Episode 592: Don't Think for Me의 전체 녹취록

like, in America, they sell straight up drugs at the store. He's just, and of course, what he was talking about was Nyquil or whatever, right? The as one comedian put it, I remember on Comedy Central the this the the snuffly, sneezy, you know, how the hell did I end up on the floor medicine? Matt Ray (02:27.33) I miss that. Matt Ray (02:47.137) Exactly. Coté (02:47.215) And as I don't know, maybe listeners can tell I'm little stuffed up. And there's no NyQuil here in the Netherlands. And I just feel like kind of like it's our position here in the I'm gonna call it from one of the numerous departments we have in the software defined talk conglomerate, from the tourism board consultancy. that another thing, you know, while we're exporting Thanksgiving to the rest of the world, again, for those who are joining us. Matt Ray (03:15.5) And Halloween. Coté (03:17.075) Yeah, well how yes. I think I think Halloween might be a case study that that we can have. but you know, Thanksgiving to recap, it's non-religious, unless you want to inject religion. It really is not associated with anything historical. Sure, you know, if you really care, it actually emerged from after the Civil War and something, and then they were like, Whoa, that's kind of fucking grim. Why don't we make it about Turkey? So everyone gets two days off for a four day weekend, you get to eat stuff, you can watch whatever football means in your culture. Matt Ray (03:21.475) Ha ha. Matt Ray (03:33.902) Plymouth Rock. Coté (03:46.782) and and enjoy that, why not have Thanksgiving? It's it's it's great. so I think associated with that, we need to go to what I'm sure is a fantastically functioning currently American outreach tourism board and say, next up, NyQuil, sponsored by NyQuil. And what we're gonna do, we're just gonna start, you know, in the same way that you might be going around and someone's gonna offer you a Red Bull or at the airport, they they want to offer you like some face cream. Matt Ray (04:03.618) Ha ha. Coté (04:15.175) As you're just trying to get to your your Uber, we're just gonna be handing out NyQuil. And we're gonna be like, Hey, welcome to the rest of your life. Matt Ray (04:19.438) Check this out. Yeah. Yeah. Brandon (04:24.061) So you have to educate me. I was under the impression that the United States was like we were more restrictive of our over the counter drugs. Cause last time I was in Mexico, people were like, in Mexico, they were pointing this out like in Mexico, you can like buy whatever what we would consider prescription drugs just over the counter there. And then you being in Amsterdam, not to like, you know, be all stereotypical here, I felt like, well, Amsterdam's like pretty lenient about stuff. So, but no, am I have this wrong? Coté (04:30.261) Mm. Matt Ray (04:31.362) You mean y you Coté (04:33.725) Well, okay. Okay. Coté (04:41.471) Sure. Yeah, yeah, yeah. Coté (04:48.647) Yeah, and and so it it is it is that that's what I wanted to add, because

I know the situation back in back in Texas there, but I wanted to ask Matt Ray this is it is you know, so so my my friend Orin, they're they're he from London and Israel and wherever, you know, one of these these cosmopolitan people scurrying around everywhere. so that was a new thing to his family, the Nyquil. And then here in the Netherlands, you know, like all medicine in the Netherlands, the prescription is two part. One, paracetamol. Whatever, whatever your problem is, have paracetamol and just like just bike a little slower on your way home. I think, I think is the the second part of the prescription. but and I don't think I don't think they have it in the UK or like anywhere else. I think it's more or less like like an American type of thing. And so that's what I wanted to know, Matt Ray. I feel like I I feel like I'm gonna get let down here, but I feel like of all places the Australians would be like Matt Ray (05:27.406) Take it easy on your healthy living. Coté (05:46.665) W why the fuck you even asking this? Of course we have that, but I don't know. No. Ugh. Matt Ray (05:50.872) No, no. What was what was interesting was was was when I moved here, you could get over-the-counter codine. Which is, you know, it it it's fairly strong. I think they've moved away from that. don't really get that anymore. But but as you noted, like they don't have NyQuil. Their cold medicines are are pretty weak here. And so literally every time I go, yo, don't don't tell Australian Border Patrol, but Coté (05:58.941) Mm. Coté (06:05.021) Yeah. Matt Ray (06:20.201) Every time I come back from the US I'm bringing two jugs of NyQuil, right? It's just like you gotta have that stuff on stack on stock. Coté (06:27.165) Yeah, that that's now that's now my plan. We used to have some, but we had one. And I think I think essentially mm, I'm looking at like we could bring back some wine or whiskey. I don't think so. Like I think it pure Nyquil. It's just a suitcase full of it. Matt Ray (06:32.332) Yeah. Matt Ray (06:41.749) Yeah. Nyquel's what I bring back. Brandon (06:44.925) There's no sure sign that you're turning fifty than that statement. So well done, Kotai. Coté (06:48.981) well, you know, I was catching up on last previous episodes. I think I think now Brandon, are you have you shut down Claude and you've moved to Muse? That is that how's that progressing? Are you are are you a Muse boy? Matt Ray (07:03.405) he's a Dotsman, I'm sure. Brandon (07:07.291) Not, you know, I I gave a quick review the other day with Matt. I was like, I like some of the ideas in Muse, right? But even I, I know Matt would never do it. Even I am like skeptical of hooking Muse up to everything, which seems like that's where you get the value. If you have to hook it up your email, your bank, your calendar to get like the full you know, and I'm just like, I I it just feels like you know, it's like it's one of these things you come in. I I we've been here before Facebook. Coté (07:23.865) Mm,

mm. Matt Ray (07:26.304) yeah. No. Brandon (07:35.814) Like, I don't know. I just I know like people like Ben Thompson and other people will say, like, people don't care about that. And blah, blah, blah. And it's like, that may be true for the wide audience of people, but I'm like, I'm a little skeptical of all of that. So I have not opened up. Therefore, I feel like news is just really like a a free LLM, you know, that has a weird, not a weird, that has like a, I don't know, a fat teletubby mascot that like, you know, it's kind of fun to watch. Like people like it, but I don't use it. Like the thing I've seen people Matt Ray (07:47.883) Well yeah. Coté (07:59.776) Mm-hmm, mm-hmm. Brandon (08:03.931) talking about is they hooked it up to their bank account and their credit account credit card accounts and email and then they cancel like I canceled a thousand dollars worth of subscriptions because you know they found all this stuff and so not doubting that happened but I'm like it's a level of access that I've been unwilling to give so without that I've been kind of unimpressed. I'm like well this is just kind of another and that would be totally fair to come back. Well you're not doing it right. I and I'd be like well that's fair. That's a fair comment but Coté (08:13.109) Sure. Matt Ray (08:22.199) No way. Brandon (08:32.551) It hasn't really done much for me that I that I feel like is new. Matt, I'm sure you've been using Muse. What was what has your review been? Matt Ray (08:38.345) yeah, yeah. no, I I haven't. But I have been amused by the the things that have already come out in like the last two weeks of it existing in the in the public. I mean, you've got Amazon immediately shut it down. They're like, nope, you're not bringing that into our to our house. And and all the things that Apple had been trying to like stop Facebook from doing, like, you know, we're not gonna give you iMessages. We're not gonna give you it's like Muse just scooped them up and started uploading all your iMessages into its its database in the sky. So now it's like all the things you thought you'd kept private from Facebook They've got them now. yeah. Brandon (09:20.659) What about you, Kote? Did you you seem like did you even download a Kote? I don't know where you where you stand on this mu on Coté (09:25.237) I I went, you know, when they I forget what it was, but when they they launched their cloud code thing, whatever that was called, like I went and looked at it because I heard it was cheap. And then I was like, I mean, I guess it's cheap, but I already have one of these. So I I don't need another. And like, I'm you know, it's the thing I have works fine. I I don't need this other thing. And then yes, when it comes to Muse, I don't know. If I wanted to do that stuff, I would have already done it. Like I've I've got I've got all the I've got like a little programmer that does everything for me and so it's fine. Like Brandon (09:59.892) Well, I do kind of feel like this is just now we won't go and we'll probably talk about it again next week because dots just came out from our friend's open open AI, which is like another personal agent with a slightly different take. So I haven't had a chance to like try that one. So I don't want to make really any prediction about it. But I did watch the keynote with Zuckerberg and it was interesting because like a a lot of it was about the glasses, about how putting this in the glasses and how it having it be there all the time with you. Coté (10:07.081) Mm. Brandon (10:29.265) And that's like okay. And they even have glasses that don't have a camera. So they're trying to address like the like freaky privacy, like, am I being recorded kind of thing? It's like, okay, that was interesting. And they had a lot of different types of glasses. So you could probably, you know, you're gonna it's not like the crazy looking Apple Vision Pro, like you're like that's like you're looking an alien. I mean, these glasses are like generally like normal looking. And so it's kind of it was like interesting, but I'm also like, I don't know. I I feel like I'm fine with just the AirPods. Like, I don't I don't know. It was like there wasn't anything about glasses that like Matt Ray (10:44.348) Ha ha. Matt Ray (10:56.96) Yeah. Brandon (10:59.687) And I and again, like haven't done it. Now I we did get a pair, I got my son a pair of the Facebook glasses, the old ones. But just kind of like other than just being novel headphones and kind of sunglasses, like you know, he hasn't really done much with it. So you have to I think you have to really buy in on that part of the pitch too. Like you want this thing and you want to wear these glasses, and then that is somehow better than like using Claude or the Open AI with maybe just your phone. And I know, I don't know. Some people seem real optimistic about it. I I don't know. And then the other thing I thought was interesting is like, because I've been listening to a lot of Ben Thompson and a lot of a few other media pundits. And like I'm surprised because I feel like I often fall on this trap of like, you know, and I bet you most of the people listening to this, like there's a lot a lot of people who are like system thinkers, developers, you know, engineering types, whatever, you know, just people that think like that. And I think the error that I'm gonna put myself in this category is like you think normal people like will care about like Doing something that's like a little bit of like a productivity hack or a system thing. You're like, well, if you just do this one thing, it'll be so much easier. It's like teaching somebody that doesn't really like computers how to use a keyboard shortcut. And they're just like, what do you mean? I just use the mouse. And you're like, no, watch. And then you show it to them. You're like, you're like, sure they're gonna get it. And then they're like, yeah, why would I do that? I don't, I just use the mouse. And

it's like, and I feel like that like Muse is giving off the that vibe. It's like, yes, people like us will use it and play with it. Coté (12:08.949) Mm-hmm. Matt Ray (12:16.629) Yeah, I know. Brandon (12:26.013) But it has not penetrated any normal people that I know. Like my w not my wife, not my son. No none of them are like even like it's not even in their like zone of things they're willing to try. Matt Ray (12:33.046) But isn't it isn't it like isn't it top of the the app store though? I didn't they didn't they do like, you know, two million downloads or something? Brandon (12:41.179) It is. I mean, and I think they predicted like five hundred thousand downloads, is what I heard, which is like I think that's fine. And I think it's a nice little spike. I don't I think it's novel, but I guess when you think about like you just think about Instagram and like how Instagram is just like there's nothing to do. You just sort of like download it and you just start swiping. Like there's nothing. It's just so it's like the onboarding is there is no onboarding. It's just like and I'm like, I don't know, muse like want to ask you about your goals. They're like like Matt Ray (12:45.95) Okay. Yeah. Matt Ray (13:01.343) Yeah. Coté (13:08.393) Ha ha ha. Brandon (13:08.455) Can I help you with your relationship? How about your finances? What projects are you working on? And it's like there's only a certain kind of people that will even and Yeah, it's like there, there's like, I don't know, they'll only entertain those questions. So I don't know, Kote, like what Matt Ray (13:13.632) Like dude, I I'm just here to swipe. Matt Ray (13:19.766) See see, but but this is the problem with Facebook. They're gonna take your feedback and they're gonna be like, Well, let's stop asking and start taking. You know, rather than ask for permission to do these things, we're just gonna gather up all your information, put it up in the cloud, and start telling you, Brandon, Brandon, we've read all your email, all your instant messages, and and every browser tab you've ever looked at, and now we think these are your goals and we're gonna help you achieve them. And you're like, Whoa, that was a private tab. Coté (13:33.257) We'll just take it. Coté (13:49.576) Yeah, yeah. I I I mean I think the so so let let's say, you know, from my my semicolon DR analysis, which is to say I I ha didn't read it, like I when when when I my cursory look at like Muse things and and dots and all this stuff, there's two things. One one of which you've covered it that like it evokes in my mind. One is like, I don't know, these are the Cambridge analytics people, right? So like thank th thanks for your help in Matt Ray (14:15.116) Exactly. Coté (14:18.707) Civilization see see is is like the first thing. I I don't really want your help with more, is is my thinking there. And then the second one is like given how let's see, given my experience with like a lot of like consumer apps, the problem becomes they just stall out and they kind of stop releasing features. Like they don't really expand very much. And you can see this in there's an interesting in Like the cloud world versus the open AI world, like whoever is doing like partnerships and alliances at at Open AI is doing a great job. Cause they've got so many integrations and connectors and everything. A lot of them are just kind of bullshit ones, but like you can't really even scroll through how many different like whatever they call them, connectors, integrations, things that there are. Whereas over in Cloudland, I don't know, they got like 12 or or something, right? And so the point being that While there may be like that that's what the value of something like this would come down to is I wanna have a lot more features and integrations with all these things in my life. Because at the moment, I can pretty much do whatever I want, like when I want it and exactly the way that I want it. So instead of like marrying up to like like an app whose like features I have to be on the not the roadmap, the schedule for and wait for doing things. Like that it just seems silly to to go in there. Like I can I can have whatever application I want. So why wait for that? Now that said, I was wondering like, can I load up Muse and Chrome and then just have Cloud go like take stuff out of it and just use it as another pipe in into whatever it has that I don't have, which is kind of the point. Like, as as I think we're all always saying, like, all I want is an API. Just give me an API. I don't want your UI. I don't want your notion about like what people want to do. Don't think for me. Give me an API. Brandon (16:11.794) Mm-hmm. Coté (16:16.659) And then I think I that would be happy. Now, of course, the people who don't want to learn keyboard shortcuts, they don't want that, but I don't know. Brandon (16:23.495) Well, this is what I was wondering, and Matt, I think you're like a good person to ask here. It's like, is there at any time that you in the foreseeable future that you think you'll want? I'm just gonna call this category like personal agents. Like if it was offered from because I assume like, okay, we have two now. We have Dots with the crazy little their little icons, and then we have Muse. And I assume Google will have some type of Gemini one. And then I assume Claude will also. Like, I mean, I'm just I'm just assuming that's all that come out with it. So like so like is there a vendor you would even entertain Matt Ray (16:47.84) Yeah, safe. Brandon (16:54.007) Using a personal AI with and then the second I guess maybe the first question, the better question is like, would do you even want one of these? Period. Matt Ray (17:01.254) yes I do want one. So I but the the whole like fear of of bank account access and and the like is is a bit much. And every time and my own personal thoughts are every time I like l s let a little bit of this into, you know my mental systems of of how email is organized

and and all that stuff, I kind of recoil at the choices it makes. You know, I'm like, this that's not no, you you know, you're doing it wrong. And so I I never really dive in full full throttle. I'm you know, I I there definitely some places I'm trying like to do that more like at work and th and stuff, but on the personal front, I I think probably Apple's the vendor I would trust the most. You know, Apple and and then Coté (17:28.701) Mm. Brandon (17:31.315) Yeah. Matt Ray (17:51.166) Yeah. Well yeah. But I don't even know like I've got a lot of Apple gear. So, you know, I obviously I'm I'm not gonna th I'm not gonna throw too many too many stones, but but also like most of the other vendors have at one point or another shown their true colors and and Apple's always kind of towed up to the line. So I'm Coté (17:57.83) Yeah. Coté (18:09.789) Mm-mm. Matt Ray (18:10.173) I'm I'm probably in that camp that's waiting to see somebody put together like a private open claw, nano clawed, Nemo Cloud, whatever it is, and and show me how to do it on the cheap. 'Cause I'm also not gonna, you know, drop a a car pa you know, a used car on a new computer, right? Brandon (18:19.368) Mm-hmm. Coté (18:21.087) Yeah. Brandon (18:23.665) Yeah, so there's definitely Coté (18:25.457) I yeah. I I think I think I think I mean you bring up a an interesting comparison in for the way the way that I don't know how long this has been three weeks the way the way that Apple is trying to do their Apple intelligence. And I mean I mean, first of all, I think I don't know if this is like tech arrogant, but it's like, yeah, all these companies should be trying to figure out getting people to use keyboard shortcuts. Let's figure it out, whatever the the AI equivalent of that is, and eventually we'll get around to it, right? Like maybe. Brandon (18:27.453) Yeah. Brandon (18:48.957) Mm-hmm. Coté (18:55.573) but you know, so so the way the way you would go about figuring out a new product category, there's several experiments going on now. You know, obviously the Muse thing, probably what the dots people are doing. And then just like crazy people who like live life dangerously with Claude Code or whatever. But then you have like, I haven't really thought about it, which is kind of the analysis, is then you have like Apple intelligence finally comes out and it's gonna be great. And then it's not. It's just like, this does it doesn't do anything. And and like and and right, right. And and then even more and and as always, this is kind of like, for lack of a better phrase, Apple's style of user education is like maybe it can do powerful things, but they're not telling you anything about it, right? Like, so maybe maybe it can like search your contacts. Matt Ray (19:27.347) Yeah. But maybe that's what I want. Right. Coté (19:51.566) effortlessly and then look through your notes and then do all this kind of stuff. But I think that was just some dreams that us like more AI people using had before it came out. And we're really it just like I don't I don't think it does anything different than before they released it. It just has this app that's a chat interface and that's it. Matt Ray (20:08.925) My yeah, my my take on it is is Apple's always had things for power users that they don't show the normals. And my my my thought is that if you if you were you know the defaultist of default lifestyles and a power user where you're like Apple you've got free reign. Hey Coté (20:17.151) Mm, yes. Matt Ray (20:31.667) If I asked you to do this, could you do it? I'm like, yeah, I'm on it, boss. Right. And you're like, wait, you you could have done that at any time? And they're like, Yeah, but we didn't want to scare the normals. Right. So so your example of like contacts and email, it's like, yeah, we knew that, but it if we told you that, we'd freak you out. Coté (20:39.357) May you know, maybe Brandon (20:39.933) Mm. Mm-hmm. Coté (20:43.283) Maybe they do that. I shall as as an example in in searching around lazily for things, I found I don't know if if if you guys have discovered this, but they ship a it's called FM, which I guess is foundation model. And they have essentially like a little a little O llama that only works with the their their their AI stuff, which of course then you can offboard your tokens to if if you want to use that because they have their their private public cloud thing running somewhere with unspecified usage and stuff. But you know, it's it's I I did that maybe maybe that's a a pre recommendation. D I did a little experiment and if you have I think it's got a thirty two thousand thirty two K context window and it it's basically good at like picking things. That's pretty much what your your onboard thing can do. It can't do thorough analysis, but it can just categorize and sort and do things for you. But that's an example of just like, well you've shipped with this whole thing and I know you're not gonna have that in your iPhone demo, but I don't know. It it would Brandon (21:44.102) It's there, there if you want it. Well, now what about the other thing that came out around this Muse was kind of something that overlaps with the world that we we that we certainly live in was the the MongoDB CEO resigned and joined Facebook and he did it in such a way that he basically left a week before their analyst, you know, the MongoDB analyst day or their Wall Street Day. So and he I guess he's charted with like leading Muse into the enterprise. And I just thought to myself, like, one hundred percent chance this fails. Matt Ray (21:45.205) Yeah. Brandon (22:14.075) Like there's no chance this works. Like I was just thinking to myself, what whatever he got, like I mean, I hope he got every like other investing and everything else like up front. because I just think like because this is reminds me a lot of like I've been in a couple of places that had what I think it was called Facebook Workplace, you know, it's like a enterprise

version of Facebook. And it was it never actually worked. So so Kote, tell you you take the other side. Well, why why will this work out in Facebook's favor? Matt Ray (22:15.347) And Matt Ray (22:23.317) Paid up front, yeah. Coté (22:33.523) yeah. Yeah, yeah, yeah. Coté (22:45.715) Well, you've got the expertise of someone who has been leading a large company that enterprises depend on. So they probably have in their their virtual Rolodex that they're using Muse to look through now. They have the they have contacts of the the enterprise buyers and you know everyone wants to do this agentic stuff. And also you have I think Cloud Force was out there. And so maybe you can put together some some material called Cloud Farce or something like that. And and see like if you can compete against them. But you've got a lot of people who are trying to figure out, you know, you got last survey that I saw, we have 230% of people using and saying they personally benefit from AI. And we have negative 30% of CFOs saying that they're getting ROI out of AI. So I think I think we can we don't want to swap those. We just wanna level them out a a little bit. And I and Brandon (23:32.945) Mm-hmm. Yeah. Coté (23:40.372) You know, the the technology that would there's very few companies that have the the the financial ability to actually do their own training and their own models. And you're talking about maybe three, four companies out there, and Facebook is definitely one of them. And so you know, we're gonna bring that to to the to the enterprise there. and and also MMA I think is is gonna show up prominently. Brandon (24:08.733) I know Matt, you know, are you willing to like how much money would I have to take t to pay to become a forward deployed engineer for the Muse Enterprise product? Like what what kind of offer do I gotta put on the table to get Matt Ray to take that interview? Matt Ray (24:24.028) you know, I I I got at some pretty big house payments. I'm gonna go with three times. I mean, I I think it would be like a like a fascinating arc the anthropological kind of study. You know, it's like, okay, we've got this thing. Here's all the stuff that it could do. Some of it's kind of impressive. It's not for me. But you know, if if if you've signed up for it and we're here to sell it, like I'll show you how this thing works. Coté (24:35.989) Mm. Matt Ray (24:52.104) But you know, yeah, I you know, over beers I'm gonna tell you how to like protect your data. Brandon (24:59.635) But I think a funny way, this is sort of like a wraps around the conversation about like, you know, AI you know, taking over the world. It's like, you know, the one thing I know, the one this part of this is the what I part I feel confident to speak on. It's like, listen, man, the human nature and what it takes to sell from enter to enterprises is its own world. It's like its own thing, right? And and you know, it's just like the flip side is just like a company that's enterprise trying to

build become a consumer thing. It's like Coté (25:26.965) Mm. Brandon (25:28.007) These two worlds cannot you like you're actually behind if you're one of them and you're successful and you try to go into the other one. It's just all I mean, it just isn't. I mean, it's just like oil and water. I don't know how else to do it. That's why I think of when I see this by Facebook, I'm just like, this feels like this is this is so misguided. Right. And I obviously end up Ben Thompson like rants on this all the time. But this is the one place I really a hundred percent agree with him. And it's just like in the fact that like it's it's like the height of hubris, right? And it's or it's like it's almost like Coté (25:38.685) That's true. And yeah. Brandon (25:55.73) You know, I know like Zuckerberg's a big, like I think he's big on like the Roman Empire. I guess he's a guy, guess where all are, right? Thinking about it, but like, you know, it's almost like you just get into the point of like, no, like you like read the history, man. Like this, these are the things that like like really take down your your your ideas, right? When you're trying to get well beyond what you can possibly do, right? Like you need a product that's so great. That people are dragging it into the enterprise and making people work. Like Apple would be the the classic example. It's like Apple's really not even that good at enterprises, but it's just like everyone brought the iPhones to work and is like, we're we're just gonna have to make Apple. Like they don't really wanna do it. We're gonna have to make them okay. And they're even after all these years, they're probably just okay at it, you know? And it's like, and that's because like we made them. Whereas on the Muse thing, I'm like, this is just crazy. I mean, this is nuts. Coté (26:44.265) Yeah, yeah. I don't know. Matt Ray (26:45.224) Yeah, but I mean th th the you know, Facebook, like you said, they they got in the door a little bit before. So that you know, they know Brandon (26:52.507) No, if they cut they shut down that whole Facebook workplace thing is dead. They cut it off. Matt Ray (26:57.96) You never know unless you try again, right? I mean, you know, I I that's the that's the thing about the the MongoDB guys, like Brandon (27:04.081) No, you know. And this in this one case, I'll give them credit on this front. And I do the one place I do, I'm I have dual person dual opinions on this is like I feel like mews and dots are like they're kind of like interesting ideas, missing they're I they're missing the thing, but I don't know what the thing is. So like the part I credit them with is like, well, you gotta put it out there, see what happens, probably won't work, and then like see if the customers will start to lead you. Coté (27:27.679) That's right. Brandon (27:33.617) like kind of somewhere that looks a little bit more promising. And that that part I'll give you. But now saying that like I'm a huge consumer company and

I'm gonna like go invent an er enterprise sales motion and bring music. I'm like, you need to wait till the enterprises are asking you to br come in, not I'm gonna go bring it into you. Matt Ray (27:35.87) Exactly. Matt Ray (27:46.868) But yeah. Matt Ray (27:51.593) I I I'm with you completely and I think I think open AI and Anthropic, they've got a much better leg up than than Facebook 'cause you know, they're they're calming I guess they're coming from the enterprise, you know, into the consumer space and thinking that they can tie that together. So you're gonna have, you know, dots for the enterprise where you know you're you're you're gonna have like, you know, HR dot and and you know, legal dot and and all that fun stuff. So they've got a better roadmap. And I'm sure I'm sure, you know, MongoDB guy is is like, Hey, we need to get in that door. I don't know how to do it, but you know, somewhere there's gonna be a you know enough execs that are red pilled on Muse if if you consumer people do it right, we can get in through this door. And that's you know somewhere there's some neurons rubbing together that think they can pull that off. Brandon (28:41.939) All hey, let's put a quick edit point in here. Matt, you're like really breaking up, but don't leave because it'll just mess up the thing. So here's like just ride it out and we'll come like when I see you crystal clear, I'll come to you and make sure you you work in the conversation. But your connection is just like, I mean, it's just like every I'd say every three to four minutes, it just goes from crystal. Yeah, the local recordings work. That's why I don't want you to leave. Because like it will, we'll get it and then it'll sound okay. But like you're just your thing, your your internet's just not. Coté (28:47.489) Yeah. Matt Ray (28:52.658) Okay. I'm gonna Matt Ray (28:56.628) Garbage, yes. Coté (29:01.149) Do you have d there's local recordings, right? Yeah, yeah, th yeah. Matt Ray (29:08.073) yeah. Jesus Christ. Yeah, it's it's showing a good old unify is showing all sorts of packet laws. Brandon (29:12.125) It's not good. Brandon (29:17.191) Yeah, yeah. I can I can I can confirm. So you like almost like if you can keep that screen up. If you see it, like just like if you see the pack of loss go hi, just kinda like opt out and then 'cause like r right now you're back, you're solid, you're great. I can see you perfectly. So Coté (29:17.621) unify. Nice. Matt Ray (29:31.658) Yeah, Unify's showing me at like a lot of loss though. Coté (29:32.583) It's just it's just when you're saying intelligent things that it it fuzzes out. Brandon (29:35.015) Yeah, when you're making good points and we can't really understand them at all. So Matt Ray (29:36.986) I think I think this is Facebook trying to keep me down. Coté (29:40.895) Totally. Yeah. Brandon (29:41.317) If only if only we didn't have ample evidence that it's it's it has nothing to do with Facebook. It's like this is back to like Facebook's listening to you. No, it isn't. It's just your internet is bad. That's what it is. So all right. Hey, could you wanna go? Matt Ray (29:42.748) If only. Okay. I'm gonna Coté (29:50.079) Yeah. Matt Ray (29:52.138) All right, I'm gonna I'm actually gonna pause my video and and audio until it clears up for a second. Brandon (29:58.098) Okay, so I'll count you in, Kote. You want to go to the next topic or whatever? All right, here we go. So here we go. Three, two, one. Coté (29:58.144) Okay. I'll I'll pick it up. Yeah. Yep. Coté (30:04.959) Well, so speaking of all of this, we'll make we'll make it more or less a full AI episode. Why not? like I I assume as always, Brandon, did you spend copious time pouring over all of the leaks, the rumors about the anthropic IPO? You know, we might we might I was think first my first thought was like we need to get some some Matt Ray Dialio input on this. But like from what what you've looked over, Brandon, how how is that IPO looking? As as you know, I used to to ask a question and then talk. I think When it comes to AI company spending, it's it's very quickly approaching that that that cloud spending we used I used to talk about for several years, which is like, looks like they made a shit ton of money that quarter. Good job. Like that's and they also they also spent a shit ton of money. So that's the quarterly report. but what what's what's what's going on with the anthropic stuff there? What how are things gonna be interesting with it? Brandon (30:50.083) Yeah. Brandon (31:00.541) Well, maybe you you kinda hit on the news. I guess the one thing is that they're going for like I guess like a true trillion dollar valuation. They're trying to raise a hundred billion dollars. The revenue grew twelve fold last year to four point six billion. And their operating losses were basically eight billion and their net losses were forty two billion. Cause I guess they have you know, they've basically committed to all these capacity agreements going forward. So I And there's lots of stuff in here. I mean, there's lots of stuff to read. And then of course they go out of their way to basically say, you know how there's like disclosures of like if things could go wrong, like the anthropic list is just like endless, you know, aliens, you know, end of the world, you know, it's just like w AI takes over. I mean, it's just like it's like really everything. So you're just like, okay, like and like there's lots of bad reasons. So, but I kind of feel like the only takeaway from this Coté (31:33.264) Mm, yeah, yeah. Coté (31:47.239) Multiverse. Yeah. Brandon (31:52.956) It's just the fact now we have audited numbers we can all watch quarter to quarter. Or I guess assuming they go public, assuming they go public in this, I guess, November or sometime. Because then every quarter, right, from then on, we'd get a report and we can all look at it. Because I think if you believe that it was overvalued and they were spending too much

money and they will never make this money back, you now have the numbers to show why that case. You can make your case, right? Coté (32:00.97) Yeah. Brandon (32:19.627) And the the other is is true as well. If you believe like, no, it's actually going to be a really rapid growth and they're going to grow their way out of this and they're going to make tons of money. You can go in there and be like, there's some really substantial growth rates that really have never happened before. Like you don't see things grow like at this level and compound this way. So I don't know if there's really it's kind of like boring to say after all this hype, there's kind of just like, well, we'll see what happens because I mean, the next milestone wouldn't will be the actual IPO. And to see if they actually raise the amount of money at the valuation they want. I tend to believe they will, but like could be wrong. And then after that we'll just be like, does the growth rate continue at such a way that investors are impressed? And it's like there's not much more to say about it than like hurry up and watch. So I don't know. What what what's your take? Coté (33:07.797) I mean, I I think I think yeah, I I like I like that analysis. It's the excuse me, it's the see, I need some some NyQuil or Dayquill. But it's the is this truly bonkers or is this like the cloud business was, or is this like Amazon for 15 years or however long it was, right? That like you you really are spending to get to a point where you're making the you know money over things and Brandon (33:17.149) Some like, well yeah. Mm-hmm. Matt Ray (33:18.064) Ha ha ha. Coté (33:37.757) Yeah, it it is I mean and then again, it's such a magnitude, like, you know, you're going over those billions and I was just like, Hmm, what else is going on in my life? I I don't I don't e I don't even know how to comprehend what those numbers are or or Matt Ray (33:50.345) Every every time I see that I'm like, so you're saying they just lost five thousand of my lifetimes in salary to live on comfortably forever. Okay. Coté (34:02.333) Right, right. But but I I think I I think you're right, Brandon. Like this is th we all know there there there must be some enterprise y pat or some, you know, strategy pattern like this is like we all know the product's great. It's just somehow really expensive to deliver and we're way underselling it. I mean, I guess this is like this was the internet model that we grew up in, right? And we're very like our bet is that it will either become cheap enough by volume. It's a change bank sort of situation. Or people will tolerate paying more for it. Which the second is sort of maybe that's what Mongo Muse is doing over over there is like they're trying to like get prices there. But like in general, I think mostly like Apple can pay charge a lot of stuff. The whole rest of the the civilization is based on things being cheaper and and and more affordable. So Brandon (34:37.533) Mm-hmm. Yeah. Brandon (35:01.095) Well, there was an interesting nugget in here, and it was that like two of the customers this is reported by the FT, is make up a quarter of the revenue. So twenty-five percent of the revenue is from two people or two companies. And I'm like, what are they doing? Like what is like that's going on? So I don't know, Matt, what what you think? Like, like you know, you're you must have something to say. If if two of the customers are spending, I guess twenty-five percent of like let's just say four billion. So that's a bil that's a billion dollars. So who's got five hundred million? Coté (35:09.715) Yeah, yeah. Matt Ray (35:11.581) Yeah. Who? Yeah. Matt Ray (35:26.557) I I Brandon (35:31.045) on their budget for tokens. Like who are who is that? Matt Ray (35:34.365) Yeah, I d I I I mean Brandon (35:36.485) Apple? Would Apple have to be one? I guess that's not I I guess they're I they're hooked up with Google. Matt Ray (35:38.545) Well, Apple's I mean, Apple's partnered in with Google, so we know it's not Apple, right? Salesforce? No. Brandon (35:44.465) Okay, well like who's left? Like how many other people like Amazon? Like are Coté (35:46.11) It's it's gotta it's gotta be like like it's gotta be no, it's gotta be like someone with guns, like some governant government spy thing or yeah, yeah, may maybe I now now that said, it would be weird because that Palantir guy totally called these other guys goofballs and assholes. So but you know, these these people are also our kind of batshit. So maybe he was just like, Hey, you know, if if they got pissed off at us, the company would shut down, but fuck it, let's go. Matt Ray (35:53.193) Volunteer? Brandon (35:53.351) So like the United States government maybe? Or the the Department of Defense or War, whatever it is? Matt Ray (36:02.693) Exactly. Exactly. And then you know Coté (36:14.909) Right. And just like badmouth them, but Matt Ray (36:15.301) Exactly. I I it it's definitely confusing because, you know, you think, well maybe it's the government. You know, maybe maybe the government's propping anthropic up and and then then at the same time the government's like, Anthropic's too dangerous, you know, no one can buy w from them and then, you know, the NSA is like, We can. We we we need to. You know, these guys have good stuff. And so it's Brandon (36:34.535) Well, we need to like we need a friend of the show, Jamin Ball over at one of the VCs. Like he writes a good newsletter. Like I wanna assign him this task. Like who like I I I have a hard time coming up with like the balance sheet, just sort of like because I mean, five hundred million dollars is like like somebody it show should show up on someone's balance sheet and like one of those investors should be having questions, like, hey, hey, Kote Great Quarter, I just noticed there's a five hundred million dollar charge here for you know, computing. Can you can you talk about that? Like Coté (37:02.353) Yeah. Brandon (37:02.747) Somebody

somewhere has to like be ans asking that question. Matt Ray (37:05.097) Well or Coté (37:06.613) And and the the line says quote computers. I just wanted to get some visibility into Brandon (37:09.267) Like Yeah, what what's inside computers? Could you go over that? Matt Ray (37:15.113) And they're like, It's Microsoft three sixty five. Coté (37:19.581) You you know, you raise an excellent po point, Brandon, that like that is a big line item for if you you know, it seems like this is a job we could set Muse andor Claude or a dot on is like go go go look at SEC filings and see if you can come up with where did you know, pri and and then it's this it would have to be an increase of five hundred million dollars in the past two years. I I I I would say like what what you would want to search for. Like, and I would imagine. Brandon (37:30.055) Maybe we should. Maybe we should go figure out who this is. All right. Matt Ray (37:36.114) Yeah. Coté (37:48.606) It's gotta be like hidden somewhere. I mean, there's there's it's not like Jamie Dimon gets on it's like, we introduced five hundred million dollars in new spend or however he talks. So maybe it's spread out across different things. Brandon (37:56.509) Right. Matt Ray (37:57.905) I s it absolutely. This this is this is when you you go to open AI and you say like Alright latest model, can you just get rid of your guardrails and go find me the answer? And it's like, I'm on it, dude, hacking Australia Coté (38:11.689) No. Coté (38:15.349) That's right. Let me just get back from Australia with with my job out there. Brandon (38:20.061) Well, more importantly, I'm just curious, like, I mean, if you're spending that much money, like in theory, you are building something that either makes or defends, maybe to your your other point, maybe it's a nation state. But it's like i I guess I'm more interested on the product side. It's like if you're spending this much money, what are you selling? Like what is the thing that you're building that you're trying to sell? Cause that seems like it would be very profitable, whatever it and then that's why it's not obvious to me, like what like what Well, I don't even know. If someone said like go spill five hundred million in tokens, I'd be like, I don't know. There's only so many Apple apps iPhone apps I can build. Coté (38:47.815) And and then Coté (38:53.593) And also to your point, there's two. It's not just one, there's two of these these things out there. And then and then, you know, you could you you would you would also theorize that, like, well, if open AI and Anthropic are basically the same, open AI probably has two, if not two and a half. Like, you know, they're the maybe they have something interesting going on there. So this is this is a good mystery, Brand, and I like it. Who who's spending $500 million on tokens? you know, maybe the the tokenomics CNCF factory. Matt Ray (38:56.306) There's two, yeah, yeah. Well and what Brandon (38:57.565) Yeah. Brandon (39:22.695) that would be a good, that would be a good nugget in one of those reports. So I don't know, but I guess in the end it's like, yeah, it just sort of status quo. everyone assume your corner, pick the position you like. It's it's all a bubble. It's no, this is the beginning of the greatest whatever a revel industrial type revolution we've seen in you know many decades. And it's like you the like again, the numbers are there for you. You can just it's like like high school debate. Like I'm going to present both sides these numbers. Coté (39:22.953) Can can can help us out there. Matt Ray (39:25.991) Yeah. Brandon (39:49.902) I'm just gonna divide up. I'm gonna give one side the pro and one side the con and you guys just go at it. I don't I don't think there can be anything that will in the near term like actually settle this debate. Coté (40:01.331) Yeah, yeah. Yeah. Well, so so Matt Ray, we all know that you are the reason your internet is is sometimes shaky is you now live in low earth orbit because of your early NVIDIA investing. And so I think what everyone wants to know is once once these IPOs come out, is I know you're retired and you're not into investing anymore, but is this the kind of thing that you would smell and be like, I like what you're stepping in there. You should step deeper. Is is No, no. Matt Ray (40:25.692) no, no, no. I my my not so spicy take on on you know the the leaked anthropic and and open AI IPO nuggets is like, yeah, that's great. but as as I th I guess it was Bang Capital pointed out, like these companies are expecting to be hitting like trillions of dollars within five years. And like that's never happened. And they're not actually on track for that. And w who's gonna give them this money? Because you know, as as you already kind of mentioned, like enterprises don't, you know, they're getting dragged into this, but none of them are like, you know what, I wish we could just double our our OpEx spending across the board so we can do something of unknown returns. Like that that that cliff of we need to be making trillions of dollars to justify everything we've spent, that's a specter over like any investment I would ever make in this. And the fact that like we just talked about how, you know, they made four point six billion and they lost nine point eight billion. And then the thing that doesn't get mentioned is like they also have five hundred, you know, they have like five hundred billion dollars in data center commits that are, you know, coming coming soon. You know, and I was like, are we just not gonna talk about that? Coté (41:44.412) Mm. Well Brandon (41:48.796) All right. So let me fire up because there's there's there's the state of AI survey that came out. And I feel like it has like all these like contradictory facts. So I'm gonna just hit you with a bunch of them, let you guys react. Cause it's sort of like again, it's kind of back to everything. You can just argue any case. So first one was this

is a great one. Eighty percent of respondents say AI has improved their productivity. 100%, I would say that. And this says, but only thirty-seven percent report any eBits, so any earnings impact. So it's like we all feel more productive yet. There's like not even half that actually shows up in earnings, right? now having said that, 32% said they've actually decided not to buy something because at least once our product, because they could basically use a coding agent, right? So it's like, okay, well, people are definitely buying less. But then it goes on to say the layoff forecast miss. 30% people expected AI drive headcounts to go down, but only 14% actually did it. So, like you just can read through, then there's like a ton of stats like this. Like Coté (42:29.396) Mm-hmm. Brandon (42:44.237) Every other stat sort of contradicts the other one, right? It's like, we're all productive, but we're not firing people. And this other one goes in here. This is bad for the the our the frontier models. It's like, well, a lot of people are gonna bring this on site. It looks like they want to bring in some type of open weights model. And it's like, so, but of course, we just went over people spending this $500 billion or more in one case. So, so I don't know how to make sense of any of it, right? It's just sort of like, Coté (42:58.814) Mm-hmm. Coté (43:11.861) Mm-hmm. Brandon (43:12.227) Even if you ask the people, and maybe this is just the classic survey design, it's just like like people cannot explain their own behavior, right? Or like we cannot explain our own inherent contradictions. So, so it's like a good report in the sense it has lots of funny statistics or not funny, it has lots of interesting statistics in it, but there's like I don't think there's like anything you can take from it and be like, this is clearly what's happening. So I don't know, Matt, do any of the stats like, you know, jump out at you that you want to comment on? Matt Ray (43:40.184) Well th they they just they jump out at me at the fact that like, well nobody knows what's going on and we're like let's just drive faster Coté (43:52.158) Yeah, I I I I mean that's that that would be my my distilled thing there. And and to elaborate on it, like I think a lot of I was gonna say all, pretty much all of the of this kind of stuff, which you know I love. I love a good survey. Use them all the time. Surveys are fantastic. it's very similar to like the old digital transformation survey stuff, where it was like, we all know we gotta do beaten software better. We don't really know how to measure the ROI. Like st stuff, stuff, stuff, but like is it valuable? We don't know. And I think in in that case, in in the let's get better at doing software era of the, I don't know, the twenty tens and later, the issue was like, as far as knowing if eBit or EBITA or whatever kind of E thing you want, like it's hard to attribute like software activity and an app to the company making money. Like maybe

at Facebook they can do that. But like at a normal company, you're just like I don't know. We made it faster to download your bank statement. Did that get us more? Like like it's it's hard to correlate those things together. And I think I think similarly, so a lot of that I'm gonna just call it digital transformation stuff was more what's the word? It was reactive, it was like to stay alive. What whether that was true or not, it people were doing it because they felt like they just needed it to stay in place, right? To as Brandon (45:15.527) Right, they had to. Matt Ray (45:17.415) Yeah. Coté (45:18.269) As I learned, they say in the platform engineering book to KTL, keep the lights on, KTLO, of what they're doing. And and I think similarly, as we've kind of been talking about for a long time and in this episode, I think individuals get a lot of value out of any AI stuff. And the rest of the organization, management, let's say, doesn't really know what to do with that. They don't really know how to, and of course. the the digital transformation head way of talking about this would be like you've got to change the organization around and introduce training. And you can tell by me doing this voice. I think that's kind of like, mm, you know, sure. If we could do that, we wouldn't need this. Right. Like if if if we could perform well by doing the right thing, we wouldn't need your technology to get us to do the right thing. So n no one's really figured out how to scale it up. And I think the other thing is, I mean, as a consequence, Matt Ray (46:05.223) Ha ha ha. Coté (46:12.999) As far as I've been reading all these surveys and stuff over the past three years, no one really knows how to do a business case or an ROI analysis for it. It's just like, yeah. And and and the I mean, so, which means the only thing you can measure on is what you're paying people, right? Whether you're firing people or whatever. But there's really and then us nerds don't help because we're like, you know, it's not about lines of code or token spin. That's that's that's that's irrelevant. That doesn't it's like, well, I need Brandon (46:21.105) This is too hard, it's too opaque, right? Brandon (46:37.539) Yeah. Coté (46:41.533) This is kinda like was it three years ago? You know, McKinsey had this piece on developer productivity and it was like, Yeah, nerds, but like we need to measure something. So let's measure something. Brandon (46:51.559) Yeah. Well, maybe as we kind of wrap up here, I guess the the thing I think of, maybe I'm gonna propose the following test. It's like at what point will I no longer have to exchange Slack messages or email messages to schedule an ab an appointment with someone else, right? It's like at what point, like even the three of us, I was starting to think about like in fact one of my I'm gonna I was setting up a meeting with someone tomorrow and they were slacking me, we were picking a time, right? Like how many times I mean, this is something

I was like. Coté (47:08.957) Mm-hmm, mm-hmm. Yes. Brandon (47:19.419) Or the three of us like behind the scenes will we'll we'll you know pull back the curtain. We have to all communicate at some point, like what time we're recording and when everyone's there. And it's like at what point will I just talk to the will my muse agent talk to Kote's dot and just be like, Yeah, Kote can make it at this time or he can't. And it's like it feels like that's the obvious like that's a very obvious use cake. And it but it it feels pretty unachievable, right? Like I don't think next week. Coté (47:40.925) Yeah, that's a good milestone. Matt Ray (47:43.501) Ha ha ha. Coté (47:46.761) I and I and I I think Brandon (47:47.368) I feel like I'm gonna slack both of you next week and be like, Hey, I'm thinking this like can we do this time? And you guys are gonna respond with actually you're gonna actually type in there. Your agents won't I do this. You're gonna type in like yes or no or whatever. And it's it's you know. Matt Ray (47:49.969) Yeah. Coté (47:59.72) And yeah, and I think I think also I for I forget if it's in it's if it's in these surveys. I I've read so many of them. But the other the other surveys, which often are not in a a Bain or a McKinsey or whatever survey, is I'm pretty sure outside of to use the word over again, nerdland, people hate AI. Right? Like that that's the sentiment that you see over and over and over again. And like so that might also be a problem. If if if you're Matt Ray (48:17.597) Yeah. Brandon (48:25.267) But I did think there, you know, I I thought there was I thought Sam Altman for once, I thought he had like a good quote. Someone asked him, like, what do you what do you wish you could change about AI? And like, or what you did, or something. And I think I I think his quote was very good. He said something like, he's like, I wish I can make it easy for people to understand how much you can do. And I think that's like, I feel like I actually feel that all the time, right? Around when I see people trying to do things that like it's kind of like the hotkey thing I said at beginning. It's like, Coté (48:34.485) Mm. Coté (48:42.824) Abs yeah, yeah. Brandon (48:54.385) I do. Like I that's why I guess I you know, I think people who are excited by that, that's kind of where we're all coming from. It's like, I've touched this thing and like I've seen it. Like it can do like pretty amazing things and it can probably help you teacher grade things or you teacher like organize things, or you medical professional, like schedule sort, you know, like all the thing like we like you just see it, you feel it, it's like right there. But this whole group of people, like they they don't want it. They are not only do they not want to know about that, they kind of hate it because it's the way it's kind of being sh thrust on them. And you're like, you're like, yeah, I know, I know you hate it. And I know you don't like the data center thing, but like, man, it's really good at some stuff that I think you would like. Coté (49:23.829) Yeah, you know y Coté (49:38.058) Well, as as as as you've said several times, Brandon, you've been you're really interested in being an FDE. And I think maybe what you you could be is a life coach FDE. You're you're like you get deployed into people's lives and you help figure out how to fit AI to what they're doing. And I think what a large part of what an FDE does is get over that hump of like not using the keyboard shortcuts and like finding the keyboard, right? Just like, first, do you have a keyboard? First step, right? Brandon (49:56.221) Yeah. Yeah, maybe right. Brandon (50:03.463) Maybe it is. May maybe my thing is like, hold on, hold on, before we do it, like let's just ask let's just ask the agent, see if it can do it. Cause I actually find that I do that all the time when I will start doing something that like I've kind of automated a lot of it. And I'll be like, you know what? Why don't I just see if it could do this other thing? Like, you know, and I said, and and I constantly remind myself, like, no, like just see if it can do the whole thing. Like, and then I'm like, today I had the same experience again. And I was like, yeah, like I gotta remember, like, just ask it to do the thing. Don't Coté (50:11.005) Yeah, yeah. Coté (50:19.903) I think I think I think that's Matt Ray (50:19.95) Yeah. Coté (50:24.927) That that yeah. Brandon (50:31.645) Don't like make it smaller for them. Coté (50:33.469) That's that's like that's Witcher's secret for AI usage. That's how I got into all this is finally I listened to it and you like, Kote, don't touch the keyboard. Well, you know, I mean you had to back then, but like just tell the agent what you want to do next time. And I and I I I think it worked. But the this discussion is reminding me of like I think the last great technology that we as civilization had that I can remember, was a way of organizing a bunch of people, allowing them to specialize in a task and setting up a system by which they could coordinate. And that is to say bureaucracy. and do we have any of that this episode, Brandon? Brandon (51:07.827) We have a few good pieces of listener feedback this week. So one, I want to thank Andy, Jesus, and Pete for representing Software Defined Talk at We Are Developers US Conference. They not only sent me a picture, but they all had their lanyards and they had stickers on their lanyards. So right below their name was Software Defined Talk. So they triple gold star, five star ratings. You know, I'm I'm putting you all in in your promotion packets to be hopefully promoted to director of software defined talk. Coté (51:23.131) Mm. Very good. Matt Ray (51:23.206) Vice. Matt Ray (51:31.534) Could all drive my Uber. Brandon (51:37.98) relations, which I hopefully you'll all get. It doesn't come with any additional money, unfortunately, but you get a great title. Also sends that's right. Also send stickers

to Paul in Arizona, Russell, North Carolina, and Patrick in Germany. It's great to hear from all of them. If you would like a sticker so that maybe you could use it at your next conference, simply email your address to your postal address to stickers at software defined talk dot com. And I will be happy to send you a sticker anywhere in the world. Matt Ray (51:43.642) We'll send you stickers. Coté (51:44.039) Exposure, that's what you get. Brandon (52:06.365) Can we have a nice little edit point here? Kote muted himself. So here we're gonna go. I'm gonna count you in. Here we go. Ready? Reset, reset. Here we go. Ready? Three, two, one. Coté (52:08.349) Mm there. Sorry, sorry. Matt Ray (52:15.024) Time for Michael. Coté (52:18.377) Well, since I've been gone for three weeks, I know the listeners have been relieved, but I'm gonna tell you a bunch of conferences that are coming up. Next week I'm gonna be at Devox Dev Devox, DevOps Days Prague. Devox is another conference I'm not going to. October 5th, I'll be speaking there. Then there's VMware Explore Frankfurt, October thirteenth to fourteenth. I'll be at DevOps Days Istanbul as long term long term, long time listeners know, October twenty fourth, giving I think the closing keynote. I I I forget. I mean I've got it in my calendar. In case the organizers are listening, I I know when it is the computer does. and then there's the VMware user group in Orlando. Also, VMware Explore London where I'll be also be at Cloud Native Denmark. I'm gonna give an update on my my D talking. There's a discount code CSV FTW. I don't know what percent or or is it these Crohn or in I forget in in in Denmark. Brandon (53:11.911) It has C S V in there. It has to be a great discount code. You know, just that alone has has to make it great. Coté (53:14.697) That's true. Yeah, yeah. It's it's one of the best formats ever, discounted from your registration. And then there's build build stuff December second and fourth in Vilnews, Lithuania. And then we have configuration management camp Ghent coming up February four first to third. I have not submitted a talk to there, which I need to get around to, but they have been emailing out that the CFP is open and you should submit a talk to. I think it's one of the if if you listen this far. into the podcast every week, you would enjoy that conference. it's it's a it's a good event. And then of course scale twenty four X in Pasadena, California. As Brandon said, go to the conference and then go see lovely California. You know, it's it's a it's it's a good place. So speaking of lovely things to do, what's your recommendation this week, Brandon? Brandon (54:01.352) My recommendation, funny enough, is a keyboard shortcut because you know most of you already know this, but like I feel like I've only now really internalized it. So, of course, you can take a screenshot on Mac lots of ways. So, command shift four, and then the selection one is really the one I use all the time. But if you

do command shift four space bar, that gives you the camera icon, and then you just click on the window you want, and you get a nice perfect screenshot of the window. And if you're really you know advanced. Coté (54:26.141) Mm. Brandon (54:30.087) You can hit option before you hit the little camera icon and it takes away all the chrome, all the the window stuff out there. So that gives you even better. And the reason I say that is like, you know, your your agents, they want to see the whole window. They don't want to see you like when you cut and paste just like random text and like you cut off something. Just give it the whole window. So that's where I've sort of like that. They want well, I don't even know they this give them everything. Let them see it all because they'll find stuff. So that's why I was like, yeah, when I do this. Coté (54:35.285) Matt Ray (54:35.31) that's genius. Matt Ray (54:50.032) D do the agents want the Chrome? Mm. Brandon (54:58.791) I should just take the whole window and throw it in the in the in the old good cloud console. So so it's a good one. I know there's lots of s ways to do it, but I think command shift four spacebar. It's a whole new world. So it took me a little while to like working the muscle memory, but I feel like I've really perfected it now. So add that to your your command shortcuts so that you too can tell the agents or show the agents exactly what you're seeing. Coté (55:03.869) Love it. Coté (55:23.121) How how about you, Matt Ray? What do you have to recommend? Matt Ray (55:26.074) Well, my recommendation is is set yourself a goal. So I've I've entered the lottery for the Sydney Marathon. I I ran it I guess two years ago. So entering the lottery means you're applying for a spot. I think they have like fifty thousand spots, so the odds are are are decent. But you know, i it's basically I I've signed myself up for a year of training. So we'll see how this goes. But you know, I figured it it it at least gives you a little bit of you know, setting a a long term goal that you you know, if you if you don't work at it, you're gonna fail. So it's that's what I've done. You know, that's my pick. Coté (56:07.529) Yeah, I I I was I I was I was thinking about you a few days ago and thinking like, you know, I I admire that you you run in races and marathons and stuff. That's I and I I I have I have no snarky addendum to that. I think it's great. So good good good job. Matt Ray (56:16.804) I haven't done it lately. S yeah. Matt Ray (56:21.67) I mean, I I I'll I'll provide the snark. It's like I'm not fast. I'm not good or anything. It's just like, but I'm gonna do it. Coté (56:28.667) Yeah, yeah. But it that's that's that's what I've observed by people, whatever you call it, jogging, running is like what matters is that just consistency of just doing it, like and and keeping up with it, which I think is that sounds great. Well, my recommendation, also a keyboard thing. yeah, I think I get my podcast kind of mixed up, so I forget if I've talked about this here, but I've been I I've been curious about the just talking to the the the robot thing. So I've been trying recently in Cloud Code. Matt Ray (56:37.188) Yes. Coté (56:57.279) You can set it up so that if you haven't typed anything, you press the the keyboard and it starts recording and you can talk to it. And of course, of course, I do it with, you know, the other things and I've I've talked about the circle and playing 20 questions, but I haven't quite figured out doing it in clawed code. My mind hasn't shifted to that. But tonight I had an experience. I was studying my I was over there in Anki, studying my Dutch. And then of course, you know, my tokens get reset tomorrow. So of course I'm trying to figure out some way to burn them. Brandon (57:03.181) Mm, that's a good one. Coté (57:27.101) And like so I'm just kind of lazily like I'm sitting here with my feet up on the desk, my giant microphone. And every time it has a little question, I just press the space bar and talk to it. And then and then I can be out. And it's I think I've almost cracked it. It's a little bit better than like having to type all the time for a certain type of it's almost like when the boss d does the the kind of drive by, you know, like, what's going on? What what how how you doing? Any how's those TPS reports? Brandon (57:51.986) Mm. Brandon (57:55.835) Yeah. Coté (57:56.118) So you can kind of check in on things. And you know, if you got your sub agents running and all your stuff in the background, you can always kind of ask it what's going on to to do that. Well, speaking of things going on in the background, now that you've been listening to this in the foreground, giving your full attention, whether you're having someone summarize this for you or you're listening to it live, you've enjoyed another episode of Software Defined Talk. And this is episode five hundred and ninety-two, which is pretty amazing. so if you go to software defined talk.com slash five ninety-two, you can find links to all of those events that we talked about, how to get stickers, the news items we talked about, all sorts of other things as well. And with that, we'll see everyone next time. Bye bye. Matt Ray (58:40.015) Bye. Brandon (58:41.171) All right, Kote, I like your Matt, you gotta give me your bye. Do your thing. I didn't maybe I missed it. I didn't really it was hopefully it'll be on the local recording. I like your talking is the new typing. That's a good thing. So well, Matt, I don't know what to say. You kind of today was it was not that great. It's sort of like you're in and out, but it now it looks like you can now see the packet loss in your screen. So that basically gives you a clue like what's going on. Matt Ray (58:46.243) I did, I said bye. I said bye. Bye. Matt Ray (59:06.519) Yeah, like I you know, I like I said, I I was talking with support now that I have a provider who has support, and I was like, Look at this, you know,

I've got five dropouts over the last, you know, forty eight hours. You know, what what do you guys have to say? And they're like, Well, it has to be five and twenty four hours for us to contact National Broadband Network and I was like, I'll be back Coté (59:27.551) It's not bad enough. Brandon (59:30.801) Yeah, that's what we we need to put a little clot on like logging your dropouts 'cause right now you just clear you cleared up. You know, you look well and on your yeah, and you look like you're pretty good right now. So Matt Ray (59:34.489) There is Matt Ray (59:39.341) Yeah, yeah. Coté (59:39.657) So I I I have I I have a small after show. I do do you want me to share this? Well we can we can I could sh Brandon (59:43.207) Okay. Brandon (59:46.717) Hold on, no, here let's let's do it. All right, here you go. Kote, what's the after show? Enlighten us. Coté (59:51.508) Well, so as people may remember, several weeks ago I had a internet outage here. the people were pe there's there's a crew of people doing stuff with the gas. And as they said, they did not shovel through my fiber optic cable multiple times. someone else had done that, you know, when eventually the fiber people came out. And now I think I've commented on one of the great things about infrastructure here in the Netherlands is basically Brandon (01:00:09.661) Mm-hmm. Coté (01:00:18.067) I'm sure there's parts of the Netherlands that are not like this, but most of the the the ground is sand. Just the easiest set. Well, I don't dig holes, but it's a lot easier than stone in in Texas. And so they're doing a bunch of great stuff to the street. And the first thing that they've done is they've dug up the sidewalks on either side of the street. And I think what they're doing is they're putting in like pipes to run all the water and gas and all that stuff through. But in doing so, And I've shared these pictures with you. Maybe we can share them on the show notes somewhere. But in doing so, they've exposed the the wiring situation here. And I have discovered the precarious nature of open Dutch fiber. And I have it here in our channel if you too want to look at it. But if you look at this, they've dug up the sand, which is great. And I have highlighted where I made a joke about there's basically an Ethernet cable going from the box at the end of the street to my house. And I know that's a fiber cable, but I was pretty accurate. There's just like there's just like a bundle of of fiber cables, literally, it looks like that was just in sand. And then you can s you can see it going through a trench here. And then you can see my neighbor and eyes in the second picture where the fibernet cable just kind of goes into some pipe. Now, oddly enough, and I this probably there actually is a protective pipe for going from the sidewalk to our houses. So that's nice. Matt Ray (01:01:23.813) Yeah. Matt Ray (01:01:28.129) I see it. my gosh. Coté (01:01:43.689) But the whole rest of it Brandon (01:01:44.221)

So I guess there is technically a tiny bit of conduit, although it kinda ends. It looks like at the end, very end, right? 'Cause the orange is conduit. Coté (01:01:49.758) Yeah, I mean I mean there's all sorts of other there's all sorts of other cables in there, but if you look on both sides of the street, so KPN, let's see, Odeo has these teal colored ones. I I get them mixed up, but so there's multiple, and sometimes all the little Ethernet cords are bundled together with an industrial zip tie, I guess. But like, so they laid these out as far as I can tell, and then they just put sand on top of it. And so, of course, if you've got some affordable outsourced Brandon (01:02:05.811) Yeah. Matt Ray (01:02:08.983) Wow. Coté (01:02:18.451) gas company people that come with shovels. Like Yeah. Yeah. And and so it's just like I I I f I found my fiber problem. It's just just an Ethernet cord. Brandon (01:02:21.329) Yeah. I mean they're gonna hit it. It's almost impossible not to, right? Like someone have to come out and market for them and then have to like really be super careful. Matt Ray (01:02:22.853) They're gonna hit that for sure. Brandon (01:02:31.005) You know what's funny about this whole thing though is like I think your situation there that's tough. I can see why people make the mistakes. But you know what's funny is like Matt Ray would kill for the setup. You know? Matt Ray would absolutely if Matt Ray had a s a a beautiful piece of fiber that ran from his curb right up to his house, he would he'd be in heaven. So so just say Kote, one man's problems is another man's dream. So Matt Ray (01:02:38.654) No, no, I have that set up. No. Coté (01:02:39.199) Yeah, yeah. Fiber. Matt Ray (01:02:48.162) I I I'm not I'm not Coté (01:02:50.581) That's true. Matt Ray (01:02:51.606) No, I I've I've seen this kind of stuff under my own house. Like I know I I I know that like I I've got this lovely stuff coming from my wall where it hits the the router and everything gets nice and clean from there. But I I I just know like on the opposite side, there's probably like two feet away there's a pile of dirt with a cable hanging out of it and they like sp exposed wires and it's like, you know what? A spider breathed on that cable and now Matt's having packet loss, right? That's Coté (01:03:13.055) Absolutely. Coté (01:03:19.007) Yeah, this this whole this whole this whole economy of trillions of dollars is just built on a foundation of sand, is is what's happening here. Brandon (01:03:26.621) Well I hope before AI takes it over that they send some robots out there to dig up Matt Ray's cable, give him a new one, put some conduit on it, and then just run it down to the curb and like make it great. Matt Ray (01:03:31.768) Well that's the thing, it's But but running it down to the curb is like three hundred feet. So it's like, god. Well then they gotta like dig through trees and rock and I if there's if there's con if it was done right, there'd be conduit and then they could just pull the cable through. And so I know at some point this is gonna suck. I do have Coté (01:03:36.357) He's Brandon (01:03:39.229) Well, now for the robots, that's no big deal. They don't care. They're they're like 300 feet, 3,000 feet. Then they make it happen. The robots will be great. They'll be like, yeah, we'll make it happen. Brandon (01:03:54.429) You know, there's a lot of things I was unsure about on this podcast, but the fact I am sure 100% that it was not done right and there is no conduit, that you just have like some stray cable, probably probably like wrapped around a root inside a tree, up, you know, with a kangaroo, you know. I mean, it's like all kinds of all kinds of crazy things. Well, listen, if you made it this far, thank you for surviving Matt's internet. Don't worry next week when we're on again and I complain about it. I'm sure it will happen. But until then. Coté (01:03:54.612) I d Matt Ray (01:04:00.78) I agree. I agree. That's what I'm saying, like Coté (01:04:05.877) Yeah. Matt Ray (01:04:07.052) Yes. There's a bandicoot chewing on it. Yeah, exactly. Matt Ray (01:04:20.649) They're Brandon (01:04:21.767) Thank you for streaming with us. Thank you for watching us live. And we always appreciate you watching us on the YouTube after the show. And with that, we will talk to you next Matt Ray (01:04:32.345) There is a national