40:24] like how manual something like that

[40:26] could be. Um I can see how LLMs could

[40:31] help with that. But also, I mean, we've

[40:33] spent 90% of the show so far talking

[40:36] about how fundamentally flawed they are

[40:39] and how often they make mistakes.

[40:41] Uh,

[40:43] this is I

[40:47] this will be abusable. I don't know if

[40:49] there will be a reward in abusing it,

[40:52] but this feels like the it is going to

[40:55] be susceptible to like Spider-Man and

[40:57] Elsa style issues. For people that don't

[41:00] understand that reference, kids YouTube

[41:03] years ago was taken over by some really

[41:06] weird content um of people like

[41:09] cosplaying as Spider-Man and Elsa and

[41:11] other characters and doing some weird

[41:14] stuff that you wouldn't really expect to

[41:16] be on the kids side of YouTube. Um but

[41:18] it it got there because of

[41:22] the framing of the video and I could

[41:25] imagine that something similar could

[41:27] happen in this type of system. Um, I

[41:30] think the reward is

[41:33] >> I mean the reward of Spider-Man and Elsa

[41:35] >> was just

[41:35] >> was always

[41:36] >> AdSense.

[41:37] >> Well, I guess yeah, AdSense. So maybe

[41:39] yeah, more reward than for just getting

[41:41] views on Instagram, but there's also the

[41:43] there's also the reward of just getting

[41:46] clout and building up followers on

[41:48] Instagram. Social media accounts are

[41:50] worth money. Um,

[41:52] >> yeah,

[41:54] >> and especially, you know, with

[41:56] automation tools, the the ability to to

[41:59] create accounts to abuse these systems

[42:02] to grow to some threshold to then be

[42:05] farmed off to, you know, scammers who

[42:08] need plausible looking Instagram

[42:09] accounts or, you know, whatever the

[42:10] model is. I I often find myself very

[42:13] surprised by the things that scammers

[42:16] will come up with. Like there was that

[42:18] recent one where there was that like

[42:20] warehouse of consoles that was farming

[42:22] like FIFA card packs or something. I

[42:24] can't remember the exact details of it.

[42:25] I'm not a I'm not a FIFA guy. Dan, do

[42:27] you remember?

[42:29] >> Oh, okay. I just assumed you were into

[42:31] football being an English person and all

[42:32] of that.

[42:33] >> Oh, cricket like a normal person.

[42:35] >> Got it. That makes sense.

[42:36] >> I don't go outside.

[42:40] >> Uh yeah, I remember that topic. I don't

[42:42] remember the specific details. I think

[42:44] it was like farming throwaway matches

[42:46] basically in order to get uh pack

[42:48] openings, like free generating pack

[42:50] openings. I don't know why that benefits

[42:52] them or anything, but yeah, it's it's

[42:55] interesting. I I can't necessarily

[42:56] imagine how you would abuse this. It

[42:58] just feels to me uh how an LLM would

[43:02] view this content

[43:05] is probably abusable in some way. I

[43:08] don't know what to what end, but it

[43:10] feels like it probably would be. Um, I

[43:12] also don't know like what kind of

[43:14] they're even saying you can sort by

[43:15] moods, which is interesting. So, like is

[43:17] it is it doing the Lionus maneuver and

[43:19] is it reading the comments? Um, so could

[43:23] could you manipulate this setup by

[43:28] convincing your audience somehow to all

[43:32] comment in a particular way? Um, I don't

[43:35] know. This almost feels like something

[43:36] like Spiffing Brit should dive into,

[43:38] although he seems to just do YouTube

[43:40] stuff. um

[43:41] >> beast.

[43:43] >> Yeah, but it'd be interesting. I mean, I

[43:47] I like the idea

[43:49] of being able to more manually control

[43:51] our feeds. I I I also respect Lionus'

[43:53] dedication to keeping his Facebook

[43:55] Marketplace weird as heck. Um but I I if

[44:00] if I could just tell Facebook

[44:02] Marketplace that I want it to show me

[44:04] weird things, that would also be nice.

[44:06] Um

[44:07] >> yeah, like one thing

[44:09] >> I think this is good. One thing I've

[44:10] wanted from YouTube for a long time is

[44:14] the ability to tell them what my video

[44:17] is about a little bit because they can

[44:19] mostly figure that out on their own. But

[44:21] what I really want is to tell them who

[44:23] it's for. Because remember remember way

[44:27] back when you made uh here I'm actually

[44:29] just going to pull up the video. Uh come

[44:32] on

[44:35] here. This one

[44:36] >> you might be able to imagine. Oh, for

[44:38] crying out loud.

[44:42] This one right here.

[44:44] >> Installing a video card. How to

[44:46] >> basics on Lionus Tech Tips 10 years ago.

[44:50] >> This video got absolutely dogpiled out

[44:55] of the gate.

[44:55] >> Hated it. People were like actually

[44:57] pissed when they launched.

[44:59] >> So angry. and and and it there was

[45:02] momentum around it because that was back

[45:04] when the like dislike ratio was visible

[45:07] publicly. So people

[45:08] >> it was very negative

[45:09] >> saw that people hated it and then they

[45:12] piled onto it even more and it it

[45:14] created this like like death spiral

[45:16] spiral of hatred and negativity around

[45:18] this video and

[45:21] if you actually just you know watch the

[45:24] video you don't say anything wrong.

[45:28] Hold on. Is it?

[45:30] >> Yeah, it was it was people hated it just

[45:32] because it was beneath most of the

[45:35] average viewers of LT

[45:37] >> at the time.

[45:37] >> Um, yeah, it it wasn't that the

[45:40] information was incorrect. It's just

[45:41] that they were like,

[45:43] >> "It's a graphics card, bruh."

[45:45] >> Yeah, you just put it in.

[45:47] >> But I I'm pretty sure we included things

[45:48] like dduing your drivers and like stuff

[45:50] like that. Like I think it was I think

[45:52] it was like decently

[45:53] >> No, you did a good job

[45:54] >> presented at the time.

[45:55] >> You did a very good job. It was it was

[45:57] lambasted because it hit the wrong

[45:59] audience like you're kind of saying.

[46:01] >> Exactly. So, what I've been asking

[46:03] YouTube for is the ability because

[46:05] because what's happened is we've kind of

[46:08] our channel has kind of Michael Bay

[46:10] itself because we can't just make a

[46:13] video on installing a CPU anymore

[46:17] because what's going to happen is it you

[46:21] bumped your mic. Careful. Uh what's

[46:23] going to happen is it's going to jump

[46:26] into the feed of everyone from someone

[46:29] who subscribed to LTT yesterday to

[46:32] someone who subscribed to LTT 17 years

[46:35] ago. And these folks, your your regular

[46:39] viewers are going to feel like like

[46:42] patronized and they're going to feel

[46:44] like this is a complete and utter waste

[46:46] of time. And what tends to happen is

[46:49] algorithmically YouTube will kind of go,

[46:51] well your channel is about, you know, X

[46:54] and the people who are interested in X

[46:57] don't like this. Therefore, this is

[46:59] probably bad content. But what I would

[47:03] actually really like is the ability to

[47:05] tell YouTube, hey, this video is not for

[47:09] my regulars. This is not

[47:11] >> This was a new viewers like only

[47:14] basically

[47:15] >> and I'm okay with that resulting in it

[47:18] potentially doing worse, especially out

[47:19] of the gate, but I don't want it to

[47:22] massively negatively affect the

[47:24] long-term potential of the video by it

[47:26] nuking a bunch of potential views on

[47:28] people who would never click on it. Cuz

[47:30] like if you know if someone's in flow

[47:32] plane chat or or they've just been you

[47:34] know I I'll meet people at events all

[47:36] the time that are like I've been

[47:37] watching since the NCIX days. like,

[47:39] "Bro, you know how to install a GPU."

[47:41] >> Exactly.

[47:42] >> Like, it's you don't you don't need

[47:44] this.

[47:45] >> Um, the the vast majority of people who

[47:47] have watched more than, you know, one to

[47:51] three LT videos or or realistically any

[47:54] other uh computer hardware tech videos

[47:56] on YouTube, um, isn't going to need a

[47:58] video on how to install a GPU. But that

[48:01] doesn't mean that a bunch of people

[48:04] don't need that. Um,

[48:05] >> so here's another great example. the

[48:07] first time you've ever built a computer.

[48:09] If you just spent a bunch of money on

[48:11] hardware and you want to build the

[48:12] computer for the first time and you

[48:13] don't have an inerson mentor, you want

[48:15] to make sure that you don't ruin this

[48:16] incredibly expensive component as you

[48:18] put it into your computer, you might

[48:20] want to watch a little video on how to

[48:21] install it.

[48:21] >> Yeah. And so we we wanted this was that

[48:24] was supposed to be a series. That's why

[48:26] the title is appended with this how-to

[48:29] basics. I had in my mind that I wanted

[48:31] to create three sort of series of like

[48:34] to really get back down to tech tips and

[48:37] do like how to basics which was kind of

[48:40] a like a play on how to basic which was

[48:43] like kind of a viral thing at the time

[48:44] and then I wanted how to intermediate

[48:47] and like how to advanced. So we would do

[48:50] these tutorials on everything from

[48:51] installing a graphics card to like you

[48:55] know reconfiguring uh ports on your

[48:58] router to uh you know setting up a

[49:02] setting up a a NAS with you know

[49:05] multiple tiers of high speed and low

[49:06] speed storage and we'd kind of we'd kind

[49:08] of segment them out and just create

[49:10] these like very evergreen guides. you

[49:12] know,

[49:14] >> sorry,

[49:14] >> the reception to it was so negative that

[49:17] we did a couple of them and we kind of

[49:18] went, well, holy crap, we can't we can't

[49:21] do this. And then ultimately, years

[49:24] later, that like dislike ratio actually

[49:26] totally turned around and the video now

[49:28] has, you know, 2.3 million views. But

[49:31] that's not nearly as many as it probably

[49:32] would have if it hadn't had its growth

[49:35] stunted right at the beginning there.

[49:39] >> Yeah. It's you, you know what's kind of

[49:41] interesting is I've heard that the

[49:43] youngans these days um to to go back to

[49:47] a previous topic actually also hate

[49:49] Google search now. Uh like Google search

[49:52] just sucks these days. The the results

[49:54] are terrible and the AI search is worse

[49:56] than the auto stuff that they had in the

[49:58] past. Like it's just Google search is

[50:00] just going completely down the hole. Um,

[50:03] but a lot of the youngans are using like

[50:05] Tik Tok for search,

[50:07] which

[50:09] >> hurt my brain a lot when I first heard

[50:11] it. Um, but apparently it's because

[50:16] they're they're using it in the way that

[50:18] a lot of us probably use YouTube, which

[50:22] is where we we look it up like, you

[50:24] know, how do I fix my washing machine

[50:28] >> or whatever.

[50:28] >> Yeah.

[50:29] >> Um, here's a weird question. Someone in

[50:32] someone in chat said that you can

[50:34] Instagram has like a trials feature

[50:36] where it won't send it to your

[50:38] subscribers or something. I I lost the

[50:41] message. Uh Andrew Candy, Instagram can

[50:43] kind of do this with trial reels. It

[50:45] doesn't send them to your followers. I

[50:47] don't know exactly how that works, but

[50:48] maybe the how-to basics series, maybe

[50:52] with a different name, not sure, could

[50:53] return, but in shorts format. I could

[50:56] see potential there.

[50:57] >> This card, but the the shorts version,

[51:00] how to install RAM, but the shorts

[51:03] version.

[51:04] >> It can also even work in long form

[51:05] though. Uh here, Dan, I'm just going to

[51:07] see if my laptop is working. Oh, yeah.

[51:09] There we go. So, I'm on my uh I'm on my

[51:11] daily driver here. So, I'm logged in.

[51:13] So, here's an example of one that kind

[51:16] of shows

[51:17] how bad this can be. Uh so we we we

[51:23] uploaded uh like just building a PC you

[51:26] know just a build guide

[51:28] >> for the first time in like ages. So

[51:31] first day performance of this video was

[51:34] 458,000 literally half of the upper end

[51:38] of our typical range like absolute

[51:41] bombed out of the gate. And then

[51:46] eventually

[51:48] the algorithm figured out, oh, okay,

[51:50] yeah. Oh, this is a highquality video

[51:52] that just the subscribers to this

[51:54] channel were not interested in.

[51:58] Now it's more than double the high-end

[52:01] of your typical range. And just, you

[52:04] know, daily as the English- speakaking

[52:06] world awakens and slepts and awakens

[52:11] again, it just gets views. Um,

[52:16] but it would be really, but it doesn't

[52:17] always work out that way. So, right, the

[52:19] other one that I was going to pull up

[52:21] was Riley did a $1,000 gaming PC. So,

[52:26] here we'll go back to Linus laptop. This

[52:28] one also bombed out of the gate in spite

[52:30] of the very likable Super Riley and then

[52:36] never found that audience. And you know

[52:39] what? Maybe we just, you know, maybe

[52:41] just didn't, you know, work. Maybe a

[52:43] $1,000 guide was not as good as a $2,000

[52:46] guide. Whatever. I don't know the exact

[52:49] answer to why one of them went and one

[52:50] of them didn't. But I've I've been

[52:52] telling Google for years. I would just

[52:54] love to be able to just tell you. I can

[52:58] tell you. I know who it's for. And as a

[53:01] as a variety channel, it's something

[53:03] that we've struggled with our entire

[53:05] existence. And I think we struggle with

[53:09] more and more now where one video on our

[53:12] channel one day is touring a data center

[53:16] and then the next day is like tearing

[53:19] down the walls of the tech house so that

[53:21] we can run conduit and and low voltage

[53:24] wiring and then the next day is a review

[53:27] of a steam controller. like we're all

[53:29] over the place. And I have I have

[53:32] noticed that algorithmically they just

[53:35] don't know what to do with us compared

[53:38] to a channel that is more more

[53:40] predictable. Uh like even within the

[53:43] tech niche, there's other channels that

[53:44] are more predictable in terms of their

[53:47] their tone from one video to the next.

[53:49] And I'd love to be able to just say,

[53:51] "Hey, this is for people that are super

[53:54] into biohacking or, you know, this is

[53:59] for people who are really into, you

[54:00] know, enterprise hardware, you know,

[54:02] whatever right?"

[54:05] >> But I can't.

[54:06] >> Yeah.

[54:07] >> There's breaking anthropic news, by the

[54:08] way, that we should probably acknowledge

[54:10] because chat is gloating.

[54:14] >> I'm working on that right now. Um

[54:18] >> basically they issued a statement um

[54:22] >> the US government shut them down.

[54:25] >> Statement on the US government directive

[54:27] to suspend access to Fable 5 and Mythos

[54:30] 5.

[54:32] Um

[54:34] yeah, citing national security

[54:37] authorities, the US government has

[54:38] issued an export control directive to

[54:41] suspend all access to Fable 5 and Mythos

[54:43] 5 by any foreign national, even those

[54:46] inside the United States, including

[54:50] anthropic employees.

[54:54] Uh

[54:58] so

[54:59] >> this feels like

[55:01] Yeah. To ensure compliance, they have

[55:03] just turned Fable 5 and Mythos 5 off

[55:08] for everyone.

[55:10] >> Uh, this Dan Dan pointed this out in

[55:13] full chat. Uh, but this is probably the

[55:16] best marketing anyone on the entire

[55:17] planet could have asked for. But I

[55:19] actually think this might be hilariously

[55:24] the whole marketing angle might not be

[55:27] the goal that the US government wanted

[55:29] because I suspect they are still in a

[55:31] bit of a tiff over when Anthropic stood

[55:34] up to them a little while ago. Um, and

[55:38] with with the IPOs coming soon, I

[55:42] suspect this might have been uh an

[55:45] attempt to hold them back actually. But

[55:47] I think just like last time, this is

[55:50] ultimately going to be just like an

[55:52] insane marketing move. Uh, promoting the

[55:56] absolute heck out of anthropic. Um,

[56:01] the the the including internal employees

[56:03] thing is is wild. We we had um Conrad,

[56:07] the the guy who makes LT store. Yeah. In

[56:10] regards to development, um was in chat

[56:12] earlier saying he used Fable 5

[56:15] personally um and said I I don't I don't

[56:18] want to misquote. I don't remember

[56:19] exactly what he said, but he said very

[56:20] positive things about it. Um

[56:24] very interesting. Honestly, um

[56:28] the Frontier models that are coming out

[56:30] right now are pretty nuts. Um

[56:35] yeah, Mari squared in fulloint chat said

[56:38] not me currently using Fable 5 to

[56:40] rewrite a whole platform. Um

[56:44] yeah wild

[56:46] wild

[56:48] in in before uh US citizens start

[56:52] reselling uh their their access to Fable

[56:55] 5 to foreign nationals. Um,

[56:57] >> yeah.

[57:00] >> If only there was if only there was some

[57:02] technology that could make it so that

[57:05] you could appear as though you were in a

[57:07] different region than you were. If only

[57:09] there was some way to pay for things

[57:12] with a, you know, US bankisssued payment

[57:16] method. If like this, I don't know how

[57:18] they could possibly hope to to enforce

[57:21] this once they actually do reenable

[57:24] Fable 5. Yeah, that's what I was going

[57:26] to say is like I I don't think it's

[57:27] going to be as simple as as a VPN, which

[57:29] is why they actually shut it down.

[57:31] >> No, but you will but you'll need but it

[57:33] won't someone will overcome it. This

[57:35] ain't going to be impossible to overcome

[57:37] no matter what.

[57:39] >> At the very least, I suspect it's going

[57:40] to be people individually reselling API

[57:43] key usage um somehow. But it's yeah,

[57:50] wild wild. Verified US ID requirements

[57:53] incoming. Maybe. Yeah.

[57:55] >> Okay. I'm not going to say who said

[57:57] this, but somebody said, "I got a notice

[58:01] on a litigation hold on all

[58:02] communications about anthropic. M DoD

[58:05] employee."

[58:07] I mean, I think you mean DOW.

[58:10] Um, but yeah, for now DoD.

[58:13] >> I've I've heard I've heard the

[58:15] Department of War thing

[58:17] isn't official. They just call it that.

[58:20] No, but it's it like went looking like

[58:23] probably official today or yesterday or

[58:26] something. Yeah,

[58:27] >> got it. Got it. Got it. Got it. Um I I

[58:31] don't I don't know what's going on with

[58:32] that, but thought that was kind of

[58:35] funny. Has to be approved by Congress.

[58:37] Okay. Yeah.

[58:38] >> Yeah.

[58:39] >> Um yeah, wild situation. Absolutely wild

[58:41] situation. I I think like um I don't

[58:44] know if we've talked about it on W show

[58:45] or not um but but PewDiePie's video on

[58:49] Odysius.

[58:50] >> Yeah.

[58:51] >> Um

[58:52] it's it's you know we've talked about

[58:55] PewDiePie a few times over the years on

[58:57] W show. Uh this is not one of the ways

[58:59] that I expected to be talking about

[59:00] PewDiePie, but he has been

[59:04] doing incredible things when it comes to

[59:06] locally hosted AI and Agentic stuff. and

[59:09] his video on Odysius uh his his project

[59:14] uh which he has open sourced um has 3

[59:17] million views now. And you know we

[59:19] talked about this last show but u you

[59:22] know local models are not are not

[59:25] necessarily you know brushing shoulders

[59:27] with these frontier models. Um,

[59:30] >> no, but they're right now,

[59:33] >> but they are very useful. And I really

[59:36] think the in the same way that with with

[59:39] the whole operating system debate right

[59:41] now, my solution is to very happily just

[59:43] dual boot on both of my systems. I'm

[59:45] dual booting on my laptop, I'm dual

[59:47] booting on my desktop, and I'm very

[59:49] happy. Um, I think the like prime

[59:52] solution for me at least is going to be

[59:54] using both. um saving your API costs on

[59:59] something big and heavy and powerful

[1:00:02] like uh Fable 5 or or any other like

[1:00:05] crazy frontier model because you know in

[1:00:07] two weeks it'll probably be something

[1:00:09] else that's in the lead. Um and then for

[1:00:12] most of your

[1:00:15] honestly probably most of your work but

[1:00:17] it really depends on who you are and

[1:00:18] what you're doing. Um, you can use local

[1:00:20] models in order to

[1:00:22] >> just like your your just dumb [ __ ]

[1:00:24] like what's the

[1:00:26] >> what's the weather going to be tomorrow

[1:00:28] just to to interact with, you know, your

[1:00:31] your DIY Google Home that doesn't

[1:00:34] actually communicate with Google's

[1:00:35] servers. Uh, like that that cute little

[1:00:38] robot thing that Nick Harris from the

[1:00:40] lab had in his AMD ultimate tech

[1:00:42] upgrade. Like paying for tokens to run

[1:00:45] something like that to turn your lights

[1:00:46] on and off is cra would be crazy.

[1:00:49] And

[1:00:50] >> yeah, there's no need to

[1:00:51] >> and you Yeah, you just don't need to

[1:00:53] because a local model can handle the

[1:00:55] voice to text can can can handle the

[1:00:58] natural language interaction just fine

[1:01:00] today. Uh

[1:01:02] >> there was that guy in chat who said they

[1:01:04] were they were using uh Fable to rewrite

[1:01:07] an entire system and it's like okay.

[1:01:09] >> Yeah. Um

[1:01:10] >> you're not doing that on your RTX 3070.

[1:01:14] >> Uh yeah, okay, sure. um you you you can

[1:01:16] you can do quite a bit with local models

[1:01:18] and and it might do pretty well. But

[1:01:20] >> yeah, Atomica says Quinn 3.6 is

[1:01:23] phenomenal

[1:01:24] >> and that happens to be uh that was

[1:01:26] actually one of the first things that

[1:01:29] Nick Harris asked when we were meeting

[1:01:31] with Nvidia at Computex was okay neat

[1:01:36] impressive demo you what are you guys

[1:01:39] using and Quen 3.6 came up like

[1:01:42] immediately. Uh it seems like it's just

[1:01:45] basically beloved right now for anyone

[1:01:48] who wants to run local AI.

[1:01:51] >> Yeah, it's uh there there's a lot of

[1:01:53] different things. If you're if you're

[1:01:55] scared of running stuff, again, I would

[1:01:56] I would highly recommend watching the

[1:01:58] PewDiePie video, but he talks about a a

[1:02:00] part of the Odysius system called

[1:02:01] cookbook, uh where it helps inform you

[1:02:05] based on your system hardware, uh what

[1:02:07] what you can run. So, if you're like,

[1:02:09] "What the heck is a what the heck is a

[1:02:11] Quen?" um it can it can kind of handle

[1:02:14] those portions for you or at least

[1:02:17] majorly simplify it. But yeah, the the

[1:02:19] local hosting stuff is was incredibly

[1:02:21] interesting and is now

[1:02:24] taken a massive leap forward in how

[1:02:26] interesting it is and the Frontier

[1:02:28] models keep being mind-bogglingly

[1:02:31] insane. Um and it's it's quite quite the

[1:02:35] time for uh terrible things to happen.

[1:02:39] Um,

[1:02:41] yeah. Anyways, topic move on.

[1:02:44] >> Oh, uh, no, it's time for us to do our

[1:02:46] creator warehouse announcements. This

[1:02:49] week, we are launching a pretty fun one.

[1:02:53] I am legitimately super excited about

[1:02:56] this one. Okay, hold on. Hold on. It's

[1:02:58] time to hopefully not have a classic

[1:03:01] Wancho nip slip here. Okay.

[1:03:05] Oh, shoot. I still have my mic on. Uh,

[1:03:07] okay. First up is the map of Siberia

[1:03:12] t-shirt. It is a fully handustrated

[1:03:16] fantasy map where someone replaced all

[1:03:20] the kingdoms with things like

[1:03:23] >> the great firewall

[1:03:25] is of when and the cryptocurrency.

[1:03:30] Get it? Cryptocurrency.

[1:03:34] >> That's pretty good.

[1:03:34] >> Hold on. Hold on. I'm going to go to the

[1:03:36] uh to the Lionus cam. Uh this is got to

[1:03:39] be Oh man, you guys can't Okay, you know

[1:03:42] what? I'm just going to have to I'm just

[1:03:43] going to have to pull it up on the site.

[1:03:45] Um

[1:03:47] >> uh yeah, I I'll bring it up on the site.

[1:03:48] I'll bring it up on the site. Uh I I

[1:03:51] have to give Lisa from the CW team like

[1:03:56] the vast majority of the credit on this

[1:03:59] one. this. She spearheaded this and just

[1:04:03] oh, I don't know when it happened, but

[1:04:09] she just like figured out our vibe as a

[1:04:14] company, our audience at some point.

[1:04:18] >> Pirate Bay

[1:04:19] >> and it culminated or maybe she's still

[1:04:22] cooking. Maybe she maybe she hasn't

[1:04:24] reached the peak yet, but oh my god, the

[1:04:28] leg access point. You have got to be

[1:04:31] kidding me.

[1:04:32] >> That is the funniest [ __ ]

[1:04:36] >> Control + C is really good. I like that

[1:04:38] one.

[1:04:39] >> Yeah,

[1:04:40] the poisoned cash

[1:04:42] >> port 80

[1:04:43] >> network bridge.

[1:04:47] Of course, this had to be there. This is

[1:04:50] This is This is quite informed actually.

[1:04:53] >> Right.

[1:04:55] >> The bits.

[1:04:56] >> That's cool.

[1:04:57] >> The bites.

[1:04:59] I asked her. I was like, "Are do the

[1:05:02] bites have exactly eight times the land

[1:05:04] mass of the bits?" And she was like,

[1:05:05] "Don't overthink it." I'm like, "Okay,

[1:05:07] all right. Fair fair enough, Lisa. All

[1:05:09] right. All right. I guess

[1:05:10] >> Hold on a second. I got

[1:05:11] >> the data

[1:05:12] >> 27032."

[1:05:16] >> Oh, interesting.

[1:05:17] >> Interesting. Don't don't don't worry

[1:05:18] about it. Don't worry about it. Um,

[1:05:21] absolutely

[1:05:23] absolutely incredible design. Super

[1:05:26] cool.

[1:05:27] >> Are these up right now?

[1:05:29] >> Yeah.

[1:05:30] >> Okay. I didn't see a link in the doc.

[1:05:32] >> Map of Siberia.

[1:05:33] >> Oh, I see. There it is.

[1:05:34] >> T-shirt available now in regular and

[1:05:37] tall. So, make sure you guys are getting

[1:05:39] in there. Oh, we don't have a lot of

[1:05:40] tall units in the uh the XL's, double

[1:05:43] XXLs, and triple XL's, at least on the

[1:05:45] global site. And great photo shoot as

[1:05:48] always. This has to be Tynen.

[1:05:51] There's no way this isn't Tynen. It's

[1:05:53] Tynen. No, it's Dan.

[1:05:55] >> That's me.

[1:05:56] >> Did Did you borrow T? Okay, this one's

[1:05:58] Titan.

[1:05:59] >> The ones outside are Tynen. Yeah.

[1:06:01] >> Okay.

[1:06:03] Amazing. I know he's into that stuff. I

[1:06:05] didn't know you were. Did you borrow his

[1:06:06] stuff?

[1:06:07] >> Uh, it was actually somebody else's

[1:06:08] stuff.

[1:06:09] >> Oh, really?

[1:06:09] >> There was some other guy who lent us a

[1:06:12] set of armor.

[1:06:13] >> Oh, that's super cool. Um,

[1:06:15] >> hold on one sec. Am I Am I getting this

[1:06:17] correctly? The olive one has the large

[1:06:20] map on the front,

[1:06:21] >> but the black one has the small logo on

[1:06:23] the front, the large map on the back.

[1:06:25] >> Yeah. So, we couldn't decide uh we

[1:06:28] couldn't decide whether to do olive or

[1:06:30] black or the here, if I switch over to

[1:06:33] the black, you'll see this one has the

[1:06:35] map on the back and then it has the uh

[1:06:39] just the cloud kingdom on the front. So,

[1:06:42] we kind of tried both and we'll kind of

[1:06:46] see what people are more into.

[1:06:49] I'm not going to spoil anything by

[1:06:50] saying who thought which way, but uh

[1:06:53] they're they're both really cool designs

[1:06:56] and

[1:06:56] >> I'm so interested how the sales of this

[1:06:59] go. I I'm so interested because I've

[1:07:01] always personally like I I think this

[1:07:04] shirt is wicked. I'm going to try, if

[1:07:06] there isn't one sitting there for me

[1:07:08] already, I'm going to try to get the the

[1:07:09] black one in XL tall. But, um, the

[1:07:12] reason why I'm going to get the black

[1:07:14] one is because I always love the, um,

[1:07:19] the small logo over the heart with Is

[1:07:24] that Is that my size?

[1:07:26] >> Sorry, it's a medium.

[1:07:28] >> Darn it. Uh, I always really like the

[1:07:30] small logo over the heart and then the

[1:07:31] big thing on the back of the shirt.

[1:07:33] That's always my favorite shirt design.

[1:07:36] But I'm wondering

[1:07:40] is that is that just a me thing or is

[1:07:42] that

[1:07:42] >> No, it's a lot of people thing. But we

[1:07:44] also felt that the way that this design

[1:07:47] doesn't pop as like Okay, here.

[1:07:50] >> No, I agree. That is a good point.

[1:07:52] >> Like this is a lot to wear on the front.

[1:07:56] This is actually not as much.

[1:07:59] >> Yeah. No, I I agree. It's very

[1:08:01] interesting. Yeah.

[1:08:02] >> And it's also not like a huge logo. Um

[1:08:07] if if it was a massive like your entire

[1:08:10] torso Linus Tech Tips logo.

[1:08:12] >> Yeah,

[1:08:13] >> that would it probably kind of suck.

[1:08:15] >> But instead, it's like a cool design

[1:08:17] full of fun little Easter eggs and stuff

[1:08:19] like it's

[1:08:20] >> and there's the you know there's the the

[1:08:22] shirt that must not be named uh that

[1:08:24] doesn't exist anymore. Um and that had a

[1:08:27] big logo on the front. Sleep is fories.

[1:08:30] >> No, it was released very recently and

[1:08:32] then taken away.

[1:08:34] >> Oh. Oh, that one.

[1:08:36] >> Yeah, that had a big logo on the front

[1:08:38] and that looked really sick as well. So,

[1:08:40] like it's obviously not a hard and fast

[1:08:42] rule.

[1:08:42] >> So, I am interested because just like

[1:08:44] you said on the olive one,

[1:08:46] >> it it does it kind of works. Um,

[1:08:49] >> and um,

[1:08:51] >> uh, well, one that I do have an XL tall

[1:08:53] here for you is this one was our best.

[1:08:56] >> See, that also works.

[1:08:58] >> This was our bestselling t-shirt in a

[1:09:00] long time. Um,

[1:09:02] >> so it's really not a hard and fast rule.

[1:09:04] It's just like a generalization that I

[1:09:06] prefer that setup.

[1:09:07] >> But like I wouldn't that pirate one

[1:09:10] >> if you just had a little parrot. It's

[1:09:12] not showing up on camera. If you just

[1:09:13] had a little parrot here and then the

[1:09:15] back said that,

[1:09:17] >> I don't think that would be better. Um,

[1:09:19] >> no, I don't think that one would quite

[1:09:20] work. So, this is like

[1:09:22] >> kind of depends.

[1:09:23] >> We've been kind of trying to find our

[1:09:25] stride um on like the t-shirt designs

[1:09:29] over the last little while. Um, like I

[1:09:33] think we had um I don't want to get too

[1:09:35] deep into like sort of the the internal

[1:09:40] stuff, but you guys might have noticed

[1:09:42] that whether it's sourcing the blanks or

[1:09:45] whether it's having designs ready or

[1:09:47] whether it's having both of those things

[1:09:49] at the same time, um, we we haven't

[1:09:52] really focused on t-shirts for like

[1:09:56] a couple of years at this point. Whereas

[1:09:59] it used to be that like every other

[1:10:00] freaking Wand show we were announcing

[1:10:02] another t-shirt and t-shirts were like a

[1:10:03] big part of LTT store and then for a

[1:10:06] long time they basically haven't been.

[1:10:08] But um I think over this last couple

[1:10:12] like I I obviously haven't seen any

[1:10:14] sales figures for this one yet but I

[1:10:16] love it. Uh the design team freaking

[1:10:18] loves it. Like we love it so we're

[1:10:20] really confident in it. But as you can

[1:10:22] see, we're still kind of mixing it up in

[1:10:24] terms of, okay, do we do little thing on

[1:10:26] the front, big thing on the back, big

[1:10:28] thing on the front? Like, what's the,

[1:10:30] you know, what's our what's our current

[1:10:31] meta going to be? So, we're going to

[1:10:33] we're going to let you guys

[1:10:34] >> ultimately decide that.

[1:10:36] >> Um,

[1:10:36] >> it's very interesting. It's I I'm I'm

[1:10:38] specifically interested in the sales of

[1:10:40] this one with my little I like the small

[1:10:42] logo thing. Um, because I think this one

[1:10:46] is a strong example of the big image on

[1:10:48] the front being cool. Yeah. Um,

[1:10:52] >> so it's it's it's like a a good test.

[1:10:54] Um, both are cool, I think.

[1:10:57] >> And but wait, there's more.

[1:11:01] >> Where is it? There's also Oh, our new

[1:11:03] arrivals could probably stand to be

[1:11:05] cleaned out a little bit. Whoops.

[1:11:07] >> A desk pad. Is that what you're talking

[1:11:08] about?

[1:11:08] >> Yeah. So, there's also a desk pad. How

[1:11:11] fun is that? Available in three

[1:11:15] different sizes. It's 900 mil by 400

[1:11:18] mil, 300 mil, or 600 mil depending on

[1:11:21] your preference. All the same price as

[1:11:23] you've come to expect from LTT store.

[1:11:26] Hello from LTT store. It also looks

[1:11:30] extremely striking. Great quality

[1:11:32] printing. Got this here. Map of Siberia

[1:11:36] desk pad. Oh, it's Luc is there. Oh,

[1:11:40] yeah. There you go. Freaking awesome.

[1:11:42] And of course, nice high quality

[1:11:44] deskpad. natural rubber stitched etches,

[1:11:46] etc., etc. Stitch etches. Stitched

[1:11:48] edges. What you would expect from us,

[1:11:52] man. They've been having way too much

[1:11:54] fun at these photo shoots lately.

[1:11:58] >> And that's a good thing. One of what?

[1:12:05] >> Oh, really? Unless it's not on there. I

[1:12:08] didn't see that one. Sorry, Dan. I

[1:12:11] didn't see it. Anyway, those are our new

[1:12:13] launches for this week and you can check

[1:12:16] them out at lmg.gg/cyiberia,

[1:12:19] which is spelled like cyber

[1:12:22] and then I A.

[1:12:25] Um, and it's a perfect time to pick one

[1:12:28] up because we're live, which means that

[1:12:30] you can send a comm.

[1:12:33] We might need to change the name again.

[1:12:35] Well, whatever. For now, they're called

[1:12:36] checkout messages coms. And the way that

[1:12:39] they work is all you got to do is fire

[1:12:40] up LTT store, add something that catches

[1:12:43] your eye to your cart, boop boop, view

[1:12:46] your cart, and in the checkout, you can

[1:12:48] type up a message. It'll go to producer

[1:12:50] Dan. There he is. Who will respond to it

[1:12:54] or just like pop it up on the stream if

[1:12:56] it's just like a shout out for someone

[1:12:57] or he will curate it for me and Luke to

[1:13:01] respond to. Uh Dan, do you want to hit

[1:13:04] us with a couple of coms? Sure thing.

[1:13:07] Lionus, your 40th level is coming up. Do

[1:13:10] you plan on celebrating it in a special

[1:13:11] way? How about writing a biography or

[1:13:13] even an autobiography? I'd read a

[1:13:16] success story. Also, more non-black

[1:13:19] t-shirts, please. Um, I'm flattered, but

[1:13:23] I

[1:13:25] am

[1:13:27] not planning to write an autobiography.

[1:13:31] I I started making notes like a couple

[1:13:33] times for something like that and then I

[1:13:36] just have not had the time to do it

[1:13:39] slashh

[1:13:44] I don't know I don't maybe maybe I've

[1:13:45] just been maybe I've just not thought

[1:13:47] about it the right way cuz I was I was

[1:13:49] really surprised when I found out that

[1:13:50] someone as private as Linus Torvalds had

[1:13:53] written an autobiography and when I

[1:13:56] asked him about it um he basically was

[1:13:58] just like well try anything once

[1:13:59] including

[1:14:03] write an autobiography apparently cuz

[1:14:06] he's a really modest guy and that didn't

[1:14:08] really he's a really modest really

[1:14:10] private guy and that didn't really align

[1:14:12] with write an autobiography

[1:14:15] but the writer actually talks about this

[1:14:18] in the autobiography um or I guess it's

[1:14:21] not really an autobiography but it's

[1:14:22] like it's got a lot of autobiographical

[1:14:26] elements but also a lot of observations

[1:14:29] from the author. It's kind of like a a

[1:14:30] mish mash of both. Um, and the writer

[1:14:33] actually talks about the process by

[1:14:34] which he convinced Torvalds to do it.

[1:14:37] Um, so maybe I don't know, maybe I just

[1:14:39] haven't heard the right pitch from a

[1:14:40] ghost writer. I also just don't think

[1:14:41] I'd be comfortable with a ghostriter

[1:14:43] writing it. I feel like I should write

[1:14:44] my own thing. I literally people ask me

[1:14:46] like, "What do you do?" And more than

[1:14:48] anything else, I think more than a host,

[1:14:50] um, I I think my greatest contribution

[1:14:53] to LTT is is writing. Uh, I don't think

[1:14:56] people realize how much of what is done

[1:15:01] at Lionus Media Group has been touched

[1:15:03] at some point by my pen. Um, not all of

[1:15:07] it.

[1:15:08] >> It's all of it.

[1:15:09] >> It's not all of it.

[1:15:10] >> Every single mistake.

[1:15:11] >> Thank you.

[1:15:12] >> It's all Lionus.

[1:15:13] >> Thank you for that.

[1:15:14] >> Even if it's just a Labs article that

[1:15:16] doesn't have him credited, if there's a

[1:15:17] mistake in it, it was Lionus. If there's

[1:15:19] not a mistake in, it was someone else.

[1:15:24] All right.

[1:15:26] >> Anyways, uh

[1:15:28] >> did I answer the question? I don't

[1:15:29] remember and I I'm having a hard time

[1:15:31] logging into the dashboard.

[1:15:32] >> No, I think that's probably good enough.

[1:15:34] >> All right, cool. Good chat.

[1:15:35] >> Hey, DL. What's the most recent book

[1:15:37] you've read or are currently reading?

[1:15:40] >> Oh,

[1:15:44] um what did I pick up recently?

[1:15:47] >> Trying to remember. I read Breakfast

[1:15:48] with my a little while ago, but that was

[1:15:50] a while ago. I think I read something

[1:15:52] more recently than that.

[1:15:56] >> I'm second here.

[1:15:58] >> I've I've reverted to my teenage state

[1:16:01] and I've been uh I've been reading old

[1:16:03] Foxtrot collections during my midnight

[1:16:06] snack breakfast cereal lately. So,

[1:16:08] that's been that's been fun. I love

[1:16:10] Foxtrot.

[1:16:11] >> Fun.

[1:16:12] >> I actually had no idea he still

[1:16:14] publishes Sunday strips online. Shout

[1:16:17] out Bill.

[1:16:21] Um, man, there's no way to go sort by

[1:16:24] recent. My goodness.

[1:16:31] I will find it. Give me a second.

[1:16:33] >> He's looking up Dan in the meantime.

[1:16:35] Read any good books lately?

[1:16:37] >> I've still got most of The Prince to

[1:16:40] finish off.

[1:16:44] >> Um,

[1:16:45] >> what?

[1:16:45] >> I read it back in college and it was

[1:16:47] like

[1:16:48] >> it was brutal and it's still brutal.

[1:16:50] reminds me a lot of the world right now.

[1:16:53] >> Have you um Have you read it, Linus?

[1:16:55] >> No, I haven't.

[1:16:56] >> Yeah, you can have my other copy if you

[1:16:58] want. It's um it's bad. It's hard. It's

[1:17:01] hard to read,

[1:17:02] >> huh? What a what a sales pitch.

[1:17:05] >> It should be read, I think, by a lot of

[1:17:07] people but

[1:17:08] >> Okay, that's a better pitch. You should

[1:17:09] lead with that.

[1:17:10] >> The prince is good.

[1:17:15] Um,

[1:17:19] >> holy crap. You know what, Lionus? I can

[1:17:21] give you something fun. Oh, there it is.

[1:17:24] Um, Tough and Competent by Eugene F.

[1:17:27] Crayons.

[1:17:28] >> Tough and Competent.

[1:17:30] >> Is that Is is that your autobiography?

[1:17:34] >> No, but thank you. Um,

[1:17:37] yeah. Uh Jean CR was uh the second uh

[1:17:43] chief flight director for NASA.

[1:17:45] >> He looks tough and competent.

[1:17:47] >> Yeah.

[1:17:50] Yeah. Yeah. Um but I I haven't actually

[1:17:54] started it. Uh but I have it and I'm

[1:17:57] excited.

[1:17:57] >> That's the first step.

[1:18:00] >> One small step for Luke's

[1:18:03] reading list.

[1:18:06] >> Yeah. Um, but okay, this is okay. I

[1:18:09] guess is that that question done?

[1:18:13] >> I'm excited to show you something. And I

[1:18:14] keep trying to jump to it and then

[1:18:15] realizing that I'm skipping the topic

[1:18:17] we're currently on. Sorry.

[1:18:20] >> Uh, this is not an ad. I don't know if

[1:18:22] these guys have ever sponsored us or

[1:18:24] not. So, my bad if they have, I don't

[1:18:26] know, whatever. But, um, and I can't

[1:18:28] send you a link to this super easily,

[1:18:30] but if you just want to Google it,

[1:18:32] you've probably already seen it. But um

[1:18:34] 8bit do ultimate three mode controller

[1:18:37] Xbox. They have a 8bit do Xbox

[1:18:40] controller that is translucent green.

[1:18:42] Have you seen this?

[1:18:44] >> Uh I thought you would like it.

[1:18:47] >> Three mode. Oh, nice. That looks sick.

[1:18:50] >> Yeah. Yeah. It's really interesting.

[1:18:53] There's a bunch of I I don't know if

[1:18:56] everyone just had the same thought at

[1:18:57] the same time or what, but there's a

[1:18:59] bunch of translucent products coming

[1:19:00] out. You can see in the background

[1:19:02] there, they have a keyboard. Um, yeah,

[1:19:04] they have an 85% keyboard and a separate

[1:19:07] 10 key, which I think is super cool.

[1:19:09] Those buttons tend to come with their

[1:19:10] keyboard. Um, so I'm assuming that's the

[1:19:13] case here as well. And then they also

[1:19:14] have the mouse. And then they also have

[1:19:16] the uh the Xbox controller. It's a sick

[1:19:19] lineup of hardware actually.

[1:19:21] >> That looks amazing. Officially licensed

[1:19:23] by Xbox.

[1:19:24] >> Awesome.

[1:19:25] >> Supported by Apple, man. 8bit Dough has

[1:19:27] just been crushed.

[1:19:29] >> I believe they're Hall effect joysticks

[1:19:32] or or or something. Um I lost the

[1:19:36] >> Hall effect joysticks. Yep.

[1:19:38] >> Yeah.

[1:19:38] >> And the triggers are Hall effect linear.

[1:19:41] >> Oh, nice. Okay. The triggers I wasn't

[1:19:43] sure, but I was trying to find that.

[1:19:44] Cool. Um and there's a switch on the

[1:19:46] back to switch which kind of device

[1:19:47] you're connecting it to and stuff. This

[1:19:49] looks wicked. Um I I just I I don't

[1:19:53] know. I'm just so stoked on translucent

[1:19:55] products coming out. Um, I think it's

[1:19:58] awesome. I love I'm so happy that it

[1:20:02] seems like we're exiting an era of

[1:20:04] everything just being either all black

[1:20:06] or all white. Um, I'm so happy that some

[1:20:09] color is entering products again and

[1:20:11] that translucent and seeing the

[1:20:13] interesting things inside the product is

[1:20:16] becoming a thing again. I just think

[1:20:17] it's so sick. Oh, dude. I um Okay, I had

[1:20:22] a similar experience today. Uh I was

[1:20:25] working on the LTT for uh Sony sent over

[1:20:28] their 115in true RGB TV. So, it's uh

[1:20:32] it's over in the theater room right now.

[1:20:34] I I had to leave for WAN show, so I

[1:20:36] didn't actually get to see it fire up

[1:20:39] and watch content on it. I have to do

[1:20:41] that on Monday. I have to wait the whole

[1:20:42] weekend before I ah Anyway, that's not

[1:20:44] the point. The point is, I was looking

[1:20:46] at the remote and it's a black remote

[1:20:50] and I was like, "Of course it's a

[1:20:51] black."

[1:20:53] No wait.

[1:20:54] >> Translucent.

[1:20:55] >> It No, no, it doesn't. It has It had

[1:20:57] little uh like teal like flexcks in the

[1:21:02] plastic like Do you remember our uh like

[1:21:04] our RGB shirts how they had those little

[1:21:07] like like uh

[1:21:08] >> like those little color tufts in them?

[1:21:12] >> Yeah.

[1:21:12] >> Yeah. It was cut. It was all one color,

[1:21:15] but it was kind of like that. Like it

[1:21:16] just had little color flexcks in it. And

[1:21:18] I was like, how fun is that? And how

[1:21:22] hard was that, you know, to bring back a

[1:21:25] little bit of

[1:21:27] color in our lives? I I just I thought

[1:21:29] it was really cool. I thought it was a

[1:21:30] nice little touch.

[1:21:33] >> Yeah. I I think that might actually be

[1:21:36] um a double benefit because I believe

[1:21:38] that's often an appearance that comes

[1:21:40] from recycled plastic.

[1:21:42] >> Yeah.

[1:21:43] Um, so it could be a recycled plastic.

[1:21:46] >> It looks great.

[1:21:47] >> Oh, I I think it looks fantastic. I

[1:21:49] think it looks fantastic. Also, I didn't

[1:21:51] know this was a thing, but just to to

[1:21:54] be, you know, fair. Um, and I'm doing a

[1:21:56] terrible job with tracking names today

[1:21:58] by East One. Um, or Eastone

[1:22:02] um posted in Flipline Chat. Xbox is also

[1:22:05] releasing their own translucent green

[1:22:08] Xbox controller. If you look up the Xbox

[1:22:11] 25th anniversary, um they have an Xbox

[1:22:13] wireless controller X25 special edition

[1:22:16] and it's a translucent translucent green

[1:22:19] Xbox controller. Um

[1:22:22] >> I'm a little confused about some of the

[1:22:24] messaging that I hear coming out of Xbox

[1:22:26] right now. Like on the one hand, they

[1:22:28] need to rejuvenate their hardware

[1:22:30] business, but then on the other hand,

[1:22:31] they're saying things like how many

[1:22:33] Xboxes they sell is gated by like

[1:22:36] production capacity, not by like demand.

[1:22:39] I just I'm I'm having a really hard time

[1:22:41] reconciling that. I'm also I am also

[1:22:44] recognizing that a lot of Xbox fans are

[1:22:48] rooting for exclusives,

[1:22:51] but I can't help but feel like if they

[1:22:54] go full exclusive and they don't come to

[1:22:57] PC that we are like we're sitting here

[1:23:00] demanding reverse progress that is

[1:23:03] actually regression.

[1:23:05] Um, as someone who has never bought an

[1:23:08] Xbox and who probably will never bother

[1:23:11] to game on an Xbox, um, making games not

[1:23:14] available for me to purchase on PC is

[1:23:16] something that I'm going to have a hard

[1:23:17] time rooting for. Uh,

[1:23:20] >> the Xbox install base being as it is,

[1:23:23] games being exclusive might just mean

[1:23:26] their death.

[1:23:27] Yeah,

[1:23:29] but I also understand that like there's

[1:23:32] it's a chicken and egg problem because

[1:23:34] you're not going to get an install base

[1:23:36] unless you give people a reason to buy

[1:23:38] the console and it sure as heck isn't

[1:23:39] going to be the Xbox first party

[1:23:42] controller. Like what's it going to be?

[1:23:44] >> I don't feel like it's going to be Gears

[1:23:47] 6 either. Um

[1:23:50] or or whatever.

[1:23:52] >> I mean, you weren't really like a Gears

[1:23:53] kid though, right?

[1:23:55] >> Not really. But

[1:23:58] I feel I feel like Ooh, is this hot? Is

[1:24:01] this a hot take?

[1:24:02] >> Here comes the hot take.

[1:24:03] >> I feel like Gears always played second

[1:24:05] fiddle to Halo.

[1:24:06] >> Well, yeah. I don't think that's a hot

[1:24:08] take at all, sir.

[1:24:09] >> And I don't feel like either series is

[1:24:11] hot right now. So, what uh Halo Infinite

[1:24:15] the second and Gears 6? Like, I don't

[1:24:20] know. uh whatever whatever that thing

[1:24:23] they're doing is with

[1:24:26] uh

[1:24:28] the the the first Halo that they're

[1:24:30] recreating and they're making new

[1:24:31] missions.

[1:24:32] >> So

[1:24:32] >> yeah, um I don't think that's going to

[1:24:34] sell consoles either.

[1:24:36] >> I was chatting with Poof about it

[1:24:38] actually right before the show and he

[1:24:41] was like super bearish on it and then

[1:24:44] just saw the latest updates and he was

[1:24:47] like, "Yeah, I'm psyched now.

[1:24:51] I haven't actually I haven't seen it,

[1:24:53] but apparently it looks great

[1:24:56] >> like visually,

[1:24:57] >> but then also, didn't they have one of

[1:24:59] the old Bungie guys come out and pretty

[1:25:01] much say like, yeah, it looks great, but

[1:25:04] it doesn't look like Halo Combat

[1:25:06] Evolved. So,

[1:25:08] I'm and I'm paraphrasing, I'm basically

[1:25:10] an AI right now. Disclaimer, what I said

[1:25:12] might not be attributable to anyone.

[1:25:14] >> I'm basically an AI right now is a

[1:25:16] hilarious line. Um,

[1:25:19] the three new missions is interesting

[1:25:21] and stuff. I

[1:25:21] >> don't trust me, bro.

[1:25:23] >> All I'm saying I'm not saying it doesn't

[1:25:24] look interesting. All I'm saying is I

[1:25:26] don't know if it's a console seller. Um,

[1:25:32] I don't know what Microsoft has right

[1:25:34] now that would be a console seller.

[1:25:37] People really like Fors Horizon 6.

[1:25:44] >> Okay.

[1:25:45] >> Yeah. They also dropped Japan, which has

[1:25:47] been a location that people have been

[1:25:49] looking for in Fors Horizon for a long

[1:25:50] time. So, they just dropped their like

[1:25:52] ace up their sleeve

[1:25:54] >> before it could be an exclusive. Um,

[1:25:58] I just spent, you know,

[1:26:01] days driving switchbacks in Taiwan. They

[1:26:04] could do one in Taiwan, but I I don't

[1:26:05] think that's going to be as big of a

[1:26:06] draw as Japan is for car culture,

[1:26:09] obviously. Um,

[1:26:14] I don't know. I don't I don't know what

[1:26:15] game

[1:26:17] Microsoft has up their sleeve. Oh, yeah,

[1:26:20] I do. Okay.

[1:26:24] I mean, if some new Xbox launches and

[1:26:27] soon after the launch, Elder Scrolls 6

[1:26:29] comes out and it's exclusive to

[1:26:32] >> consoles, that'll move some consoles.

[1:26:35] >> That'll move some cons. Well, assuming

[1:26:37] it's not a

[1:26:39] complete dumpster fire, it'll move

[1:26:41] >> I think it'll still move some consoles,

[1:26:43] honestly. It's just a question of how

[1:26:45] many sir.

[1:26:46] >> Yeah, totally. Yeah, exactly. Um, don't

[1:26:49] curse us like that. Yeah, sorry. That's

[1:26:51] a good point, but sorry. Um, happy 8th

[1:26:54] birthday to the trailer of ES6. I didn't

[1:26:58] believe that trailer for so long. I

[1:27:00] thought it was fake. I thought it was a

[1:27:02] fan edit. Uh, I didn't watch that game

[1:27:05] show, whatever one it was announced in.

[1:27:07] So, when I saw it go on YouTube, I I

[1:27:09] legitimately thought it was just like a

[1:27:11] fan edit. um and didn't believe it for

[1:27:14] so long and then people were talking

[1:27:15] about it like I think years later and I

[1:27:19] was like wait what that was real

[1:27:22] um that was wild. I can't believe that I

[1:27:25] still can't believe they did that. Um

[1:27:28] yeah I I don't think Oh man I I'm kind

[1:27:33] of strongly of the opinion that it's

[1:27:34] been it's been downhill since Morwind.

[1:27:37] Um, but the peak of Morwind was so high

[1:27:40] that the slope was still really, really

[1:27:42] good. Um, but like Fallout 4,

[1:27:47] even compared to Skyrim,

[1:27:50] even compared to Oblivion, I think was

[1:27:52] like a continual

[1:27:55] down. Fallout 4 was still really fun. It

[1:27:57] was a great game, but it felt like the

[1:27:58] worst one out of those. I feel like as

[1:28:01] someone who came into the series on

[1:28:03] Oblivion,

[1:28:04] >> it's tough and then tried to play

[1:28:06] Mororrowind because Oblivion was so

[1:28:09] great and I wanted more.

[1:28:11] >> You can't go back. You You have to look

[1:28:13] at Morowind in the time capsule of when

[1:28:15] it was released.

[1:28:16] >> Sure.

[1:28:16] >> Um I'm not I'm not suggesting you go

[1:28:18] back and play it now, but when it was

[1:28:20] released, Morowind was an insane game.

[1:28:22] Like there was

[1:28:24] >> there was nothing like it. Um,

[1:28:27] it it was wild. Um, and the the the way

[1:28:31] that I'm describing this like downward

[1:28:33] slope is basically like

[1:28:36] um like the weapon systems were reduced

[1:28:38] for Oblivion. The the you'd be you might

[1:28:41] be surprised having played Oblivion, but

[1:28:42] the magic system was was reduced for

[1:28:45] Oblivion. Um things just got simpler and

[1:28:47] simpler and simpler. Oblivion was a

[1:28:49] fantastic game. I think I think in a lot

[1:28:51] of respects you can say it was better

[1:28:52] than Morwind but um

[1:28:56] yeah under understanding the climate in

[1:28:58] which Morwind released I I still think

[1:29:01] Morowind's better if that makes sense.

[1:29:03] Um

[1:29:06] anyways and and Skyrim was a like

[1:29:08] massive simplification of things as

[1:29:10] well. Um

[1:29:13] I don't know. I can understand people

[1:29:14] might not agree with that argument, but

[1:29:16] I do think especially beyond Skyrim,

[1:29:19] Fallout 4 and

[1:29:22] Fallout

[1:29:24] 76, I think is what it's called. Um, and

[1:29:29] Starfields and some of the mobile games

[1:29:31] that they did and things like that. It's

[1:29:35] been

[1:29:38] I don't know. I still don't think Elder

[1:29:40] Scrolls 6 and I I talked big game about

[1:29:42] how I thought Starfield was going to be

[1:29:43] great. So, hey, maybe I'm going to be

[1:29:45] terribly wrong again. Um, but I still

[1:29:48] don't think

[1:29:48] >> Can you still call yourself a Bethesda

[1:29:50] fan?

[1:29:57] Do you can you be fully rational and

[1:30:01] logical about Bethesda or will you still

[1:30:04] have naive hope in spite of all evidence

[1:30:07] from the last 10 years

[1:30:11] cuz I think that's the line right is is

[1:30:13] >> that's tough though cuz like I did enjoy

[1:30:15] Fallout 4. Um, I thought it was a lot

[1:30:20] weaker than New Vegas and I thought it

[1:30:22] was weaker than three again.

[1:30:24] >> So, note that I said 10 years that it

[1:30:26] launched,

[1:30:27] >> Luke, I said 10 years.

[1:30:28] >> I'm trying to remember.

[1:30:29] >> Fallout 4 came out 11 years ago.

[1:30:33] >> Jesus.

[1:30:34] >> No, it didn't.

[1:30:35] >> Yes, it

[1:30:37] >> actually 10 and a half years.

[1:30:38] >> No, it didn't. Shut up.

[1:30:41] >> Holy crap. That

[1:30:43] >> I just broke him. I broke him down. I'm

[1:30:46] sorry.

[1:30:48] >> You want to know something funny? You

[1:30:49] know how I've like taken a break from

[1:30:51] streaming for a long time?

[1:30:53] >> I went through my most recent streaming

[1:30:56] assets folder.

[1:30:58] It's all Fallout 4 stuff. My taking a

[1:31:02] break has been a decade. Oh my god.

[1:31:06] Oh my god.

[1:31:07] >> He's dying, Dan. Holy crap.

[1:31:14] Okay. Um,

[1:31:17] >> oh man. Should I break him again?

[1:31:21] >> I How old is Mororrowind?

[1:31:23] >> Oblivion. I know the Xbox.

[1:31:24] >> Oblivion was 20 years ago.

[1:31:27] >> Yeah. Yeah. Yeah. Cuz Xbox 360 like I

[1:31:29] knew that one. For some reason, the gap

[1:31:31] between

[1:31:32] >> Oblivion and Fallout 4 seems so much

[1:31:35] bigger than between Fallout 4 and now. I

[1:31:37] don't know why that's true, but it feels

[1:31:39] like it.

[1:31:40] >> Um, okay. So the the question of the

[1:31:43] last 10 years thing is is tough.

[1:31:46] >> Yeah, it's tough.

[1:31:48] >> Even even if we look like locally,

[1:31:50] right? If we look at Landis Media Group,

[1:31:52] like Landis Media Group looking five

[1:31:55] years back is so much different in what

[1:31:57] we're doing, who works here,

[1:31:59] >> our capabilities are, all that kind so

[1:32:01] different. Go back go back seven years,

[1:32:05] >> completely different thing. Um,

[1:32:08] so it's okay. It's been an incredibly

[1:32:12] long time since Skyrim, but the only

[1:32:14] mainline game that they've released in

[1:32:16] that time. I don't I don't necessarily

[1:32:20] think I believe 76 is a mainline game.

[1:32:24] Is that fair to say?

[1:32:29] I don't know.

[1:32:31] >> This was meant to be kind of a yes or no

[1:32:33] answer Luke.

[1:32:34] >> Yeah. Sorry.

[1:32:35] >> Are you still a fan of Bethesda?

[1:32:37] Well, I I what I think I'm saying is I

[1:32:39] can't conclusively say either way

[1:32:42] because they haven't really released

[1:32:44] anything in the series that I liked in

[1:32:46] that period of time. They haven't

[1:32:47] released a mainline Fallout game. They

[1:32:49] haven't released a mainline Elder

[1:32:51] Scrolls game. Those are the two games

[1:32:52] that made me love Bethesda. So, I don't

[1:32:54] think I can say

[1:32:56] >> I I'm I'm going to still naively hold

[1:32:58] out hope though. Okay, then it sounds

[1:33:00] like you can still be at least a little

[1:33:02] bit of a fan

[1:33:03] >> cuz if you can naively hold out hope for

[1:33:06] something. I think

[1:33:07] >> I am reviews and buying ES6 day one

[1:33:10] guaranteed cuz I I have to experience

[1:33:13] it. I must there's no other like

[1:33:16] >> I think that's the answer. I think

[1:33:18] that's the answer then.

[1:33:20] >> Um I want to jump into a couple sponsor

[1:33:23] spots so that uh because I'm sure you

[1:33:24] want to get on with your day at some

[1:33:26] point. Uh, the show is brought to you

[1:33:27] today by ODO. Having to cycle between a

[1:33:30] dozen different tabs and apps when

[1:33:32] operating a business can have you

[1:33:33] wanting to jump into a bottomless pit of

[1:33:35] despair, but not with our sponsor, ODU.

[1:33:38] ODU is the all-in-one business

[1:33:39] management software that has just about

[1:33:41] any app your business could need

[1:33:42] integrated into one platform. You and

[1:33:45] your team can get projects off the

[1:33:46] ground with ODO's designated app, so

[1:33:48] everyone can stay organized and on the

[1:33:50] same page. Once your project gets moving

[1:33:52] along, their CRM tool can help you send

[1:33:54] quotes and schedule meetings in a simple

[1:33:56] drag and drop interface. And the

[1:33:58] invoicing app can help you get paid when

[1:34:00] the time comes with things like sending

[1:34:02] invoices and tracking open accounts. And

[1:34:04] they can be as automated as you'd like.

[1:34:06] Something that makes ODO even cooler is

[1:34:08] that if you only need a single app, you

[1:34:10] can just use it for free. Use our link

[1:34:14] for a 15-day trial with no credit card

[1:34:16] required or book a demo with their

[1:34:18] expert team to learn how ODU can help

[1:34:20] your business grow. The show is also

[1:34:22] brought to you by Op manager Nexus. Op

[1:34:25] manager Nexus is a full stack

[1:34:26] observability platform that gives you

[1:34:28] and your IT team a single clear view of

[1:34:31] your entire infrastructure. There's no

[1:34:33] need to start from scratch. It

[1:34:34] seamlessly plugs into the manage engine

[1:34:36] ecosystem and the tools that your team

[1:34:38] already relies on. Whether your org

[1:34:40] works from the cloud or prefers an

[1:34:42] on-remise setup or a mix of the two, Op

[1:34:44] manager is designed to work the way you

[1:34:46] prefer. It's built to be deployed

[1:34:48] quickly and uses customizable dashboards

[1:34:50] so each member of your team gets the

[1:34:52] reports and information they need the

[1:34:53] most. Plus, with a global network of

[1:34:55] localized data centers, everything stays

[1:34:58] highly available because they know just

[1:35:00] how much reliability matters. It's

[1:35:02] backed with strong data encryption

[1:35:04] adhering to industry and regional

[1:35:06] compliance rules so your data stays

[1:35:07] protected. Op manager Nexus, stop

[1:35:09] guessing and start knowing with full

[1:35:11] stack visibility. We'll have a link for

[1:35:13] you down in the video description.

[1:35:15] >> All right, Luke, want to pick a topic or

[1:35:17] two?

[1:35:19] >> Uh, yeah, sorry. I was talking to Chad

[1:35:21] about Bethesda things.

[1:35:22] >> Oh my god.

[1:35:22] >> Um,

[1:35:23] >> definitely fan. Still a fan.

[1:35:25] >> You got me started, bro. You got me

[1:35:26] started. It's not my fault. Um,

[1:35:29] >> yeah, let's do this one. We were talking

[1:35:32] about AI stuff before we did the the

[1:35:35] jump off topic section, so we'll go back

[1:35:37] to it for some reason. Um, New York

[1:35:40] State could ban data center development.

[1:35:44] New York State has put a temporary hold

[1:35:46] on any further data center development.

[1:35:48] Governor Kathy

[1:35:51] Ochel uh will have the final say and

[1:35:55] must decide before December if the state

[1:35:57] will enact a permanent ban on all data

[1:36:00] centers that demand 20 megawws or more

[1:36:03] of power. Uh as as a quote uh this is

[1:36:06] one of the first times that we're really

[1:36:08] drawing a line in the sand and saying

[1:36:10] that as a state legislature we have the

[1:36:13] responsibility to make sure New Yorkers

[1:36:15] are in the driver's seat. uh said New

[1:36:18] York Senator Kristen Gonzalez. And then

[1:36:21] again to quote, "Big tech has been used

[1:36:25] uh has been used to writing their own

[1:36:26] rules or not having rules that they have

[1:36:29] to play by, which is yeah, I mean, fair

[1:36:31] enough." Um

[1:36:36] I feel like Yeah, it's interesting. I

[1:36:38] wonder what the

[1:36:42] >> just for context a 20 megawatt data

[1:36:44] center would be approximately 133 SpaceX

[1:36:49] AI satellite data centers.

[1:36:53] >> H

[1:36:55] just throwing that out there.

[1:36:59] >> That is pretty fun for context actually.

[1:37:02] Um,

[1:37:03] >> in case anyone was wondering why some

[1:37:07] people are pushing back on the idea of

[1:37:09] AI satellite data centers, um, it would

[1:37:13] take 133 of those to build something

[1:37:16] that's big enough that the state of New

[1:37:17] York, it wants to ban it. Wants to ban

[1:37:20] it.

[1:37:22] Not the orbit of New York. Um,

[1:37:27] yeah. Interesting. I'm trying to rack my

[1:37:30] brain. What are the primary complaints

[1:37:33] about data centers being near you?

[1:37:35] There's power obviously. There's water

[1:37:38] >> that it's pulling out.

[1:37:40] >> Ugly.

[1:37:40] >> Noise. I guess some people have

[1:37:42] mentioned the really high pitch noise as

[1:37:43] well.

[1:37:44] >> Ugly. Y

[1:37:45] >> I'm sure there's like ecological

[1:37:47] destruction

[1:37:47] >> cost. Uh so that's related to the power,

[1:37:50] but also

[1:37:51] >> the power cost thing. The power cost

[1:37:53] thing feels solvable

[1:37:55] >> um

[1:37:55] >> by making the power for the data center.

[1:37:58] I think part of it too much is like the

[1:38:01] economic argument for data centers is

[1:38:03] all the the jobs that they create, but

[1:38:05] that's only while they're building them.

[1:38:07] >> Um, once they're built,

[1:38:09] >> yeah, they don't have much afterwards.

[1:38:10] >> Modern data centers require very little

[1:38:12] in the way of personnel compared to

[1:38:14] while they are being constructed.

[1:38:17] >> Um,

[1:38:17] >> totally. So, it's one of those things

[1:38:19] where when they talk about all the jobs

[1:38:20] that like AI infrastructure is creating,

[1:38:22] it's like, yeah, as long as we're

[1:38:25] chasing infinite growth, then it will,

[1:38:27] you know, continue to support those

[1:38:30] jobs. But the second that there's a

[1:38:32] cutback or bubble burst or the the

[1:38:34] buildout is stalled, then they kind of

[1:38:36] go away.

[1:38:42] >> And it's it's worth noting actually. I

[1:38:44] think there's it's man it's it's kind of

[1:38:48] funny to kind of I I don't know I don't

[1:38:51] have uh I don't think I've taken a

[1:38:53] strong position either way on the the

[1:38:55] building out of data centers and it's

[1:38:58] really it's really interesting to see

[1:39:00] both the good points that that both

[1:39:02] sides of this issue make and also the

[1:39:05] somewhat seemingly flawed points that

[1:39:08] both sides make. Like there's uh there's

[1:39:10] a lot of talk about the water

[1:39:12] consumption of data centers. Um and

[1:39:15] during construction that is that is like

[1:39:18] super true. They use a ton of water. Um

[1:39:22] but a big part of that is that it takes

[1:39:24] a ton of water to like make concrete

[1:39:27] construction. Um

[1:39:30] like back when um uh back when I toured

[1:39:33] u Intel's FAB, they were they were

[1:39:35] showing me the site where they were they

[1:39:38] were building like an extension to the

[1:39:39] fab and they literally like this the

[1:39:42] scale of these kinds of projects. They

[1:39:44] had built their own concrete production

[1:39:46] facility on site so that they could so

[1:39:50] they could mix and then logistically

[1:39:54] move around the concrete much more

[1:39:56] efficiently. And like people don't

[1:39:58] necessarily recognize like how much