WEBVTT 00:00:00.000 --> 00:00:02.340 Hello, can you hear me? Hey Vincent. 00:00:02.340 --> 00:00:08.040 How was your weekend? Yeah, good, good, uh, and yourself. 00:00:08.100 --> 00:00:08.520 How you doing? 00:00:09.420 --> 00:00:15.180 Uh, what did I do? Oh, also, I decided that 00:00:15.180 --> 00:00:20.040 I did not want to touch the computer over the weekend, or at least, at least not 00:00:20.040 --> 00:00:23.940 for anything work related and just for whatever I wanted to do. 00:00:24.000 --> 00:00:29.940 So I ended up on Friday working until, um, whenever, 00:00:29.940 --> 00:00:35.040 It's, um, you know, working until I reach the milestone that I set out for the 00:00:35.040 --> 00:00:38.820 week, right? Normally on Friday afternoon, it's easy to say, oh, you know what? 00:00:38.880 --> 00:00:40.560 I'll just catch up on the weekend. 00:00:40.620 --> 00:00:44.340 And then you end up, you know, getting distracted and then spending on me. 00:00:44.400 --> 00:00:49.740 So I sent, I decided to do it, finish it, and then I ended up working until 5 AM, 00:00:49.740 --> 00:00:53.340 Saturday morning. And then I was used this all weekend. 00:00:53.460 --> 00:00:58.500 I've embraced the other approach by just letting AI into my life at any point. 00:00:58.680 --> 00:01:05.220 And uh, with Orca, I'm on the phone, driving my home 00:01:05.220 --> 00:01:11.760 assistant, Automations and uh, improving them and, 00:01:11.760 --> 00:01:17.340 I mean, to be honest, Astra 6 has changed my, well, it sounds 00:01:17.340 --> 00:01:21.180 a bit too crazy to say this, but it's, it's pretty, it's pretty much changed my 00:01:21.180 --> 00:01:23.760 life in the last, uh, like 10 days. 00:01:23.820 --> 00:01:28.740 I mean, Astra 6 is, I'm really enjoying OpenAI’s, Astra 6. 00:01:29.400 --> 00:01:32.460 The results are just amazing to me. 00:01:32.520 --> 00:01:38.220 I'm just, I'm just in awe. So I'm like, you 00:01:38.220 --> 00:01:43.140 know, I have like more than a 100 GitHub projects on my, uh, on my GitHub profile. 00:01:43.200 --> 00:01:44.220 I't know if you've ever looked there. 00:01:44.280 --> 00:01:47.700 I mean, it's a dark place. It's a lot of... 00:01:47.700 --> 00:01:52.620 100 what on your on your... 100 what? 00:01:52.680 --> 00:01:55.500 There's a 100 like get repositories on my GitHub. 00:01:55.620 --> 00:02:01.380 Oh, like origin, not fork. Yeah, many of them, many of them suck, 00:02:01.380 --> 00:02:07.080 but like it's so fun and rewarding to just say, hey, Astrok, how would you improve 00:02:07.080 --> 00:02:11.340 this project? And it goes, yeah, no, no, no, no, no, no, and then it goes through 00:02:11.340 --> 00:02:16.440 it. You know, it really gives some great recommendations and I say, yeah, I think 00:02:16.440 --> 00:02:18.120 you should do this and then you should do that. 00:02:18.180 --> 00:02:23.820 Um, yeah, so, I'm, I'm, I'm loving Astra 00:02:23.820 --> 00:02:27.120 6 is incredible. Okay. interesting. 00:02:27.180 --> 00:02:33.180 I didn't try Astra 6 Astra 6. Um, I, 00:02:33.180 --> 00:02:39.120 I'm still, you know, content with Claude, 00:02:39.120 --> 00:02:44.820 and to be honest, it, it was quite an amazing because my laptop, 00:02:44.820 --> 00:02:49.080 my MacBooks, they run cafe, Nate, which keeps them awake and they don't sleep. 00:02:49.140 --> 00:02:55.080 Um, yeah. And so that actually kills, if they are not on power, 00:02:55.080 --> 00:02:57.000 they die, they drain very quickly. 00:02:57.060 --> 00:03:02.580 So I brought home another power adaptor today so that both of them 00:03:02.580 --> 00:03:07.260 can stay online. And what happens is that Claude Code, like I said, I think last 00:03:07.260 --> 00:03:11.340 time it automatically enabled remote control, but that does mean that whenever 00:03:11.340 --> 00:03:16.560 I, wherever I am, I get, I get a notification, like, Claude, Claude is stuck. 00:03:16.620 --> 00:03:21.660 And the only things that I, that stop cloud right now is it'll be as temporary 00:03:21.660 --> 00:03:27.480 credentials expiring. Yeah, to be honest, I've 00:03:27.480 --> 00:03:29.580 gone back to access keys for that very reason. 00:03:29.640 --> 00:03:32.100 And I feel like I feel like someone's kind of... 00:03:32.100 --> 00:03:33.240 It's a skis with AI. 00:03:34.140 --> 00:03:40.020 Oh my god. Oh, no, I've got I'm totally red pilled 00:03:40.020 --> 00:03:41.700 or something like that. 00:03:42.600 --> 00:03:43.980 Yeah. 00:03:44.940 --> 00:03:46.920 my personal account anyway. It's just whatever. 00:03:47.640 --> 00:03:51.240 I should have an agent to watch the spend it every any one time. 00:03:52.200 --> 00:03:57.960 The, The last podcast, I think I was bemoaning Orca 00:03:57.960 --> 00:04:02.100 doesn't have a way of me using it from another machine, but it actually does. 00:04:02.160 --> 00:04:08.040 So I run Orca, with all my stuff on this machine, which is permanently plugged 00:04:08.040 --> 00:04:13.860 in to power and ethernet. Then I use my other laptop when I'm at home, whatever, 00:04:13.860 --> 00:04:19.380 and then I just connect to all my sessions through the desktop app. 00:04:19.440 --> 00:04:25.080 And as well as, as well as the uh, the orca, uh, iOS app. 00:04:25.140 --> 00:04:30.000 So I have like, I have this. I have this, I have like a really nice, like, I'm 00:04:30.000 --> 00:04:35.220 organised feeling, because usually I'm working on a client laptop and I have like 00:04:35.220 --> 00:04:41.040 a 1000000 tabs open and I'm like, and I'm like losing track of my, of my life. 00:04:41.100 --> 00:04:45.180 Now Orca has completely organised everything and it's just amazing. 00:04:45.240 --> 00:04:45.780 I love it. 00:04:47.400 --> 00:04:52.200 Interesting. Orca. But like Orca to me sounds like Kiro. 00:04:52.260 --> 00:04:58.080 Kind of. You gotta try. I mean, there's a lot, there's 00:04:58.080 --> 00:05:03.480 a lot, the orca offers, I feel, that, that just, I mean, it's just got a lot of 00:05:03.480 --> 00:05:05.340 polish. I wouldn't say it's got a lot. 00:05:05.400 --> 00:05:08.280 It's just like when you use it, it's just streamlined. 00:05:08.340 --> 00:05:14.100 It's nice. I love it. Okay. Yeah, then the 00:05:14.100 --> 00:05:20.280 most powerful thing for me was that serendipitously or coincidentally, when 00:05:20.280 --> 00:05:25.620 I 1st tried Orca, it has like a, like, it opened a Claude tab and then opened up 00:05:25.620 --> 00:05:28.560 a Codex tab, and it was just like one click, and I, and I did that. 00:05:28.620 --> 00:05:33.720 I opened up a Claude tab, opened up a Codex tab, and then I just tried the same prompt 00:05:33.720 --> 00:05:39.480 in both, and I was just, and then when it does that, it opens up a separate isolated 00:05:39.480 --> 00:05:42.120 work tree. And I was, I was stunned. 00:05:42.720 --> 00:05:48.120 Absolutely stunned that like that some, in some cases, Codex was doing better than 00:05:48.120 --> 00:05:51.120 Claude. And in some cases, Claude was doing better than Codex. 00:05:51.180 --> 00:05:56.340 So then, so now I feel like I got a superpower, because now with Orca, I can say, 00:05:56.340 --> 00:05:58.860 oh, I want I want these changes, not these changes. 00:05:58.920 --> 00:06:00.420 I want these changes, not these changes. 00:06:00.480 --> 00:06:03.540 And, yeah, without getting into details. 00:06:03.600 --> 00:06:08.520 I'm just putting the models against one another, and that can be, that's just amazing 00:06:08.520 --> 00:06:14.340 to me. Cool. Um, yeah, I guess, like the reason that I mentioned Kiro 00:06:14.340 --> 00:06:19.200 and remote control on Claude Code is because, first, yeah, Kiro, I'm not really 00:06:19.200 --> 00:06:23.220 using it a lot. using it at all, actually, and I feel I'm missing out. 00:06:23.280 --> 00:06:25.200 I'm still having a lot of Hermes. 00:06:25.260 --> 00:06:27.240 Like Hermes is doing all the reviews for me. 00:06:27.600 --> 00:06:30.900 I just added stacked PR support. 00:06:30.960 --> 00:06:36.600 That means that if PRs get stacked, Hermes is reviewing each PR as 00:06:36.600 --> 00:06:40.860 part of the stack because otherwise it would review the like the 3rd PR in the 00:06:40.860 --> 00:06:42.720 stack against the page. What is this? 00:06:42.900 --> 00:06:44.520 I don't think I've used stack PR. 00:06:44.580 --> 00:06:50.220 So stack PRs. So basically what happens with AI is that your PRs 00:06:50.220 --> 00:06:55.920 get really big, right? Um, and and Opus already like months ago 00:06:55.920 --> 00:07:00.300 was able to slice down the commits really nicely, but ultimately when you're reviewing 00:07:00.300 --> 00:07:05.820 a PR as a as an open source or a project maintainer, the PR is really just like, 00:07:05.820 --> 00:07:08.880 you don't usually go commit by commit. 00:07:08.940 --> 00:07:13.560 I mean, it depends, right? Well, I get my slot centres to carve it up because it 00:07:13.560 --> 00:07:16.200 can't go over 100 lives. Enough about that. 00:07:16.260 --> 00:07:22.080 You don't need that anyway. So anyway, what I was saying is that like Opus is slicing 00:07:22.080 --> 00:07:26.760 up comets pretty nicely. they're reviewable and but I don't think a lot of people 00:07:26.760 --> 00:07:29.400 review commits like commit by commit. 00:07:29.520 --> 00:07:35.340 Um, so what GitHub released. is a concept of stacking PRs. 00:07:35.400 --> 00:07:40.020 That means you can create multiple PRs and one PR stacks on top of the other. 00:07:40.080 --> 00:07:42.540 And you can get up CLI stack in it. 00:07:42.600 --> 00:07:47.460 It's an extension, actually, you have to get up extensions, install the stack extension, 00:07:47.460 --> 00:07:52.140 and it's an API capability. So you can do it on the UI as well, but you can use 00:07:52.140 --> 00:07:53.820 it with the get up CLI as an extension. 00:07:53.880 --> 00:07:57.540 You can then you need a stack and then you can stack multiple PRs together. 00:07:57.600 --> 00:08:03.180 That means the 1st PR, and then you see the PRs as part of a collection. 00:08:03.300 --> 00:08:06.600 Like if you go into, I can show you... 00:08:06.600 --> 00:08:11.700 I got deja vu. The problem I have with that is that is that AI tends to make put 00:08:11.700 --> 00:08:16.260 huge descriptions in. So then aren't you going to have the problem with like, like 00:08:16.260 --> 00:08:21.900 10 PR stacked with like, Tolstoy novel inside 00:08:21.900 --> 00:08:24.960 the descriptions. So let's look here. 00:08:25.020 --> 00:08:28.440 Here you have a 6 stack PR, which is... 00:08:28.440 --> 00:08:29.520 Wow, I've never seen this before. 00:08:29.580 --> 00:08:33.120 This is a 2 stack PR. So this one's maybe a little bit more interesting. 00:08:33.180 --> 00:08:37.740 This is where after we did a release, there were a couple of errors reported in 00:08:37.740 --> 00:08:40.440 sentry, and I wanted to fix them. 00:08:40.500 --> 00:08:46.020 I had related ghetto issues, but I wanted like this to go 00:08:46.020 --> 00:08:49.140 in, and then the other one kind of dependent on this. 00:08:49.200 --> 00:08:51.420 So the other one couldn't merge until this one was in. 00:08:51.540 --> 00:08:57.480 So this one, if you go into the bottom here, you will actually see the 00:08:57.480 --> 00:09:01.980 stack, right? This is Maine. This is the 1st PR that you're looking at, and this 00:09:01.980 --> 00:09:04.380 is the 2nd PR that is stacked on this PR. 00:09:04.440 --> 00:09:09.900 So if you were to go in here, you again, at the top, you see this is 2 of two. 00:09:10.080 --> 00:09:12.420 You can see the full stack. 00:09:13.560 --> 00:09:15.600 There's a lot going on here. Vincent. 00:09:15.660 --> 00:09:18.480 And then you can see at the bottom here, this is the 2nd one as well. 00:09:18.540 --> 00:09:24.720 So isn't it better? I mean, I'm just I'm just comparing this approach to just breaking 00:09:24.720 --> 00:09:28.620 up your commits. I mean, Yeah, that's what I started with, right? 00:09:28.680 --> 00:09:34.260 I said, like, Opus already breaks up your commits, but, um, I think a 00:09:34.260 --> 00:09:39.840 PR on its own should be able to merge, um, and commit on 00:09:39.840 --> 00:09:44.100 its own might not. So I think the unit of work is different. 00:09:44.160 --> 00:09:49.440 Um, it also allows the team to look at like, I don't know if people actually go 00:09:49.440 --> 00:09:51.720 and look at like each individual commit. 00:09:51.780 --> 00:09:57.300 Um, it depends. I guess what GitHub clearly had the demand for 00:09:57.300 --> 00:10:03.240 this. I think the main the main demand here was to allow agentic 00:10:03.240 --> 00:10:08.880 workflows to create like larger, larger sets of work and then break them down. 00:10:09.300 --> 00:10:13.440 But then I have this 6 stack here, which honestly, I don't think is much easier 00:10:13.440 --> 00:10:14.160 to review. 00:10:17.100 --> 00:10:20.640 Can I assume that all 6 of them need to go in, right? 00:10:20.820 --> 00:10:24.420 Yeah, so I think I can. I cannot. 00:10:24.720 --> 00:10:29.340 Yeah, I cannot. I cannot merge an individual commit, it looks like. 00:10:29.400 --> 00:10:32.220 Okay, PR. I cannot merge on individual PR. 00:10:32.280 --> 00:10:37.800 So this one has like 5 commits. And all it does is it updates 00:10:37.800 --> 00:10:42.180 a dependency with a new one, like several major versions ahead. 00:10:42.240 --> 00:10:46.380 Um, and it's quite complicated and it sets up for the next PR. 00:10:46.440 --> 00:10:50.640 But what's good is what I find really good is that once this is green, right, once 00:10:50.640 --> 00:10:55.920 this is approved, you can just ignore because then you don't need to like rebase 00:10:55.920 --> 00:11:00.000 everything else again. Like this one is done and now you can focus and handle the... 00:11:00.060 --> 00:11:02.880 Wait a minute, that approval was done by an agent or... 00:11:02.880 --> 00:11:07.740 Yeah, it doesn't matter. It's clearly not approved by a human yet. 00:11:07.800 --> 00:11:12.360 So then there's like this one, it's also like, I'm not going to look at this again 00:11:12.360 --> 00:11:17.880 until, like, I guess the other maintainers are not reviewing it 00:11:17.880 --> 00:11:21.960 yet because they're waiting for the full stack to go green. 00:11:22.020 --> 00:11:24.480 So there's a couple of weird stuff, though. 00:11:24.540 --> 00:11:30.240 Like, this PR sets up some things and then the review says like, okay, this sets 00:11:30.240 --> 00:11:33.000 up for its load bearing for the next PR. 00:11:33.060 --> 00:11:36.300 So, like, it might not be important for this PR, but actually, it is very important 00:11:36.300 --> 00:11:40.260 if you take into context the fact that an XPR depends on some of these capabilities 00:11:40.260 --> 00:11:41.940 and therefore they should be there. 00:11:42.000 --> 00:11:44.700 They should be part of this PR. So, yeah. 00:11:44.760 --> 00:11:49.080 I think that's kind of an interesting way of slicing up your work. 00:11:49.140 --> 00:11:52.200 I think sex PR. So when I, how do we end up in this? 00:11:52.200 --> 00:11:54.120 Because you started, you keep talking about Orka. 00:11:54.180 --> 00:11:58.380 I said, for me, Hermes is still the workhorse that is doing most of my stuff. 00:11:58.440 --> 00:12:01.740 I just landed PR stacked PRs support. 00:12:01.800 --> 00:12:04.200 We've talked long enough about stack PRs. 00:12:04.260 --> 00:12:09.840 What I wanted to get to was that Claude Code, remote code, uh, remote control, kind 00:12:09.840 --> 00:12:15.060 of has given me, like I, when I'm on my phone, there's 2 things. 00:12:15.120 --> 00:12:19.980 Either it's Hermes pinging me on discord with a new review of multiple like PRs 00:12:19.980 --> 00:12:24.180 coming in, or it's Claude Code pinging me in the Claude Code app. 00:12:24.240 --> 00:12:25.800 And the Claude Code app is amazing, actually. 00:12:25.860 --> 00:12:31.860 Now, the web app has this, when you look at an agent, you can see the quote diff 00:12:31.860 --> 00:12:36.540 as well in the browser as well. I had my laptop at home working on something. 00:12:36.660 --> 00:12:39.840 I was in the office and I was able to see the work going. 00:12:39.900 --> 00:12:45.660 That was pretty cool. I noticed like that sometimes the code 00:12:45.660 --> 00:12:47.640 sessions sometimes can show the diffs. 00:12:47.700 --> 00:12:51.960 But on the topic of discs, I've, I've been in the habit of using this programme 00:12:51.960 --> 00:12:57.420 called TIG, TIG, so... I switched to Lazygit, though. 00:12:57.480 --> 00:13:02.760 Ah, it's got too much. Lazy get, I like, I think it's just a pager and lazy get 00:13:02.760 --> 00:13:07.140 is like a whole ecosystem, which I can't be bothered to learn the key strokes for. 00:13:07.200 --> 00:13:09.420 But Tig didn't work on certain architecture. 00:13:09.480 --> 00:13:12.600 Like they didn't build and publish certain architectures, if I remember. 00:13:12.720 --> 00:13:18.480 Yeah, tickets kind of old. And, and so basically, I use take 00:13:18.480 --> 00:13:25.020 and take status to basically base up to stage and push my 00:13:25.020 --> 00:13:30.540 hunks or whatever, and do my commits, everything with this, which is take status. 00:13:30.600 --> 00:13:34.560 But then when I installed Omarchy, which uses Lazygit, and I didn't want to do 00:13:34.560 --> 00:13:40.440 additional work of installing Tig, I just aliased the Lazygit 00:13:40.440 --> 00:13:44.460 status, and I've got pretty much exactly the same thing that I was doing with Tig. 00:13:44.520 --> 00:13:46.380 Maybe I just need to learn something new. 00:13:46.440 --> 00:13:50.700 There's another one that I use and take would take blame and then the path, and 00:13:50.700 --> 00:13:52.740 then I can see what changes were made by who. 00:13:52.860 --> 00:13:58.380 But that, I mean, VS code gives me that with the, um, if you, if 00:13:58.380 --> 00:14:00.420 you use the git lens and extension. 00:14:00.480 --> 00:14:02.580 Oh, I can't stand that extension. 00:14:02.640 --> 00:14:04.740 It's too much. It's too much for me. 00:14:04.800 --> 00:14:08.640 You don't need to. Like, the only thing I use it for is whenever I go over a line, 00:14:08.640 --> 00:14:13.380 it hoovers up, which, which author, who is the last one for the set line, when 00:14:13.380 --> 00:14:17.340 you hoover it, you can see the PR that touched the line, you can go to the file 00:14:17.340 --> 00:14:20.940 history and it goes on into the separate extension. 00:14:21.000 --> 00:14:25.020 I do like that. But I just can't be bothered with extensions anymore. 00:14:25.620 --> 00:14:29.160 Oh, let's talk about... It bothers with VS code. 00:14:29.220 --> 00:14:33.960 Anyway. Yeah, so that's another good thing about Orca is that I don't have to jump 00:14:33.960 --> 00:14:39.540 into VS code. It has like a basic sort of VS code editor differ thing 00:14:39.540 --> 00:14:42.420 in there and it works for me. I'm happy. 00:14:42.480 --> 00:14:48.240 I feel like I feel like the web application of Claude Code now kind of starts to 00:14:48.240 --> 00:14:54.000 be exactly what I used an ID for, like which was to look at the diffs that 00:14:54.000 --> 00:14:57.780 the agent had been doing so that I can have an overview of what it's been doing. 00:14:57.840 --> 00:15:03.780 And I mean, I was presently surprised today when I had a laptop at home building 00:15:03.780 --> 00:15:08.760 something and I was able to just go to the web app or my phone, but my phone is 00:15:08.760 --> 00:15:10.440 not very nice to look at the get tips. 00:15:10.500 --> 00:15:14.520 But still, it's quite interesting that they have collapsed everything and you can 00:15:14.520 --> 00:15:18.000 really dive into, hey, I want to really see what's the difference on this particular 00:15:18.000 --> 00:15:21.360 file. This experience, yeah. Yeah. 00:15:21.420 --> 00:15:26.760 So anyway, that was the whole topic of Orca and being remote and using remote. 00:15:26.820 --> 00:15:31.440 Yeah, using, I think they're called agent development environments, ADEs. 00:15:31.500 --> 00:15:35.160 I mean, I mean, Claude has has, I mean, Anthropic has their own one. 00:15:35.220 --> 00:15:38.940 I think I think Codex is pretty much caught up. 00:15:39.600 --> 00:15:42.300 I was pretty upset with my colleague this morning. 00:15:42.360 --> 00:15:46.320 I told him this is actually, this turned my laptop into open claw. 00:15:46.380 --> 00:15:51.900 Cloth code, just behind my back, turned my laptop into open cloth. 00:15:51.960 --> 00:15:53.760 The whole reason I never install open cloth. 00:15:53.820 --> 00:15:57.540 on my laptop is because I don't want an agent to be managing my machine. 00:15:57.600 --> 00:16:03.120 And what happens is now I'm at office and I have an 00:16:03.120 --> 00:16:08.460 agent telling me, yeah, I ran it on my laptop, it's like, that's not your laptop, 00:16:08.460 --> 00:16:13.740 that's my laptop. You know, you just, you just completely owned my laptop and I 00:16:13.740 --> 00:16:18.720 never approved remote control. So I think I think all of the organisations like 00:16:18.720 --> 00:16:24.240 Anthropic and Codex, they're moving and even Kiro CLI, they're turning these CLIs 00:16:24.240 --> 00:16:28.020 into basically agents that run on your machine and take ownership of your machine, 00:16:28.020 --> 00:16:32.160 which I don't like. I mean, I don't like, but I thought you were going to say take 00:16:32.160 --> 00:16:38.040 ownership of your life because surely when you when you clock off from work, as 00:16:38.040 --> 00:16:42.540 I was saying last week, when you're brushing your teeth, you get a ping for to 00:16:42.540 --> 00:16:47.160 do some work thing or, you know, and then you're going, nah, I don't want to be 00:16:47.160 --> 00:16:50.640 bothered with this sort of stuff while I'm trying to look into my children's eyes. 00:16:51.180 --> 00:16:54.840 You're brushing your teeth while looking into your children's eyes. 00:16:54.900 --> 00:17:00.660 That's is that a power move? I'm trying to say that these AI notifications are 00:17:00.660 --> 00:17:03.060 coming out of our hours and it's a bit insane. 00:17:06.060 --> 00:17:11.160 But I feel... Okay, let's talk about Omarchy. 00:17:11.280 --> 00:17:14.880 I think we're done with ADEs. We've brought it up. 00:17:15.000 --> 00:17:18.000 going into a sensitive area there with too much. 00:17:18.060 --> 00:17:23.880 Yeah. A colleague shared with me some crazy neo-cities 00:17:23.880 --> 00:17:27.300 page about how people are anti, um, a march. 00:17:27.540 --> 00:17:31.380 Is this real because DHH shared it and I thought it was a mockery. 00:17:31.440 --> 00:17:32.700 I'm not sure if it's is it real? 00:17:32.760 --> 00:17:35.640 I don't know. I mean, it looks kind of real. 00:17:35.700 --> 00:17:39.720 It's hard to know nowadays I saw it posted either on, either he posted it on LinkedIn 00:17:39.720 --> 00:17:43.200 or he posted on the on March's sub on Reddit because I'm, I followed that sub because 00:17:43.200 --> 00:17:48.960 I'm crazy about what they do. There was also Justin Garrison, who I kind 00:17:48.960 --> 00:17:51.120 of really, you know, appreciate it. 00:17:51.480 --> 00:17:56.580 Yeah, he does a lot of like, he has a lot of visualisation, visual explanations 00:17:56.580 --> 00:17:59.340 on how communities, like pod scheduling works. 00:17:59.400 --> 00:18:03.600 I think he works on talos. I'm not sure if I'm confusing people now, but I respect 00:18:03.600 --> 00:18:09.840 him. And he posted a very public message about, well, 00:18:09.840 --> 00:18:13.020 I'm cancelling DigitalOcean because they decided to sponsor a racist. 00:18:13.620 --> 00:18:19.020 I'm like, so I had this whole conversation with ChatGPT about where is it, like, 00:18:19.020 --> 00:18:26.400 acceptable to attribute political opinions to an open source contributor, 00:18:26.400 --> 00:18:31.320 and where do you separate political, you know, visions? 00:18:31.320 --> 00:18:37.320 Like, so basically ChatGPT helped me understand that if, so if you 00:18:37.320 --> 00:18:42.960 supporting that project results in financial gain, for somebody expressing such 00:18:42.960 --> 00:18:46.980 visions and you do not agree and you do you want to protest, then you should indeed 00:18:46.980 --> 00:18:52.500 not like avoid using it. But in the case of a mark she, there 00:18:52.500 --> 00:18:56.760 is no financial gain, right? So that was the, this, that was the hold on. 00:18:56.820 --> 00:18:58.380 Well, I mean, that was the hold on. 00:18:58.440 --> 00:19:02.880 Hold on. That was the state 4 months ago, right? 00:19:03.000 --> 00:19:08.940 So, so DHH, he built a marchy, you know, basically Arch Linux, um, bunch 00:19:08.940 --> 00:19:13.200 of shell scripts, literally a bunch of shells, uh, and a bunch of Golang and maybe 00:19:13.200 --> 00:19:17.700 some agents. MD for your agents to help you manage your your operating system. 00:19:17.760 --> 00:19:19.920 It's not much, right? And a separate maybe. 00:19:20.520 --> 00:19:26.280 Very, very good, uh, Very good design, if you ask me, 00:19:26.280 --> 00:19:29.460 very good implementation in a way, like the way. 00:19:29.640 --> 00:19:32.880 Or much is the thing, right? Is it a distro? 00:19:32.940 --> 00:19:36.240 I'm not sure. It's kind of like it's, it's, it's a bunch of shell scripts that 00:19:36.240 --> 00:19:38.820 you installed a couple of, of, of Arch Linux. 00:19:38.880 --> 00:19:44.520 And it makes, he said that his goal for his company was 00:19:44.520 --> 00:19:47.940 like, we don't want to pay Apple for overpriced hardware. 00:19:48.000 --> 00:19:53.520 We can get nice hardware and Linux is ready for the big, 00:19:53.520 --> 00:19:58.380 like, extreme. And, and, and he basically said, you, oh, it sucks. 00:19:58.440 --> 00:20:02.880 macOS sucks. Here you go. You know, you use, what is it? 00:20:02.940 --> 00:20:08.580 Wayland, window manager. Well, I just, I just love how, like, Codex 00:20:08.580 --> 00:20:11.760 is so much faster on my Arachi laptop. 00:20:11.820 --> 00:20:17.160 Another thing, I think there's a lot of like benchmarks on file systems or performance 00:20:17.160 --> 00:20:22.860 on Arch Linux compared to macOS that actually the 00:20:22.860 --> 00:20:25.380 performance on macOS file system is abysmal. 00:20:25.440 --> 00:20:27.540 I think it's again... It's terrible. 00:20:27.600 --> 00:20:30.480 I mean, it's the same story with Windows, actually. 00:20:30.720 --> 00:20:35.340 Okay, so we agree that Omarchy is kind of a good thing in a way that it really promotes 00:20:35.340 --> 00:20:39.900 Linux on the laptop and, you know, it kind of, you know, for people that are maybe 00:20:39.900 --> 00:20:43.080 not not, it's an entry, um, it's a gateway. 00:20:43.140 --> 00:20:48.060 It's a gateway for people that might not be inclined to adopt arch because they 00:20:48.060 --> 00:20:50.520 don't want to tell everyone they use arch, by the way. 00:20:51.360 --> 00:20:57.060 Because they spent several 4 weekend, 4 weekends to make it work. 00:20:57.120 --> 00:20:59.760 So now they have to make it make it visible. 00:20:59.820 --> 00:21:05.460 So, so, Omarchy kind of, like, burns an, an ISO, you pop in 00:21:05.460 --> 00:21:09.660 an USB stick, and before you know it, you, you, you have a useable machine, and 00:21:09.660 --> 00:21:12.780 all of your macOS capabilities are kind of at your fingertips. 00:21:12.840 --> 00:21:17.760 You didn't have to do it. Yeah, yeah, the way I understand it is DHHH, uh, doesn't 00:21:17.760 --> 00:21:22.800 want to be a slave to Apple, and he has, he runs a company, a successful company. 00:21:22.860 --> 00:21:28.440 He has employees, and if he can run his own OS, which he 00:21:28.440 --> 00:21:32.280 and get everyone to use it, then he's winning. 00:21:32.340 --> 00:21:35.460 And I think that's a, that's like, why doesn't other people do this? 00:21:35.520 --> 00:21:41.040 Yeah, yeah. And I mean, and and kind of what makes me interested in it is 00:21:41.040 --> 00:21:43.800 that he stepped forward and says, look, every single person in my organisation 00:21:43.800 --> 00:21:46.860 is using this. So it is for wording. 00:21:46.920 --> 00:21:49.500 Okay? So to me, it's like great. 00:21:49.560 --> 00:21:53.340 I want the simple Linux setup because I don't want to run a whole virtual machine 00:21:53.340 --> 00:21:54.900 just to have a Docker demon. you know? 00:21:54.960 --> 00:22:00.960 Yeah, I'm sure. I'm sure things will change quite quickly, like, because like 00:22:00.960 --> 00:22:05.400 device management, MDM is required by any bigger organisation to be compliant. 00:22:07.680 --> 00:22:12.480 Yeah, so back to the, because you, you, you talk about on, on Archie, you want 00:22:12.480 --> 00:22:17.220 to talk about the Archie? I step back, where I say the political vision, as long 00:22:17.220 --> 00:22:22.680 as it is not financial gaining to someone, you can use the project without guilt. 00:22:22.680 --> 00:22:26.700 And I was happy with ChatGPT, I was like, great, I can use Comachi, right? 00:22:26.760 --> 00:22:29.220 Not that I would care too much. anyway. 00:22:29.280 --> 00:22:34.980 So now this whole Omacom thing happened, 00:22:34.980 --> 00:22:39.120 right? called Omacom, right? Because I have Oma, Omacom, uh, it's it's like a 00:22:39.120 --> 00:22:40.620 community fund. What's it called? 00:22:40.680 --> 00:22:46.320 Yeah, I haven't... I wrap my head around the the terminology, 00:22:46.320 --> 00:22:50.400 yeah. But there's some backing. Yeah, some money. 00:22:50.460 --> 00:22:56.340 So Omacom is a foundation. I take a non-profit foundation that sponsors on 00:22:56.340 --> 00:23:02.040 Omarchy. Um, because they have Omar Cube, which was the Ubuntu version, 00:23:02.040 --> 00:23:06.360 and then you have Oma Omar Chi, which was the Arclinus version, then that they 00:23:06.360 --> 00:23:10.680 have Omacom, which is the foundation that, you know, manages the communities of 00:23:10.680 --> 00:23:13.380 the different Oma, um, flavours. 00:23:13.440 --> 00:23:17.580 And now they created a foundation that has several companies backing it. 00:23:17.640 --> 00:23:21.540 People pledging. I think six. I think initially when they announced they had 8 00:23:21.540 --> 00:23:26.580 or six, um, like a six, each one a million, like DHH within a 1000000, Spotify 00:23:26.580 --> 00:23:29.220 puts in a million. I think another one. 00:23:29.280 --> 00:23:32.640 don't know. And then DigitalOcean stepped up and also says we're committing like, 00:23:32.640 --> 00:23:34.620 I don't know how much to this foundation. 00:23:34.680 --> 00:23:39.540 Wow. And this is where Justin Garrison posted on on LinkedIn. 00:23:39.600 --> 00:23:42.540 I guess DigitalOcean is fine with sponsoring racism. 00:23:42.600 --> 00:23:48.120 Yeah, this... And then a picture, a screenshot of him cancelling 00:23:48.120 --> 00:23:51.120 the account. And then I went back to my ChatGPT, you know? 00:23:51.120 --> 00:23:56.400 I was like, how does this, like, how does, how, why, why is this person that I 00:23:56.400 --> 00:24:02.040 respect and follow? Um, so tightly combining, like coupling whatever DHA says, 00:24:02.040 --> 00:24:07.740 which I don't agree with, um, then does not want to even 00:24:07.740 --> 00:24:12.060 use DigitalOcean because they are backing a nonprofit foundation. 00:24:12.120 --> 00:24:16.320 And ChatGPT was very quick to tell me, but now he's financially gaining from it. 00:24:16.380 --> 00:24:20.100 Right now they are having employees and now they are paying people. 00:24:20.220 --> 00:24:23.100 So therefore, we have crossed the threshold. 00:24:23.220 --> 00:24:29.280 And I was like, oh, man, but... Yeah, 00:24:29.280 --> 00:24:31.800 but the but that seems to be... 00:24:32.760 --> 00:24:38.880 What do you call it, perpendicular or adjacent to people calling DHS DHH 00:24:38.880 --> 00:24:42.900 some sort of right wing, um, you know, like. 00:24:43.140 --> 00:24:50.100 You know, like, you know, the blog post he wrote is, you know, is, 00:24:50.100 --> 00:24:55.380 I mean, if you go back to Nazi, uh, Nazi Germany, and you look at the propaganda 00:24:55.380 --> 00:25:00.240 that the Nazis put out against the Jews, it's the same verban. 00:25:00.300 --> 00:25:02.820 It's the same words. Come on. I mean... 00:25:02.880 --> 00:25:06.180 What is it? Wolves among sheep. It's like... 00:25:06.180 --> 00:25:10.620 I think a lot of, I think a lot of text, even, like, I find it weird. 00:25:10.680 --> 00:25:16.500 Like, there's a lot of political discourse nowadays, like, like, 00:25:16.500 --> 00:25:21.780 you deserve this. And I never forget when I visited a concentration camp, which 00:25:21.780 --> 00:25:27.240 was a harrowing experience, um, there was a German thing like uh, Arbeit Macht 00:25:27.240 --> 00:25:30.600 Frei. Well, no, about my fry? I don't know, I don't know. 00:25:30.660 --> 00:25:34.020 I don't know what the Germans were, but in English it was something like, you get 00:25:34.020 --> 00:25:35.460 what you deserve or something like that. 00:25:36.420 --> 00:25:41.940 Hmm. And so basically, I feel like, yeah, let's 00:25:41.940 --> 00:25:47.880 not go, but the thing that shocks me with the marchy is that, people just are just, 00:25:47.880 --> 00:25:53.460 very negative against it. And, and, uh, they might, 00:25:53.460 --> 00:25:56.640 they might say the right, it might, they might say, oh, he's got right wing thing. 00:25:56.700 --> 00:26:01.080 Uh, they might say, like, what was the other thing that uh, someone was complaining 00:26:01.080 --> 00:26:04.800 about, okay, the money thing. I mean, that's that's a difficult thing. 00:26:04.860 --> 00:26:09.900 I think there was a colleague of mine who was like, oh, I don't like DHH anymore 00:26:09.900 --> 00:26:12.240 because he's anti-tester and development. 00:26:12.600 --> 00:26:18.000 And I'm like, um, like, you know what I mean? 00:26:18.060 --> 00:26:23.640 The problem I have is that, like, people, To be honest, if I 00:26:23.640 --> 00:26:28.260 don't agree with somebody. I would say like, okay, cool, you're crazy. 00:26:28.320 --> 00:26:30.660 You know, that's probably the worst thing. 00:26:30.720 --> 00:26:32.880 You told me that. You said, you said, I was crazy. 00:26:32.940 --> 00:26:34.440 That means you didn't agree with me. 00:26:34.500 --> 00:26:38.940 It means I don't agree with it. No, I don't agree with you on that every point. 00:26:39.000 --> 00:26:40.980 That's not very respectfully calling Creek. 00:26:41.040 --> 00:26:43.860 Yeah, well, okay, it's not very respectful, but that's the worst thing I can say 00:26:43.860 --> 00:26:49.380 that you're mad, okay? But like, I wouldn't, I wouldn't 00:26:49.380 --> 00:26:51.240 say that like, hey, I don't like you. 00:26:51.300 --> 00:26:53.700 I just think it's so immature. 00:26:54.780 --> 00:27:01.560 And that's, yeah, it's like, So, 00:27:01.560 --> 00:27:06.240 so I was about to comment on Justin's Garrison post something like, um, I think 00:27:06.240 --> 00:27:11.820 we should be able to decouple the political, uh, vision or, or, like, political 00:27:11.820 --> 00:27:16.560 statements from an open source contributor, honestly, um, who has created many 00:27:16.560 --> 00:27:19.800 open source projects that a lot of people uh, benefit from. 00:27:19.860 --> 00:27:24.600 A lot of people get into programming through Ruby on Rails, before AI, people created 00:27:24.600 --> 00:27:26.580 their businesses on Ruby on Rails. 00:27:26.640 --> 00:27:28.860 It was a nice, very quick way to start. 00:27:28.920 --> 00:27:32.880 Exactly. It was a gateway for programming. 00:27:32.940 --> 00:27:38.460 It was a lot of people. And now Omarchy is a gateway for people that want to 00:27:38.460 --> 00:27:42.060 get rid of windows and of macOS and just get like an operating system. 00:27:42.120 --> 00:27:44.100 DHH is doing God's work in a way. 00:27:44.640 --> 00:27:49.920 Wow. In one slice, right? Wow. Wow. 00:27:49.980 --> 00:27:54.240 Okay, sorry. He's not mad is what I wanted to say. 00:27:54.300 --> 00:27:58.560 He's very insane. Okay. I'm not gonna say anything. 00:27:58.620 --> 00:28:01.320 I have a lot of opinions about what you just said, but okay. 00:28:01.500 --> 00:28:06.600 Um, so... And I was about to comment. 00:28:06.660 --> 00:28:08.040 So I went through all the comments in it. 00:28:08.100 --> 00:28:11.040 And then I read, like, people were confused, like, what are you talking about? 00:28:11.100 --> 00:28:14.100 And then I, I, some people were like, I think it's related to this and that and 00:28:14.100 --> 00:28:18.840 then, and then I saw a comment, which was like, to be honest, and I kind of agree 00:28:18.840 --> 00:28:22.860 with that. And I saw this a lot from other people talking about the Marchi. 00:28:22.920 --> 00:28:26.040 It's just a bunch of shell scripts, not even good ones at that. 00:28:26.100 --> 00:28:28.740 That's what they were saying. It's a fact. 00:28:28.800 --> 00:28:34.320 I mean, honestly, some I think he's made a product out of a whole bunch 00:28:34.320 --> 00:28:37.320 of packages. And the product is impressive. 00:28:37.380 --> 00:28:41.460 I was, I'm impressed. It's fast to install, everything works. 00:28:41.520 --> 00:28:46.620 Yeah, agree. agree. And like having in front behind it because he also put his 00:28:46.620 --> 00:28:49.080 money and and created... Yeah, I mean, what's it called? 00:28:49.140 --> 00:28:53.400 I haven't wrapped my head around like the plug-in ecosystem, which seems to be 00:28:53.400 --> 00:28:58.380 a marketplace in itself. And fine, you know, fine, fine, fine. 00:28:58.440 --> 00:29:04.020 Uh, though, yeah, I think, yeah, I think we've talked enough about 00:29:04.020 --> 00:29:07.620 it. It's a great gateway and I just don't understand. 00:29:07.620 --> 00:29:12.420 Yes, some people's negativity. So the 1st part was it's a bunch of shell, and most 00:29:12.420 --> 00:29:17.400 people that install it end up moving on, saying like, hey, that was great, but 00:29:17.400 --> 00:29:22.080 you know what? I'm actually using I think I might be better off using art directly. 00:29:22.140 --> 00:29:24.240 I don't like all of the pre-bundled stuff that comes with it. 00:29:24.300 --> 00:29:25.980 And I think that's a great way on its own, right? 00:29:26.040 --> 00:29:29.820 I mean, now you've got a working machine and you start exploring further. 00:29:29.880 --> 00:29:33.000 Yeah exactly. Maybe we should be using PewDiePie. 00:29:33.120 --> 00:29:37.140 version of it then. It's pretty, but I may have OS now. 00:29:37.200 --> 00:29:42.600 Yeah, PewDiePie built the whole AI machine like you need to own your own inference. 00:29:42.660 --> 00:29:48.720 He says we shouldn't be slaves to, but he had his, he started with Linux moding, 00:29:48.720 --> 00:29:50.640 right? He started with going hard into Linux. 00:29:50.700 --> 00:29:55.260 After being famous for games. He went hard into Linux, and then he went hard... 00:29:55.320 --> 00:30:01.260 I like PDP a lot too. Yeah, I mean, I'm, it's, it's something I'm, I'm struggling. 00:30:01.320 --> 00:30:06.420 I tried to, you know, it, in the intro of my YouTube channel. 00:30:06.480 --> 00:30:09.960 I think I try to make a video, but I wasn't I didn't really like the video, so 00:30:09.960 --> 00:30:15.720 I I removed it or something. But like, my whole philosophy, with, with my computing 00:30:15.720 --> 00:30:21.660 has to be, has been, uh, you know, trying to be independent and trying to, um, 00:30:21.660 --> 00:30:27.180 foster... Well, I don't, I just, I just want to be 00:30:27.180 --> 00:30:29.580 independent. Like, I don't want to be a slave to Microsoft. 00:30:29.640 --> 00:30:32.400 I don't want to be a slave. Good luck with AI. 00:30:32.460 --> 00:30:37.740 Apple. But at the same time, I'm a bit of a heretic because I guess I'm just looking 00:30:37.740 --> 00:30:43.740 around. I have my Dell, I have my Dell monitors, but my phone and all the 00:30:43.740 --> 00:30:49.320 laptops inside are on Mac. Of course, out of sight, I have my thing pad and 00:30:49.320 --> 00:30:51.300 my raspberry pie and all that sort of stuff. 00:30:52.260 --> 00:30:55.440 Maybe I'm just, what do you call it? 00:30:55.500 --> 00:31:01.080 an absolute hypocrite, but I'm a big, Believer having your 00:31:01.080 --> 00:31:06.840 own infrastructure, and knowing how, you know, if you're on a desert island, 00:31:06.840 --> 00:31:11.760 like how, how to build it from scratch, is the way I think of it. 00:31:11.820 --> 00:31:14.280 Like, what do you need? Like, you need this, you need that. 00:31:14.340 --> 00:31:16.140 You need your terraform pipeline. 00:31:16.260 --> 00:31:19.620 No, you don't need your terrible pipeline. 00:31:19.680 --> 00:31:23.760 Or a different island. Stay away from terrifying. 00:31:23.760 --> 00:31:28.860 Of course, like, what I'm saying here in the light of the AI infrastructure podcast 00:31:28.860 --> 00:31:33.900 probably doesn't make a lot of sense to people because I'm pretty much addicted 00:31:33.900 --> 00:31:38.100 to frontier models by anthropic and OpenAI at this point. 00:31:38.160 --> 00:31:40.320 Now how does that make you independent? 00:31:40.380 --> 00:31:45.540 No, Kai. You're sucking on the teat of big corporate. 00:31:45.660 --> 00:31:47.940 So the... 00:31:51.060 --> 00:31:56.880 Yeah, no, I, I, um, when you were saying that, like being independent and and 00:31:56.880 --> 00:32:01.680 I was at the computer store this weekend to pick up a laptop that we dropped off 00:32:01.680 --> 00:32:05.520 a while ago to get fixed, like a motor port issue, then a battery issue, many issues 00:32:05.520 --> 00:32:09.360 with never buying. I mean, I've had my share of Lenovo problems. 00:32:09.420 --> 00:32:15.180 Oh, Anyway, um, I don't know for Legion. 00:32:15.240 --> 00:32:20.760 And basically on the desk, at the maintenance desk, there 00:32:20.760 --> 00:32:24.960 was this machine, which was a tower, but without a tower. 00:32:25.020 --> 00:32:28.260 It was just a motorboard with like the liquid cooling on one side. 00:32:28.380 --> 00:32:35.580 The GPU on the other side, and then the coolers on the GPU had these band 00:32:35.580 --> 00:32:40.200 displays. So you know how these, like, 0 my god, iPhone, I don't want to talk about 00:32:40.200 --> 00:32:44.880 iPhone 2. But you have these band displays now at all the rage, right? 00:32:44.940 --> 00:32:46.980 And it was a band display. Oh, yeah, yeah, yeah. 00:32:47.040 --> 00:32:48.240 Sorry. No, no, what do you mean? 00:32:48.300 --> 00:32:50.160 Bent displays. Okay, sorry, I was... 00:32:50.160 --> 00:32:53.640 Yeah, display that's not flat, but like... 00:32:53.640 --> 00:32:56.340 I'm pretty you were talking about cooling or something, but yeah, I don't get it. 00:32:56.340 --> 00:32:59.340 Oh, yeah, so on top of the cooling, they had this display. 00:32:59.400 --> 00:33:05.040 There was like playing videos of of like the Nvidia logo and other 00:33:05.040 --> 00:33:10.680 things. And the machine was, I don't know, it was 00:33:10.680 --> 00:33:14.280 360000000 Vietnam dong. How much is that in US dollars? 00:33:14.280 --> 00:33:19.980 Give me that ding dong. Jesus, Kai 00:33:19.980 --> 00:33:24.900 was... Oh, that's close to 14,000 US dollars, 14. 00:33:24.900 --> 00:33:27.060 Okay. 00:33:27.780 --> 00:33:31.020 And I wanted to ask them, like, how much of that? 00:33:31.140 --> 00:33:35.280 total is the GPU card? Just the GTX, right? 00:33:35.340 --> 00:33:40.800 RTX. And, um, my son asked them in Vietnamese and they explain, they told him it's 00:33:40.800 --> 00:33:46.800 like 30%. So 30% of that of the cost is the GPU. 00:33:46.860 --> 00:33:50.820 I guess the other, you know, 10% will be the memory. 00:33:50.880 --> 00:33:54.600 And then the rest is the motorboard, the CPU who cares about that and power supply. 00:33:54.660 --> 00:34:00.540 about the CV. Anyway, so so, and then what I thought was, and 00:34:00.540 --> 00:34:03.600 even then, for that amount of money, what do you get? 00:34:03.660 --> 00:34:07.020 I think you get maybe 22 per 2nd or what? 00:34:07.080 --> 00:34:12.060 No, you should definitely. I mean, I get 50 toes per 2nd on this thing. 00:34:12.120 --> 00:34:13.980 Of course, I'm not, I'm running Qwen. 00:34:17.040 --> 00:34:20.580 Yeah, well, apparently 50 tokens per 2nd for one. 00:34:20.760 --> 00:34:26.340 LLM call, right? For one inference provider, whereas if you are running 00:34:26.340 --> 00:34:29.580 Cloud, you're running 90 agents in parallel. 00:34:29.640 --> 00:34:35.460 I mean, like, I guess when it comes to freedom and independence 00:34:35.460 --> 00:34:37.260 and all that stuff that I'm aiming for. 00:34:38.160 --> 00:34:44.340 At the same time, I'm, I guess I'm a hypocrite because like, of course, I'm 00:34:44.340 --> 00:34:47.580 not going to buy 14 K machine and host at my home. 00:34:47.640 --> 00:34:53.460 Of course, I wouldn't mind, but like, The reality is, is that, you know, when I'm 00:34:53.460 --> 00:34:55.860 not using that machine, other people need to use it. 00:34:55.920 --> 00:35:00.120 The only, the only way, uh, this technology is gonna scale so that everyone can 00:35:00.120 --> 00:35:05.700 use it. Like this, I was, I was thinking of a guy I know 00:35:05.700 --> 00:35:09.960 in Singapore. I hope he doesn't mind me saying his name, Chian. 00:35:10.020 --> 00:35:13.980 He did, he's done, he's done hacker web. 00:35:14.040 --> 00:35:18.240 He done rail router, trainmap.sg. 00:35:18.300 --> 00:35:22.800 He's done all these, he's done like the trees.SG, where you can see all the trees 00:35:22.800 --> 00:35:28.320 in Singapore. He's like one of the most talented, Developers I've ever 00:35:28.320 --> 00:35:30.540 come across in my, in my career. 00:35:30.600 --> 00:35:37.200 He, his UX and his, his, his product, Ownership 00:35:37.200 --> 00:35:42.900 levels are next level. So I don't know the reason why exactly, but he's been unemployed 00:35:42.900 --> 00:35:47.580 for a while. He's kind of like been living off, like, I don't know how he's been 00:35:47.580 --> 00:35:52.740 living, to be honest, but he's office savings and GitHub, sponsorships and things 00:35:52.740 --> 00:35:57.600 like this. And, and like, I was just thinking, I mean, I'm probably, I shouldn't 00:35:57.600 --> 00:36:01.560 mention him by name, but like, I was just thinking to myself, like, he's probably 00:36:01.560 --> 00:36:07.500 not using AI as much as me because he doesn't have the resources, resources 00:36:07.500 --> 00:36:13.380 that I have using, you know, I basically have a, a Claude 00:36:13.380 --> 00:36:18.000 Max, and I have an OpenAI pro or whatever it's called. 00:36:18.060 --> 00:36:21.480 If he has open source project, you who have a sponsorship, you don't know that. 00:36:22.020 --> 00:36:24.480 But why what makes you say that? 00:36:24.540 --> 00:36:29.280 Like, do you think there's, like, he would, you would see more, more, um, code 00:36:29.280 --> 00:36:35.100 gen? Yeah, like this. I mean, I'm probably, I'm probably making the wrong... 00:36:35.100 --> 00:36:38.220 A lot of assumptions, yeah. I'm definitely making the wrong conclusions. 00:36:38.280 --> 00:36:43.020 But like, I do... But I think the thought is still valid, that like there's some 00:36:43.020 --> 00:36:46.680 people who are not using AI as much as me, because they just don't have the subscription. 00:36:46.980 --> 00:36:50.160 Mm-hmm. And that's, that sucks. 00:36:51.120 --> 00:36:55.980 It's why open quote says that they're most popular in like the other count, the 00:36:55.980 --> 00:37:01.560 um, East Asian countries because of uh, cheap, cheap models like Deep Sea Penguin, 00:37:01.560 --> 00:37:02.820 right? I think we mentioned that before. 00:37:02.880 --> 00:37:04.980 Oh, is it easy to set up at the box? 00:37:05.040 --> 00:37:06.840 So what would you have to set up an open router? 00:37:06.900 --> 00:37:10.560 OpenCode is their most... 00:37:12.600 --> 00:37:17.940 I guess because they are independent of OpenAI and an entropic. 00:37:18.000 --> 00:37:23.700 They are also a gateway for engineers in Southeast Asia to 00:37:23.700 --> 00:37:29.520 use Quinn and DeepSeek, and they had another model access to the one. 00:37:29.580 --> 00:37:32.220 So doesn't code do the LLM provision too? 00:37:32.280 --> 00:37:35.160 How does it all work? I mean, some money... 00:37:35.220 --> 00:37:37.320 They have, that's demonetisation, I guess, right? 00:37:37.380 --> 00:37:41.700 Their their their agent, their their harness is open source and free. 00:37:41.760 --> 00:37:45.120 They have their, um, they have their service. 00:37:45.180 --> 00:37:49.320 Yeah, when you when you when you start OpenCode, you can just go and put some 00:37:49.320 --> 00:37:54.840 like money on your wallet and then use the subscriptions via their routers. 00:37:55.200 --> 00:38:00.600 You were impressed from light. was it light LLM that got acquired for, like, 1000000000s 00:38:00.600 --> 00:38:04.260 of dollars and how easy... Open Ridge is the one I enjoyed. 00:38:04.320 --> 00:38:06.240 I mean, I was impressed by. 00:38:07.680 --> 00:38:13.680 Yeah, I get that. So OpenCode, what, one thing that, I've 00:38:13.680 --> 00:38:18.480 been pushed in, uh, maybe it's because I'm an ex-user, is uh, that that Groc build 00:38:18.480 --> 00:38:21.000 product, and it's like really expensive. 00:38:21.060 --> 00:38:23.820 It's like, I think it's like, it starts at $200 a month. 00:38:23.820 --> 00:38:29.700 And I'm wondering to myself, Yeah, a lot of people said croc is extremely expensive. 00:38:29.820 --> 00:38:35.520 But you know, when I'm on Instagram or Facebook or anywhere, I 00:38:35.520 --> 00:38:39.120 get bombarded with Groc advertisement. 00:38:39.180 --> 00:38:44.640 They have these social influences, usually pretty girls talking about how they 00:38:44.640 --> 00:38:48.960 build this or that. And that's not my feet, right? 00:38:49.020 --> 00:38:51.240 That's my son used my Instagram. 00:38:51.300 --> 00:38:56.940 Yeah, yeah, of course. So, so the, you know, talking about how they build this 00:38:56.940 --> 00:38:59.880 or that product, um, using Robbot. 00:38:59.940 --> 00:39:05.520 Often it's very simplistic. Like, oh, I just had this problem and I have 00:39:05.520 --> 00:39:09.180 growth pot, you know, do the research and build it for me and blah, blah, blah. 00:39:09.240 --> 00:39:14.160 And I get it, come bombard it, where I'm like, 0 my god, you know the reason? 00:39:14.220 --> 00:39:19.860 No, like, I hate it because I, I, I, I run agents myself and, 00:39:19.860 --> 00:39:25.920 and I guess I've never seen advertisement for Kiro Crew or, um, Hermes, 00:39:25.920 --> 00:39:27.120 well, yeah, they're open source. 00:39:27.240 --> 00:39:30.720 you know, like for people to get like the gateway. 00:39:30.780 --> 00:39:35.760 I guess it's a smart move by by X or by, you know, by Musk. 00:39:36.180 --> 00:39:37.500 So basically, hey, you know what? 00:39:37.560 --> 00:39:40.140 Most people don't know how to run a bot or an agent. 00:39:40.200 --> 00:39:46.980 Right? I actually, I give you an ex- when I build the Kero Crew OpenAI, um, ACP 00:39:46.980 --> 00:39:52.680 bridge, I send you a propo, um, like a project plan of like, hey, if 00:39:52.680 --> 00:39:56.880 I provide this as a service where people just click a button and very quickly, 00:39:56.880 --> 00:39:59.700 you know, you sign in with your ChatGPT subscription, which a lot of people have, 00:39:59.700 --> 00:40:03.420 then you get like, here you go. There's a little node running. 00:40:03.480 --> 00:40:08.100 And if you want to run it on your own infra, you click a button and they cut sports... 00:40:08.160 --> 00:40:12.360 This this definitely provide us last time I looked when open claw was a thing. 00:40:13.200 --> 00:40:16.500 Yeah, open. Yeah, there's plenty for open claw to do that. 00:40:16.560 --> 00:40:20.220 I just liked Kiro Crew because of the UI and everything. 00:40:20.340 --> 00:40:24.720 really nice. And it's also... So, you know, that's probably something that you 00:40:24.720 --> 00:40:30.240 and I or you or I or both, we should, we should probably be building it 00:40:30.240 --> 00:40:30.780 because like, 00:40:31.980 --> 00:40:34.620 But no, Musk did it, right? And then he's blasting it everywhere. 00:40:34.680 --> 00:40:38.400 Instagram, Facebook, X. And then, you know, I think it's amazing. 00:40:38.460 --> 00:40:43.020 I think it, I think this is the, this is, I said this to, to, to my team because 00:40:43.020 --> 00:40:47.460 we're building like agent workflows, workflows where you can invoke agent tasks, 00:40:47.460 --> 00:40:49.380 and they're triggered by by events. 00:40:49.440 --> 00:40:55.560 Um, but I think the real, um, future of agentic, the 00:40:55.560 --> 00:41:00.840 agentic future is every employee has a couple agents and the agents are acting 00:41:00.840 --> 00:41:04.440 on their own identity, but on behalf of the user. 00:41:04.500 --> 00:41:10.020 So they have their own identity and they are identified as bots, but they 00:41:10.020 --> 00:41:13.380 belong to this user. And I see this with Kiro Crew, and I think that's what EDBS 00:41:13.380 --> 00:41:17.040 says. That's how they, like, they saw Kiro Crew adoption. 00:41:17.100 --> 00:41:20.520 Like after somebody built it internally, as they called it, kiru claw. 00:41:20.580 --> 00:41:23.340 It was an open clock. I mean, there's so much momentum. 00:41:23.400 --> 00:41:27.540 It's like, I don't even feel the need to build it because it's coming. 00:41:27.600 --> 00:41:31.620 I mean, I kind of have it already and it's just the way the whole industry is moving 00:41:31.620 --> 00:41:35.760 in a way. Yeah, but I think that's how enterprise adoption is going to go as well. 00:41:35.820 --> 00:41:39.720 Like every employee within an organisation will have a couple of agents that run 00:41:39.720 --> 00:41:43.680 tirelessly, even when the user is not on their laptop. 00:41:43.740 --> 00:41:45.300 It runs on its own infrastructure. 00:41:45.360 --> 00:41:48.540 It has like its own node that it has full control over. 00:41:48.600 --> 00:41:54.420 It has an identity that it acts on as, you know, a contributor and get, a 00:41:54.420 --> 00:41:58.320 participator in Slack. There's some projects like beehive or something. 00:41:58.380 --> 00:42:03.300 There's some slack alternatives where, you know, humans and boats are both... 00:42:03.360 --> 00:42:08.580 I'm not actually, I was, I think I would have agreed with your, the identity thing. 00:42:08.640 --> 00:42:13.740 Now I'm thinking the identity must be attached to a person so that someone owning 00:42:13.740 --> 00:42:15.300 it and being successful, but yeah. 00:42:15.360 --> 00:42:19.560 That's what I said. I said there will be a couple of agents that have their own 00:42:19.560 --> 00:42:23.340 identity, but there are, you know, there's a human responsible for them. 00:42:23.400 --> 00:42:25.380 Exactly, yeah, yeah, yeah. Okay, yeah. 00:42:25.380 --> 00:42:28.740 So when you look at Kiro Crew, if you go to the repository, if you create a poor 00:42:28.740 --> 00:42:33.480 request, there will be a couple of both responses, and often these PR comments 00:42:33.480 --> 00:42:38.280 will show, I am a bot, and my responsible human is this person. 00:42:38.340 --> 00:42:43.920 So each one of them, like they have a few, and they're 00:42:43.920 --> 00:42:48.600 acting on behalf of the humans, they have their loops, they have all kinds of, 00:42:48.600 --> 00:42:53.760 you know, I mean, I'm willing to accept this future. 00:42:54.060 --> 00:43:00.000 Though just to rewind it a little bit, I mean, I 00:43:00.000 --> 00:43:01.980 feel like this is going to be out of reach for a lot of people. 00:43:02.640 --> 00:43:08.100 Because it's an expensive amount of infrastructure, really. 00:43:08.280 --> 00:43:13.800 And the fact that, uh, I mean, I'm, we're whining that Grock Builders 00:43:13.800 --> 00:43:15.780 join a, it starts a join a bucks a month. 00:43:15.840 --> 00:43:19.020 It's probably just going to get more and more expensive, the time goes on really. 00:43:22.080 --> 00:43:25.140 So, it's a bit, it's a bit weird, isn't it? 00:43:25.200 --> 00:43:30.480 Like, like, to get started with computing when I was a kid, was expensive, I think, 00:43:30.480 --> 00:43:36.300 when I was a kid, my dad, like, spent a truckload of money getting me my 1st PC, 00:43:36.300 --> 00:43:41.220 like more than £1000 and that must be like £10,000 and in today's money. 00:43:41.880 --> 00:43:45.240 Yeah, it's... Yeah, it's insane. 00:43:45.300 --> 00:43:50.280 And but like, but like, there's so many factors now, like, 00:43:51.900 --> 00:43:56.280 Like, one reason why I had a computer back in South Africa is my dad used to go 00:43:56.280 --> 00:44:01.740 buy fabric in Singapore, he used to go to Sim Lim, buy me hardware, because South 00:44:01.740 --> 00:44:04.740 Africa had sanctions and bring them back to South Africa. 00:44:04.800 --> 00:44:08.100 So what I'm trying to say here is like the future is looking pretty wild because 00:44:08.100 --> 00:44:13.320 like, you know, China is supposed to not have the, you know, the anthropic and 00:44:13.320 --> 00:44:14.520 OpenAI models. 00:44:16.200 --> 00:44:17.940 You're supposed to have a credit card, I suppose. 00:44:18.000 --> 00:44:23.040 I mean, how does a young kid have a credit card and is able to pay 200 bucks a 00:44:23.040 --> 00:44:24.600 month or something? That's impossible. 00:44:25.440 --> 00:44:31.200 There's all these like exclusionary, uh, or there's all these barriers 00:44:31.200 --> 00:44:34.980 to entry when it comes to becoming an AI engineer, right? 00:44:36.180 --> 00:44:40.860 Anyway, that's just mine. But, but, so I hadn't talk. 00:44:40.920 --> 00:44:45.420 Well, my kids have access to my accounts, my subscriptions. 00:44:45.480 --> 00:44:51.060 Oh, your kids have like a, what I, I had a discussion with someone 00:44:51.060 --> 00:44:56.640 over the weekend talking about, because my son studied uh, algorithms and and 00:44:56.640 --> 00:45:03.000 data structures and, and, you know, studied a bunch of, my, I, I got in brilliant.org, 00:45:03.000 --> 00:45:06.360 so he understands like how an index in the database works. 00:45:06.480 --> 00:45:10.440 Like, Brilliant Org has a really nice course explaining parabolism. 00:45:10.500 --> 00:45:16.620 what, sorry? Brilliant.org has a nice, has a nice set of courses around, 00:45:16.620 --> 00:45:20.460 like engineering fundamentals. talking about the difference between true put and 00:45:20.460 --> 00:45:26.040 parallelism, the difference between like how would a librarian organise 00:45:26.040 --> 00:45:28.980 books and like how do you look up the books? 00:45:29.040 --> 00:45:32.880 Like if you sort them by this by author, but then you want to get them by category, 00:45:32.880 --> 00:45:37.140 you must have some type of look up mechanism to then go to the right shelf. 00:45:37.200 --> 00:45:40.020 So basically explaining how indexes work in databases and stuff. 00:45:40.080 --> 00:45:41.940 It's really nice. really experiences me well. 00:45:42.060 --> 00:45:46.380 I never quite understood the Dewey system, my local library when I was a kid. 00:45:46.440 --> 00:45:52.020 Yeah. So I kind of pushed my son hard into engineering. 00:45:52.080 --> 00:45:57.600 into software engineering. My wife is now pushing him hard to do something more 00:45:57.600 --> 00:46:01.620 like going into NBA or logistics or something else. 00:46:01.680 --> 00:46:07.560 Are you for real? Uh, yeah, so I was talking with someone who isn't a 00:46:07.560 --> 00:46:10.380 software, like, engineering team director. 00:46:10.440 --> 00:46:13.320 I think CTO were the company here in Vietnam. 00:46:13.380 --> 00:46:17.400 And we were talking about where do you see the future, like of engineering. 00:46:17.460 --> 00:46:22.440 And I think I read quite a few articles about companies that laid off engineers, 00:46:22.440 --> 00:46:28.200 believing that AI can build the code, have largely gone, like, ended up with a 00:46:28.200 --> 00:46:34.740 bunch of non-architected solutions that are hard to maintain and upgrade and operate, 00:46:34.740 --> 00:46:40.320 and there will be resurgence in hiring engineers to do, like, a proper architectural 00:46:40.320 --> 00:46:44.460 review and clean up all of the things that has been delivered. 00:46:44.580 --> 00:46:49.800 what I've been doing. Yeah. And that's what a friend of mine also said, like, very 00:46:49.800 --> 00:46:52.740 hands-on. Um, they acquired the company. 00:46:52.800 --> 00:46:57.900 They're like a young company and they acquired another startup and he said that 00:46:57.900 --> 00:47:03.420 was basically AI AI slop and the performance issues and 00:47:03.420 --> 00:47:05.280 bugs and things like that were insane. 00:47:05.340 --> 00:47:11.100 So he spent weeks and weeks of, you know, refactoring and re-engineering 00:47:11.100 --> 00:47:13.980 the whole solution. You can do that a lot faster now with AI as well, right? 00:47:14.040 --> 00:47:20.820 You can you can set up basically black box testing to achieve, 00:47:20.820 --> 00:47:23.340 you know, to keep the characteristics that you like. 00:47:23.400 --> 00:47:27.480 Yeah, right. or whatever. Yeah, I like that. 00:47:27.540 --> 00:47:33.180 So I think that's a future. I think, um, I think that that puts, that puts 00:47:33.180 --> 00:47:36.660 engineers back on the map and I was telling, I was thinking I need to, you know, 00:47:36.660 --> 00:47:39.720 see if my son really likes software engineering and if he really likes it, he should 00:47:39.720 --> 00:47:42.960 go further. You can wear a line, like... 00:47:42.960 --> 00:47:45.120 Being an AI engineer, we need more of them. 00:47:45.180 --> 00:47:47.820 There's a future for AI engineers. 00:47:47.880 --> 00:47:49.920 I mean, there's a fantastic future for AI engineers. 00:47:50.160 --> 00:47:53.280 I think it's, it's, this is the hard bit. 00:47:53.340 --> 00:47:56.400 It requires a high level architectural thinking. 00:47:56.460 --> 00:47:59.640 I don't think, like, AI engineer is just a hype term. 00:47:59.700 --> 00:48:04.260 I think it's an architectural, like an engineer that knows and understands architecture, 00:48:04.260 --> 00:48:10.740 um, of software engineering, designing for what change, and the ability to, um, 00:48:10.740 --> 00:48:15.480 consider operational costs. I think that was always the case with like, you know, 00:48:15.480 --> 00:48:19.920 SRE and DevOps. They always say that those are people that's supposed to have years 00:48:19.920 --> 00:48:23.700 and years of experience of operational hand on, that are not just only looking 00:48:23.700 --> 00:48:28.620 at delivering a feature to production, but actually knows on the operational aspect. 00:48:28.680 --> 00:48:30.660 Like, you don't just throw the code over the wall. 00:48:30.720 --> 00:48:32.580 You're also responsible for how it operates. 00:48:32.640 --> 00:48:38.580 I agree. I mean, the... I think it was said by Hashimoto 00:48:38.580 --> 00:48:42.180 a week or 2 ago, and it's something that gels well with me. 00:48:42.240 --> 00:48:48.000 is like, if someone has autonomy, has has the autonomy to build the code, 00:48:48.000 --> 00:48:54.300 run the code, evolve the code, do all those things with AI, 00:48:54.300 --> 00:48:59.460 and it's, it's, uh, it's actually quite rare. 00:48:59.520 --> 00:49:05.220 I find. Super red to someone who has the determination to just drive 00:49:05.220 --> 00:49:08.760 through problems. Yeah, I think it's a combination. 00:49:08.820 --> 00:49:13.620 It needs to be, you have the idea of building a product and delivering that product. 00:49:13.680 --> 00:49:19.380 So for my son who has some coding experience, some AI 00:49:19.380 --> 00:49:23.580 agent experience. I think for him most importantly is to own a product. 00:49:23.640 --> 00:49:28.920 and to be able to identify what are the features that I need to build and basically 00:49:28.920 --> 00:49:33.780 learn how to deliver product features in an, in an, in an, you know. 00:49:33.840 --> 00:49:36.240 my next question is, surely, hopefully he has a product. 00:49:36.300 --> 00:49:39.120 Has he built something? I give him assignments. 00:49:39.240 --> 00:49:44.220 I, we have, I, we have now an Lenovo laptop, which is no longer his because he 00:49:44.220 --> 00:49:47.940 he got a new one, and I told him, we have 2 controllers. 00:49:48.000 --> 00:49:51.600 We have a sofa and I couldn't find proper co-op sofa games. 00:49:51.660 --> 00:49:54.780 I have 2 controllers and I was like, we can play it. 00:49:54.840 --> 00:50:00.660 It takes two. We can play cult of the, I always, 00:50:00.660 --> 00:50:03.960 I always say it wrong on purpose to, to, to piss up my kids. 00:50:04.020 --> 00:50:08.400 It's like cult of the, It's not cultural that goat, which is a really old hacker 00:50:08.400 --> 00:50:12.120 group. It's it's cult of the that something else. 00:50:12.180 --> 00:50:15.600 Um, it's called of the goat, or cult of the cow. 00:50:15.660 --> 00:50:18.240 Anyway, there's this game that's that's like multiplayer. 00:50:18.300 --> 00:50:23.700 There's, um... There's a few. I don't know entangle, but there's carpet. 00:50:23.760 --> 00:50:30.000 Anyway, not enough. Not enough. So I told him, I'm going to install Omar on 00:50:30.000 --> 00:50:34.080 this Linux on this desktop. Sorry, on this laptop. 00:50:34.140 --> 00:50:39.660 And you're gonna deliver by Christmas to co-op sofa 00:50:39.660 --> 00:50:45.000 games because he likes playing games and he knows a real engine and how to build 00:50:45.000 --> 00:50:50.880 games. I must say, OpenAI’s, Astra is amazing for generating 3D 00:50:50.880 --> 00:50:54.900 stuff. amazing. And my feed on Reddit for the last week since Astra came out is 00:50:54.900 --> 00:50:57.180 nonstop. Astra built this in a day. 00:50:57.240 --> 00:51:00.480 Astra Building 14 hours. I've been modelling my new home with it. 00:51:00.540 --> 00:51:02.100 It's incredible. It's incredible. 00:51:02.160 --> 00:51:08.220 Yeah, so that's why. So anyway, how is he doing with the 2 00:51:08.220 --> 00:51:09.360 co-op games? 00:51:10.200 --> 00:51:15.720 He hasn't built any yet. So that's why you're asking me, does he have a 00:51:15.720 --> 00:51:17.940 product to own? Does he have something to deliver? 00:51:18.000 --> 00:51:19.560 Okay, so that's that's the problem there. 00:51:19.620 --> 00:51:24.120 Maybe he has December. You also need to scratch your own itch, right? 00:51:24.180 --> 00:51:26.460 That's that's an important aspect of autonomy. 00:51:26.520 --> 00:51:30.720 Yeah, he has this idea for a game, but then it never materialises. 00:51:30.780 --> 00:51:35.040 I send him, there's an AI, game death, subreddit, and they talk about how to generate 00:51:35.040 --> 00:51:39.480 assets, which models are good with which work... 00:51:39.540 --> 00:51:43.200 One of the 1st major things I did in my career was make a game, but I'm not sure 00:51:43.200 --> 00:51:45.960 I can recommend it because I'm just traumatised by making that game. 00:51:47.580 --> 00:51:53.580 So maybe something easier, like, uh, Oh, 00:51:53.580 --> 00:51:56.340 I offered them so many options. Like we used to play Valorant together. 00:51:56.400 --> 00:51:58.500 Like, what if you were in a lobby? 00:51:58.560 --> 00:52:02.220 Ask him to come with the idea, Vincent, come on. 00:52:02.940 --> 00:52:05.220 He doesn't come up with anything. 00:52:05.280 --> 00:52:06.840 So I said to him, it's very simple. 00:52:06.900 --> 00:52:12.300 You know, scrape this website. I, I, I, I gave him this idea. 00:52:12.360 --> 00:52:16.440 Like if you're in a lobby waiting for a map, the map loads and you see your lineup. 00:52:16.560 --> 00:52:22.320 Um, so you have, you know, you have valorances 00:52:22.320 --> 00:52:25.080 with agents, a bit like um, overwatch. 00:52:25.140 --> 00:52:28.440 So every agent has to different capab- like abilities. 00:52:28.500 --> 00:52:32.700 Oh, yeah. So one agent. Each, like they're different roles. 00:52:32.760 --> 00:52:37.860 So usually you have attackers, I forgot the names of the roles. 00:52:37.920 --> 00:52:41.940 Anyway, so the point is that when you were in the team, you have certain abilities 00:52:41.940 --> 00:52:44.580 and certain agents that combine in certain ways on certain maps. 00:52:44.640 --> 00:52:50.160 So what I wanted is the moment you're in a scrim and you get kind of random, you 00:52:50.160 --> 00:52:53.880 should be able to very quickly come up with like, oh, this is the map, this is 00:52:53.880 --> 00:52:57.660 the agents. these are the lineups that you can do with those 2 agents on this part 00:52:57.660 --> 00:53:00.060 of the map, they can do this lineup together to take over the site. 00:53:00.120 --> 00:53:02.400 I told do something simple like that. 00:53:02.460 --> 00:53:05.700 You know, with your phone, take a picture, and boom, immediately you get a couple 00:53:05.700 --> 00:53:09.600 of lineups. On that note, which is pretty bizarre. 00:53:09.660 --> 00:53:15.240 But like, in the real world, what agent, like if you were delivering a 00:53:15.240 --> 00:53:18.600 generic software product, what agents do you think you need? 00:53:18.660 --> 00:53:24.180 Like, you'd probably want a security reviewer agent, you probably want a, a developer 00:53:24.180 --> 00:53:26.820 agent, you probably want a reviewer agent. 00:53:26.880 --> 00:53:29.580 Are we back to these these roles? 00:53:29.940 --> 00:53:35.580 I actually never use them. And the funny thing is, in Claude Quote today, still, 00:53:35.580 --> 00:53:41.100 whenever I ask, um, you know, that security review, uh, actions pretty 00:53:41.100 --> 00:53:46.080 handy. I use it all the time. But then I'm playing, I'm playing the security engineer 00:53:46.080 --> 00:53:52.500 hat, in my, in my employer. I always just tell Opus, approach 00:53:52.500 --> 00:53:57.540 this from this angle and then load up your, like, launch an agent with this, with 00:53:57.540 --> 00:54:01.140 the context tuned for this type of activity and then it just builds it. 00:54:01.200 --> 00:54:03.180 I don't have dedicated agent profile. 00:54:03.240 --> 00:54:08.400 agents. uh, oh, I don't know what, this is.agent format, isn't there? 00:54:08.460 --> 00:54:10.980 I haven't used it, but I know it's there. 00:54:11.040 --> 00:54:14.520 And I know that that by defaulted chips and explorer and except ships a couple 00:54:14.520 --> 00:54:19.920 of others. So you don't see the, you don't see the point of being having a sort 00:54:19.920 --> 00:54:22.920 of a role, role-playing or something like this? 00:54:22.980 --> 00:54:27.960 I mean, I, I, I, I, I'm on the fence about it myself, really. 00:54:28.200 --> 00:54:33.720 I can understand that loading a context with a certain set of rules, and 00:54:33.720 --> 00:54:40.380 I do a three-way review, um, The 1st product that I built was 00:54:40.380 --> 00:54:42.960 an gentic review with 3 different angles. 00:54:43.020 --> 00:54:49.440 So it created 3 agents reviewing it from 3 different perspective, from an ops operational. 00:54:49.440 --> 00:54:54.720 Adversarial and insecurity wise, operational wise, and I forgot the 3rd one, architectural 00:54:54.720 --> 00:54:58.560 wise, maybe on the code. And those were the 3 reviews that came back. 00:54:58.680 --> 00:55:02.520 And then the final agent would aggregate the report and post it as a comment on 00:55:02.520 --> 00:55:07.020 the PR. That was the 1st thing I built for the as part of my probation. 00:55:07.800 --> 00:55:10.800 But it wasn't with like Asian profiles. 00:55:10.860 --> 00:55:12.540 It was just like loading up the context. 00:55:12.600 --> 00:55:16.740 I think the biggest mistake you can make is to tell, to start an agent profile 00:55:16.740 --> 00:55:20.520 with like, you are an expert, security reviewer. 00:55:20.880 --> 00:55:26.640 Why do you think that? Because the model we work out of memory, it 00:55:26.640 --> 00:55:28.620 will assume that it shouldn't ask any questions. 00:55:28.680 --> 00:55:33.180 It will not verify. It will, it will not go and and say like, Well, get confused 00:55:33.180 --> 00:55:38.520 what role it is. Yeah, because if you, if you're an expert, you are, you already 00:55:38.520 --> 00:55:40.320 know everything, you don't have to go and look. 00:55:40.380 --> 00:55:44.400 So the worst thing you can do is tell an LLM that is an expert. 00:55:44.460 --> 00:55:46.980 And I told you this, you're an expert. 00:55:47.040 --> 00:55:49.500 Vincent. I told you this months ago. 00:55:49.560 --> 00:55:55.800 Yes, I did. I'm pretty sure if you go and search on your on your transcript summary 00:55:55.800 --> 00:55:58.620 or whatever, even if the full transcript, you will find me saying that this is 00:55:58.620 --> 00:56:01.380 the worst thing you can do is creating roles. 00:56:01.440 --> 00:56:05.400 Well, thanks for bringing it up again, because I think your argument is really 00:56:05.400 --> 00:56:07.140 good, actually. It's not my argument. 00:56:07.200 --> 00:56:09.720 I read it somewhere. I read it. 00:56:11.040 --> 00:56:13.380 But what I really wanted to talk about. 00:56:13.380 --> 00:56:15.660 I think about arguments that I had to pull you up on. 00:56:15.720 --> 00:56:19.440 It's like, we were talking about rewrites and you said, no, that's one way to kill 00:56:19.440 --> 00:56:24.120 a product. Why? The, uh... when was that? 00:56:24.180 --> 00:56:26.940 I think I was just trolling you, I was probably in a bad mood. 00:56:27.420 --> 00:56:32.460 Okay, I think we talked enough. I need to answer some messages on Slack, to be 00:56:32.460 --> 00:56:34.740 honest. To be honest, you never talk. 00:56:34.800 --> 00:56:36.840 I told you I have one topic I want to address. 00:56:38.160 --> 00:56:40.080 Am I going to help you? 00:56:41.280 --> 00:56:47.040 I shared you 2 articles. I mean, we talked about one article last time, and it 00:56:47.040 --> 00:56:53.160 was about the idea of agents running a lot of verifications and self verifications. 00:56:53.220 --> 00:56:58.560 Therefore, why does the pool request, um, CI, containers integration, need to rerun 00:56:58.560 --> 00:57:00.420 every command that the agents already ran. 00:57:01.140 --> 00:57:02.760 Remember that? 00:57:03.420 --> 00:57:06.600 Sorry, repeat, I got distracted while looking for your message. 00:57:07.080 --> 00:57:10.920 Ah, there's a ton of WhatsApp, probably people. 00:57:10.920 --> 00:57:14.220 Let me try and share it. Do you know what? 00:57:14.340 --> 00:57:16.680 I have I have a notifications disabled everywhere. 00:57:16.740 --> 00:57:20.760 So every time I'm looking at my WhatsApp was the name like, oh, okay, there's a 00:57:20.760 --> 00:57:25.140 bunch of things. Oh, not this. So there was this article from... 00:57:25.260 --> 00:57:26.400 Oh, the one that you shared this morning. 00:57:27.840 --> 00:57:33.060 Uh, the one that I shared this morning was just a follow-up that I say, hey, this 00:57:33.060 --> 00:57:35.460 kind of like belonging the same bracket, right? 00:57:35.520 --> 00:57:41.040 So here we have the original article that we discussed last time, which is 00:57:41.040 --> 00:57:46.860 CI is now is the bottleneck, and it was posted on the 26th of August by 00:57:46.860 --> 00:57:50.700 one of the swamp creators, Paul's Tech. 00:57:50.760 --> 00:57:56.460 And he talks about the concept of having a verification, basically an 00:57:56.460 --> 00:58:00.660 attestation. I remember, we struggled with the word, say the word, I hate saying 00:58:00.660 --> 00:58:00.960 that word. 00:58:01.740 --> 00:58:07.320 Basically, a very, like, a signed a certificate, like, an 00:58:07.320 --> 00:58:10.020 agent. It can be fully deterministic, right? 00:58:10.080 --> 00:58:15.060 The agent can only get, um, this this manifest or this certificate that it has 00:58:15.060 --> 00:58:18.120 ran the test. if you had really ran the test, right? 00:58:18.180 --> 00:58:21.300 So there's some type of verification that runs. 00:58:21.300 --> 00:58:26.640 And then you push that up. And basically the CI says, great. 00:58:26.700 --> 00:58:31.200 The signature valid is valid, I don't need to run a bunch of tests. 00:58:31.260 --> 00:58:33.360 This is confirmed. The agent runs it. 00:58:33.420 --> 00:58:39.360 So when I saw this article today on 00:58:39.360 --> 00:58:43.920 September 14th, not today, today, yesterday, well, today. 00:58:43.980 --> 00:58:49.740 which is agenda coding is training the CI, which is pretty much the same title 00:58:49.740 --> 00:58:53.820 here, right? I broke the CI. I was like, great, they're they're going to like... 00:58:53.820 --> 00:58:55.140 was ahead of the curb by a whole month. 00:58:55.320 --> 00:59:00.360 Yeah, so I thought they're going to like, you know, explore this and go further. 00:59:00.420 --> 00:59:04.380 And they talked about how, and I like this diagram, right? 00:59:04.440 --> 00:59:08.160 It's kind of cool. Um, you know, building was a bottleneck, then the review became 00:59:08.160 --> 00:59:10.800 the bottleneck, but now you have agantic reviews. 00:59:10.860 --> 00:59:15.960 So now CI is the bottleneck. And they talk specifically on how many engineers, 00:59:15.960 --> 00:59:19.080 like they, the code base grew 10 times. 00:59:19.140 --> 00:59:22.980 The CI jobs increased by 25 times. 00:59:23.040 --> 00:59:28.740 That's a big number. And so basically how they took a bigger 00:59:28.740 --> 00:59:31.440 machine, which gave them a bit of a runway of 10 weeks. 00:59:31.500 --> 00:59:36.000 But then it broke. So then they started charting the jobs so that they are spread 00:59:36.000 --> 00:59:41.280 across multiple machines. But that only gave them like 29 days instead of 10 weeks. 00:59:41.280 --> 00:59:46.560 And it's... I like how the progression goes, like, step one increased CPU, step 00:59:46.560 --> 00:59:52.080 two, shard. So I expected this article to end with like, we don't need CI 00:59:52.080 --> 00:59:54.300 anymore, right? We have verification by the agent. 00:59:54.360 --> 00:59:57.300 Yeah, we have exactly what Paul Stec told them to do. 00:59:57.360 --> 00:59:59.280 Unfortunately, that's not the case. 00:59:59.340 --> 01:00:04.920 Um, they just, I think the end of it is, um, 01:00:04.920 --> 01:00:05.580 where is it? 01:00:06.240 --> 01:00:12.060 This is the redesign. The redesign is to basically build 01:00:12.060 --> 01:00:16.440 a database and memory data store and blah, blah. 01:00:16.500 --> 01:00:20.100 Honestly, I'm not impressed by these articles by entropic, to be honest. 01:00:20.160 --> 01:00:25.740 The technical details and the like last time when 01:00:25.740 --> 01:00:29.340 they they shared about the AI is DLC, I was very disappointed in that article. 01:00:29.400 --> 01:00:32.220 I think there's a lot more interesting articles out there. 01:00:32.460 --> 01:00:37.020 We talked about the... 01:00:38.160 --> 01:00:43.680 On anthropic and his article. Like, uh, inside my, my employer, there's a 01:00:43.680 --> 01:00:47.580 PDF called uh, 2026 state of AI agents report. 01:00:48.180 --> 01:00:53.580 And inside that PDF, which you have to like, unfortunately, use a form to download, 01:00:53.580 --> 01:00:55.560 you know what I mean? Like, you have to give your email address away. 01:00:55.800 --> 01:01:01.620 The number one bottleneck there is integrations. 01:01:02.880 --> 01:01:08.400 But I find like, like Claude Anthropic is putting out a lot of, like, what 01:01:08.400 --> 01:01:08.940 do you call it? 01:01:10.020 --> 01:01:15.120 Slop. Yeah, slop for want of a better word. 01:01:15.180 --> 01:01:19.440 Like, it's putting out, like, a fair bit of messaging, like, where you're, what 01:01:19.440 --> 01:01:23.400 you need to optimise for what you need to think about, but I do find it strange, 01:01:23.400 --> 01:01:27.480 like, in one, in one report, it's like, integrations is the bottleneck. 01:01:27.600 --> 01:01:31.560 And then another report, no, CI is your ball neck. 01:01:31.620 --> 01:01:35.160 And it's like, okay, bottleneck, bottleneck, bottleneck. 01:01:35.220 --> 01:01:40.980 You're selling AI. Exactly. You need more agents to 01:01:40.980 --> 01:01:46.920 solve the problem. Add agents. So, so I'm disappointed 01:01:46.920 --> 01:01:50.580 by his article. I send it to you and I thought this would, you know, correlate 01:01:50.580 --> 01:01:54.180 with the ideas of CI having a different, you know, job. 01:01:54.240 --> 01:01:58.980 Um, but they're just talking about building a proper, like, CI being a product 01:01:58.980 --> 01:02:00.720 with um, I guess. 01:02:01.440 --> 01:02:07.500 I guess... The same with, like, organisations 01:02:07.500 --> 01:02:13.200 like, um, who was acquired, again, windsurf, no, um, by 01:02:13.200 --> 01:02:18.060 Devin. Devin. Cursor. Maybe cursor. 01:02:18.120 --> 01:02:24.480 There was this company that is a terraform consultancy company that 01:02:24.480 --> 01:02:29.640 talks about how they, Yeah, so they came into an organisation like Devon or Cursor 01:02:29.640 --> 01:02:32.340 to optimise their infrastructure as code. 01:02:32.340 --> 01:02:37.680 And they described like teralytic terraform configurations on terraform cloud and 01:02:37.680 --> 01:02:43.260 were costing 1000s of dollars for all the resources that were in them and basically 01:02:43.260 --> 01:02:48.960 ran hours and were like very brittle and and very painful and so 01:02:48.960 --> 01:02:53.940 on and so on. All the basic stuff that, you know, if you do infrastructure is code, 01:02:53.940 --> 01:02:57.960 if you know terraform, um, and then they they highlighted this. 01:02:58.020 --> 01:03:01.740 I think these articles go onto spacelift because they migrated cursor onto space. 01:03:01.800 --> 01:03:03.840 This is all this talk stuff, yeah. 01:03:03.900 --> 01:03:06.900 Yeah. I think when I started this podcast. 01:03:06.960 --> 01:03:10.200 We're talking about talk. Tacos. 01:03:10.260 --> 01:03:14.700 Tacos, tacos, automation, and collaboration software. 01:03:14.760 --> 01:03:20.280 But, um, yeah, I mean, maybe it's because we are in this, well, 01:03:20.280 --> 01:03:24.720 we're doing DevOps, we're doing CICD, we're building mono repos, we're building 01:03:24.720 --> 01:03:27.240 uh, task runners, we're building caches. 01:03:27.300 --> 01:03:29.640 We understand the infrastructure's code. 01:03:29.700 --> 01:03:33.000 We're building, we're optimising for delivering change quickly to production. 01:03:33.060 --> 01:03:38.400 And then reading organisations that are pure agentic, that are building an IDE 01:03:38.400 --> 01:03:43.740 or building an inference product that have a lot of researchers, building AI, um, 01:03:43.740 --> 01:03:48.480 and then, you know, even reading the OpenAI hack, of code artefact being such 01:03:48.480 --> 01:03:54.180 a, um, not code effect artifactory, having so many 0 days that were able 01:03:54.180 --> 01:03:59.520 to to being exploited, and then basically no proper uh, sandboxing on the Cubinitis 01:03:59.520 --> 01:04:04.920 cluster. I felt like these organisations that are at the frontier of AI have like 01:04:04.920 --> 01:04:09.060 hired the best people. They hire the best infro people, the best security people, 01:04:09.060 --> 01:04:12.360 the best dev-ops people. They would have a proper CICD system. 01:04:12.420 --> 01:04:15.000 They would have a proper sandboxing mechanisms. 01:04:15.060 --> 01:04:18.420 They would have air gapped agent training facilities. 01:04:18.480 --> 01:04:24.120 But, well, that comment is maybe wrong because, I mean, I agree that if 01:04:24.120 --> 01:04:26.880 an agent has to solve a task we should have access to download packages. 01:04:27.060 --> 01:04:32.700 So anyway, I just I'm just surprised to read that these these 01:04:32.700 --> 01:04:36.720 labs. Yeah, it's very much on our turf. 01:04:36.780 --> 01:04:42.300 I mean, I'm, I do get like uh, this feeling of FOMO that I should be 01:04:42.300 --> 01:04:45.660 working on uh, bigger, bigger challenges in a way. 01:04:45.720 --> 01:04:51.840 Because, I mean, I'm just, I'm working in an enterprise with, with quite familiar, 01:04:51.840 --> 01:04:57.780 Challenges, I would say. I mean, I I 01:04:57.780 --> 01:04:59.520 would like to work on the frontier, that's for sure. 01:04:59.580 --> 01:05:03.900 I think it would be... If you compare this, okay, let's compare this to an article 01:05:03.900 --> 01:05:08.820 from Tinder on how DNS broke their Cubinities cluster, or if you compare this from 01:05:08.820 --> 01:05:14.460 the tech team of at Reddis, sorry, Reddit on how they had to migrate their communities 01:05:14.460 --> 01:05:18.300 and configuration, these tech block articles are amazing. 01:05:18.360 --> 01:05:23.880 These infrastructure and technical engineering blog 01:05:23.880 --> 01:05:28.560 articles are really detailed. Okay, I need you need to maybe share them with me, 01:05:28.560 --> 01:05:32.160 but like, if I see a headline like how I broke my Kubinichi's cluster. 01:05:32.220 --> 01:05:34.020 I mean, I'm rolling my eyes because it's easy. 01:05:34.740 --> 01:05:40.320 That's easy. I've done that. You can you can find a lot of these here. 01:05:40.380 --> 01:05:46.320 I think, um, cute. I mean, to be honest, I think I'm so, 01:05:46.320 --> 01:05:51.360 I'm so, like, bored of Cubanese in a way, I just, I find it oppressive. 01:05:51.420 --> 01:05:54.720 Oh, it wasn't Tinder, it was okay, Cupid, I think. 01:05:54.780 --> 01:06:00.540 Okay, Cupid. Cupid? Dating websites have 01:06:00.540 --> 01:06:02.820 better tag teams than OpenAI. 01:06:05.880 --> 01:06:10.620 Wow, what a statement. Yeah, I didn't go there yet. 01:06:10.680 --> 01:06:16.380 But like if you see thumbnail systems on, on, on porn websites and, 01:06:16.380 --> 01:06:19.740 and the forefront, you know, they're ahead of YouTube and a lot of these things. 01:06:19.800 --> 01:06:24.480 There's some crazy technology. I know some sides use, which I... 01:06:24.540 --> 01:06:29.520 Did you watch the Spotify miniseries on Netflix? 01:06:30.060 --> 01:06:34.680 Uh, I think maybe a few minutes. 01:06:34.740 --> 01:06:37.440 like one of the Spotify guys. I actually know personally. 01:06:37.860 --> 01:06:44.580 So I had a bit of an issue watching it because the character was 01:06:44.580 --> 01:06:48.420 so, what do you call it, typecast or so weird. 01:06:48.480 --> 01:06:52.020 It's not like the real person. So I'm I'm not gonna. 01:06:52.200 --> 01:06:58.740 So there's a Wikipedia article about the network protocol that Spotify 01:06:58.740 --> 01:07:04.860 developed for delivering music as 01:07:04.860 --> 01:07:09.300 quick as it can. So that I think this episode three, I think it's 4 episodes. 01:07:09.360 --> 01:07:12.480 One is about the impact Spotify has for artists. 01:07:12.540 --> 01:07:16.560 First is about the founder on, like, where he come from and how... 01:07:16.560 --> 01:07:20.340 It goes through each person. Yeah, from like different tech guy. 01:07:20.400 --> 01:07:24.900 The tech guy. The tech guy episode is really interesting because I watched it and 01:07:24.900 --> 01:07:29.700 I was like, I didn't want to say it, but I actually know that I know the tech guy. 01:07:30.360 --> 01:07:33.060 Which, that guy, because they're like four. 01:07:33.180 --> 01:07:36.000 Uh, I know Andreas N. 01:07:37.560 --> 01:07:43.080 Um, but is it's, there's somebody right now that is being 01:07:43.080 --> 01:07:48.600 very, um, like, criticised because he does a lot of articles and 01:07:48.600 --> 01:07:53.460 posts about their agentic adoption and their delivery of product features with 01:07:53.460 --> 01:07:56.640 agents and so on. They also build it backstage. 01:07:56.700 --> 01:08:01.200 I think if you adopted backstage, now that gives you a big, big leg up, right? 01:08:01.260 --> 01:08:06.300 Um, because of um, basically the context engineering part for agents. 01:08:06.360 --> 01:08:07.560 I think that that really allows you. 01:08:08.820 --> 01:08:13.380 Really? Everybody hates backstage because it's, it's a mess. 01:08:13.440 --> 01:08:18.300 Everybody hates it, but I think today it deserves another look because as far as 01:08:18.300 --> 01:08:23.940 I understand, it has rebuilt itself around MCP and it has rebuilt itself around 01:08:23.940 --> 01:08:28.980 exactly what agents need, which is like what developers needed from backstage, 01:08:28.980 --> 01:08:32.460 a front page on where are the services who are the owners, where is my data dog 01:08:32.460 --> 01:08:34.920 metrics? How can you say this? reports. 01:08:34.980 --> 01:08:39.300 What is your latest? What is your latest, what do you call it? 01:08:39.360 --> 01:08:42.660 experience with the backstage? Does your employer run it? 01:08:42.780 --> 01:08:48.480 No, um, how can you, how can you say this, Vincent, you're shocking me to my core. 01:08:49.200 --> 01:08:52.440 Next thing you're going to start recommending, I'm out of Microsoft. 01:08:52.680 --> 01:08:58.740 Wow. That's insulting. I actually shared this with you. 01:08:58.800 --> 01:09:02.100 Maybe maybe back in July or maybe May? 01:09:02.160 --> 01:09:09.120 Um, basically, Backstage. It's 01:09:09.120 --> 01:09:14.760 supposed to be this internal developer platform that gives you a view across all 01:09:14.760 --> 01:09:18.900 of your organisations, all of the microservices and products that are there, where 01:09:18.900 --> 01:09:20.520 does the code live, who are the owners? 01:09:20.580 --> 01:09:24.060 What are the services and production, where is the dashboards, all the operational 01:09:24.060 --> 01:09:26.400 data, exactly what an agent needs, right? 01:09:26.460 --> 01:09:29.520 So a lot of it, they have rebuilt as MCP service. 01:09:29.580 --> 01:09:34.080 So when you have backstage already, and then you create MCP servers, and then you 01:09:34.080 --> 01:09:39.480 have, like, basically it becomes a backend for all of your MCP data, then your 01:09:39.480 --> 01:09:42.120 agents can fetch everything they need whenever they need to work on something. 01:09:42.180 --> 01:09:47.040 If they need to build code for a service, they can find all the consumer services, 01:09:47.040 --> 01:09:49.620 they can find information about the operationals. 01:09:49.680 --> 01:09:56.340 So I, I, I was looking into backstage as a context provider for agents months 01:09:56.340 --> 01:10:01.980 ago. Mostly because somebody pointed it out at me that they had adopted backstage 01:10:01.980 --> 01:10:07.500 after I left and MCPs were like a massive enabler 01:10:07.500 --> 01:10:09.540 on top of vaccine. Oh yeah, for sure. 01:10:09.600 --> 01:10:14.100 So that's where... So when we talk about Spotify. 01:10:14.160 --> 01:10:18.180 I know that there's somebody that's very publicly talking about how they are delivering 01:10:18.180 --> 01:10:21.300 with agents, and I think it's a name that you mentioned, and then you have people 01:10:21.300 --> 01:10:24.900 like... No, not very active nowadays. 01:10:25.020 --> 01:10:30.600 So Hashimoto, he mentioned that Spotify is very publicly talking about 01:10:30.600 --> 01:10:34.740 their identic adoption, but that Spotify is probably the least preferred service 01:10:34.740 --> 01:10:36.960 that he's using. Like there's tweets out there like that. 01:10:37.020 --> 01:10:40.200 Like people are complaining. They're talking about how they are delivering quickly 01:10:40.200 --> 01:10:43.800 with agents, but people are saying the quality and the errors in the bugs. 01:10:43.860 --> 01:10:46.080 Honestly, I'm using Spotify. I don't have any issues with it. 01:10:46.440 --> 01:10:53.520 But anyway, the whole point I feel as a product hasn't really wowed me in the last... 01:10:53.580 --> 01:10:57.300 The reason we talk about Spotify is because there's this interesting Netflix series, 01:10:57.300 --> 01:11:02.460 they have an actual protocol, because we were talking about these deep technical 01:11:02.460 --> 01:11:06.660 engineering blocks that are interesting, and there's, of course, this Cubnitis 01:11:06.660 --> 01:11:11.160 failure stories, which is a very interesting collection of those, but as you can 01:11:11.160 --> 01:11:14.400 see, it's kind of outdated because the last one is 2023. 01:11:14.640 --> 01:11:17.940 But like, nobody cares about pre-AI. 01:11:18.000 --> 01:11:24.540 B-A before AI and after AI. That's that's 01:11:24.540 --> 01:11:27.180 the new timeline, right? It's kind of sad. 01:11:27.240 --> 01:11:32.820 It's kind of sad. It's kind of sad that big frontier eye laps are putting out engineering 01:11:32.820 --> 01:11:38.580 articles that are not really interesting to read compared to pre-AI articles. 01:11:38.640 --> 01:11:40.020 it. 01:11:42.180 --> 01:11:47.700 Well, it's, I mean, it occupied us 01:11:47.700 --> 01:11:51.900 for a good half hour. I mean, it is interesting in a way what they're spewing out. 01:11:51.960 --> 01:11:57.600 It, it, it, the bottle. It occupied us to rent for 30 minutes about how low 01:11:57.600 --> 01:12:00.720 quality those articles are. For the last 30 minutes. 01:12:01.320 --> 01:12:03.180 Yeah, I agree. 01:12:04.260 --> 01:12:09.900 I was saying that there's better articles from the community, and I mean, I am 01:12:09.900 --> 01:12:15.720 not a fan of swim. I do not understand whites need it, but 01:12:15.720 --> 01:12:18.060 this is actually a very interesting read. 01:12:18.120 --> 01:12:24.240 I think anyone that basically puts that right, like a caveman, 01:12:24.240 --> 01:12:30.000 some right, their own words from their own human mind onto the 01:12:30.000 --> 01:12:32.340 screen. It's probably doing a better job. 01:12:33.540 --> 01:12:39.420 Someone who's being like, who's owning and being accountable for the 01:12:39.420 --> 01:12:42.240 stuff that they're writing. Wait a minute. 01:12:42.240 --> 01:12:47.100 Oh, no, don't quote me now. Wait a minute. 01:12:47.220 --> 01:12:49.320 Where's your enterprise struggle article? 01:12:49.380 --> 01:12:53.880 The struggle is real. You're going to point out some AIisms. 01:12:53.940 --> 01:12:57.720 I'm not pointing all the items. You said that it was the Iisms. 01:12:58.620 --> 01:13:03.180 I mean, this is actually, I'm proud of this document. 01:13:03.540 --> 01:13:09.180 Yeah. Um, I remember DHC saying 01:13:09.180 --> 01:13:13.020 something interesting. Oh, yeah, I saw this article. 01:13:13.080 --> 01:13:16.860 I think I saw his his his podcast before this article. 01:13:16.920 --> 01:13:20.040 Um, Or maybe I did see it from here. 01:13:20.100 --> 01:13:22.320 Let's give you credit. I probably saw this from yours. 01:13:22.380 --> 01:13:23.400 Oh, thanks, Vincent. Wow. 01:13:24.300 --> 01:13:30.120 Um, but I immediately sent, So remember how I feel spectriven development 01:13:30.120 --> 01:13:36.300 was not built for, um, like team collaboration, that I was like, multiplayer? 01:13:36.360 --> 01:13:39.720 Yeah. So D Chase is the opposite. 01:13:39.840 --> 01:13:43.500 I don't know. I'm going to disagree with the HH here. 01:13:44.400 --> 01:13:48.960 Well, the the, the multiplayer aspect is still not sold. 01:13:49.020 --> 01:13:54.960 There was, I, um, do you use Claude artifacts because I 01:13:54.960 --> 01:13:56.880 only just realised, I don't know who told me. 01:13:56.940 --> 01:14:00.960 Was it you? I only just realised the other day with Claude artefacts, you can, 01:14:00.960 --> 01:14:04.260 you can comment on them. Collaboratively. 01:14:04.320 --> 01:14:09.180 And then you can send the comments essentially back to the Claude instance, and 01:14:09.180 --> 01:14:14.640 then you can iterate like that. So, for example, earlier this week, I was working 01:14:14.640 --> 01:14:20.100 on a mermaid diagram, and then I got some comments that were shared on artefact. 01:14:20.160 --> 01:14:24.300 Yeah, you see that, just to the left of share. 01:14:24.360 --> 01:14:27.780 You can highlight different objects and then comment on them. 01:14:27.840 --> 01:14:31.920 Hmm. And then you can send, and then you can send it to Claude. 01:14:31.980 --> 01:14:35.580 So you can get multiplayer experience this way. 01:14:35.640 --> 01:14:37.860 It's a bit low-fi, but it works. 01:14:37.920 --> 01:14:39.960 And I was like, yes, this is cool. 01:14:40.980 --> 01:14:44.400 This is cooler. If you can do this. 01:14:47.460 --> 01:14:50.280 Oh. No, that's not how it works. 01:14:50.760 --> 01:14:56.580 Oh, okay. should explore. No, I showed you that you could do this on on Kiro 01:14:56.580 --> 01:14:57.180 artefacts. 01:14:58.200 --> 01:15:02.100 that you can you can leave. But I didn't know that you could do it. 01:15:02.160 --> 01:15:05.880 Kira takes some setting up, especially if you want to use OpenAI with your own 01:15:05.880 --> 01:15:12.240 key. This is out of the box, Claude code, you know, you don't even have to like... 01:15:12.300 --> 01:15:14.040 Well, it's my argument against Orca. 01:15:14.100 --> 01:15:20.820 Of course, your argument for Orca is I can use Codex and I can use... 01:15:20.880 --> 01:15:26.580 I can use cloud or codex. But my argument for against Orca is 01:15:26.580 --> 01:15:29.460 that Claude Code is giving me all of this out of the box already. 01:15:31.080 --> 01:15:36.360 I mean, you might be happy living in your anthropic world, but I see the value 01:15:36.360 --> 01:15:41.220 of using both. Or more than one ecosystem. 01:15:42.660 --> 01:15:44.880 I'm a heterogeneous kind of guy. 01:15:44.940 --> 01:15:47.580 I'm sad. I'm sad that it's one by one. 01:15:48.240 --> 01:15:50.640 Can you send them all at the same time? 01:15:50.700 --> 01:15:55.260 It's not great. It's very low-fi, as I mentioned, but it's... 01:15:55.260 --> 01:15:59.040 So this is something that... Oh my god, this is this is frustrating. 01:15:59.100 --> 01:16:00.840 This is something that Kiro does, right? 01:16:00.900 --> 01:16:05.340 Kiro allows you to put a bunch of comments, queue them up and send them all to 01:16:05.340 --> 01:16:07.800 Kiro all at once. Instead of like one by one. 01:16:07.860 --> 01:16:10.860 by the way. Yeah. Well, it's getting there. 01:16:10.920 --> 01:16:13.800 I guess I guess Claude is working towards... 01:16:13.860 --> 01:16:15.960 I mean, and like, it's it's so simple. 01:16:16.020 --> 01:16:16.980 Don't know why it takes them so long. 01:16:16.980 --> 01:16:20.700 A web URL, but a colleague and asks for some feedback. 01:16:20.760 --> 01:16:25.800 I think this is probably something that's gonna be more common. 01:16:25.860 --> 01:16:27.480 I mean, I hope it's going to be more common. 01:16:27.540 --> 01:16:30.000 I mean, I think I think most better. 01:16:30.060 --> 01:16:33.060 It's really difficult to get feedback from people in some ways. 01:16:33.240 --> 01:16:37.140 So display not depth did this like back in March, right? 01:16:40.260 --> 01:16:45.960 You publish your article. You produce your artefact, and then you get, you can, 01:16:45.960 --> 01:16:51.120 you put it behind SSO, so you can share it, and then you can sign in, and then 01:16:51.120 --> 01:16:55.200 you can aggregate the comments. Like as you can see, multiplayer comments, and 01:16:55.200 --> 01:16:58.080 then you can send it across. So they did this back in March. 01:16:58.140 --> 01:17:03.000 Well, I'm afraid to say that Anthropic is caught up with this. 01:17:03.300 --> 01:17:08.280 Yeah. Because entropic, I always felt entropic was going to do this. 01:17:09.000 --> 01:17:11.520 as an enterprise offering, right? 01:17:11.580 --> 01:17:14.520 Artefacts always felt like you should be able to do this. 01:17:14.580 --> 01:17:16.860 And initially, when you did share, it was not available. 01:17:16.920 --> 01:17:19.740 You couldn't do it on team plan. 01:17:19.800 --> 01:17:22.380 You could only do it if you were in an organisation. 01:17:22.740 --> 01:17:27.240 Oh, no, I can. Oh, no, I can share it with people in my in the team as well. 01:17:27.300 --> 01:17:31.980 Yeah, I'm not sure about that. Anyway, let's end the podcast. 01:17:32.040 --> 01:17:35.280 I really need to get on back onto some things. 01:17:35.340 --> 01:17:37.380 Well, we cover what I wanted to cover. 01:17:37.440 --> 01:17:42.900 Um, hey, have you, I know, I think there was a few comments on the last podcast 01:17:42.900 --> 01:17:46.440 I uploaded. I don't know if you responded to them, but I would appreciate if you 01:17:46.440 --> 01:17:48.360 could respond to the comments. 01:17:49.080 --> 01:17:52.680 And thank anyone who got this far. 01:17:52.740 --> 01:17:56.160 I thank you and I and please be in touch. 01:17:56.220 --> 01:17:58.920 We'd love to hear from you I respond right now. 01:17:58.920 --> 01:18:01.260 Vincent can respond to your questions. 01:18:01.560 --> 01:18:04.920 There is, there was one comment you linked to me. 01:18:05.040 --> 01:18:07.200 There was one comment you linked to. 01:18:07.260 --> 01:18:12.900 Ah, I remember there was a comment asking why don't we just use something 01:18:12.900 --> 01:18:17.220 else? Yeah. I didn't haven't used something else. 01:18:17.280 --> 01:18:22.080 And I was like, but that wasn't the point because Hiro AC like Kiro ACP allowed 01:18:22.080 --> 01:18:26.340 me to use Codex and I guess it only makes sense if I published it. 01:18:26.400 --> 01:18:30.660 Unfortunately, I think I missed the window to get my contribution to Kiro Crew 01:18:30.660 --> 01:18:33.720 merged as I... What does ACP stand for again? 01:18:34.620 --> 01:18:39.240 Agent client protocol. So the client is a dashboard. 01:18:39.300 --> 01:18:44.940 The agent is the is the agent. Oh yeah, yeah, yeah, yeah. 01:18:45.060 --> 01:18:48.540 It's a standard by, yeah. I forgot about that. 01:18:48.600 --> 01:18:51.840 Oh man. So what's happening. Okay, cool. 01:18:51.900 --> 01:18:55.260 I'm gonna publish this podcast. What number would it be? 01:18:55.800 --> 01:19:00.360 You figure it out later. You comment below and think of it. 01:19:00.420 --> 01:19:03.000 Give me an idea what the title of this one should be. 01:19:03.060 --> 01:19:06.720 What did we talk about? Um, renting. 01:19:08.040 --> 01:19:12.300 God, okay. I'll ask AI to come up with the title. 01:19:12.360 --> 01:19:15.000 Okay. See you. Thanks, Vincent. Bye.