1 00:00:00,129 --> 00:00:02,349 AI is really shaking up everything. 2 00:00:02,676 --> 00:00:07,296 I think putting software as a entire, like industry a little bit in a different 3 00:00:07,296 --> 00:00:12,976 light now, and, I think this does not make local-first at all, less relevant. 4 00:00:13,246 --> 00:00:16,366 I think this is, as for most things, a huge tailwind. 5 00:00:16,786 --> 00:00:21,629 But given that, both of us have been really like, immersed in that new AI 6 00:00:21,629 --> 00:00:26,393 space as well, now, I think we can come with a perspective where we can speak 7 00:00:26,393 --> 00:00:31,883 as insiders from the local-first space, but also as insiders from the AI space. 8 00:00:32,183 --> 00:00:35,783 And with that fresh perspective, I think we wanted to give this year's 9 00:00:35,783 --> 00:00:37,963 Local-First Conf a new, code of paint. 10 00:00:39,343 --> 00:00:41,143 Hey Adam, so great to have you back. 11 00:00:41,143 --> 00:00:41,833 How are you doing? 12 00:00:42,133 --> 00:00:42,593 Hey Johannes. 13 00:00:42,613 --> 00:00:43,123 I'm good. 14 00:00:43,213 --> 00:00:44,908 Looking forward to the arrival of Spring here. 15 00:00:46,423 --> 00:00:50,863 So I was just joking that the two of us have been on the last 16 00:00:51,073 --> 00:00:56,413 localfirst.fm episode since it hasn't really been on a regular cadence 17 00:00:56,413 --> 00:00:57,883 over the last couple of months. 18 00:00:57,883 --> 00:01:02,756 I'm taking a bit of a, pause from the podcast to reorient myself 19 00:01:02,756 --> 00:01:07,916 in this new crazy world of AI and coding agents, et cetera. 20 00:01:07,916 --> 00:01:11,396 And like, I wanted to get my own hands even more dirty again. 21 00:01:11,696 --> 00:01:16,916 And yeah, little bit recalibrate, and I think that's also gonna be quite the theme 22 00:01:16,946 --> 00:01:19,429 for, an announcement we're making today. 23 00:01:19,863 --> 00:01:22,083 And to foreshadow a little bit. 24 00:01:22,273 --> 00:01:25,663 What this year's Local-First Conference will be about. 25 00:01:25,793 --> 00:01:29,800 So with that out of the way, yeah, curious what's on your mind, Adam? 26 00:01:31,090 --> 00:01:33,220 Well, obviously as we ramp up to this Conference and we're, 27 00:01:33,250 --> 00:01:35,590 we're doing something pretty different this year, I think so. 28 00:01:35,966 --> 00:01:36,250 I don't know. 29 00:01:36,250 --> 00:01:40,057 Maybe I can give some, some history for, I espect a lot of your listeners have either 30 00:01:40,057 --> 00:01:41,767 been to the Conference or know about it. 31 00:01:42,055 --> 00:01:48,483 but just for context, this will be our third year and we started in 2024. 32 00:01:48,731 --> 00:01:52,230 we really weren't sure honestly, if there was enough demand for it. 33 00:01:52,255 --> 00:01:56,075 You sort of had the idea of like, Hey, I think there's space for or 34 00:01:56,075 --> 00:01:59,595 interest, you know, a latent interest in local-first and we could build a 35 00:01:59,595 --> 00:02:03,455 conference around it on data ownership and CRDTs and all that goodness is 36 00:02:03,475 --> 00:02:07,685 things that yeah, need a venue and a place to meet in person and share ideas. 37 00:02:08,135 --> 00:02:09,935 And we were genuinely really concerned. 38 00:02:09,935 --> 00:02:12,215 We just wouldn't sell enough tickets and we'd lose money on the 39 00:02:12,215 --> 00:02:13,355 venue and all this kind of thing. 40 00:02:13,595 --> 00:02:17,255 Instead, what happened was, I think our venue held like 150 or 180, something 41 00:02:17,255 --> 00:02:18,635 like that, we sold it out within a week. 42 00:02:18,965 --> 00:02:22,085 It was actually kind of embarrassing because we had to be turning wonderful 43 00:02:22,085 --> 00:02:24,995 people away and plus ones for our speakers and things like that, just 44 00:02:25,025 --> 00:02:27,605 'cause we were at the venue capacity, so. 45 00:02:27,845 --> 00:02:32,855 And for me it was also a big question about bringing together eclectic mix 46 00:02:32,855 --> 00:02:36,905 of people, you had both academic people who had been working on CRDT theorems 47 00:02:36,905 --> 00:02:43,295 and things for a decade and more kind of fringy, you know, Brett Victor, Alan Kay. 48 00:02:43,295 --> 00:02:48,072 Tools for thought, perspectives with more practical software 49 00:02:48,072 --> 00:02:51,909 engineers build React apps, just looking for the newest and greatest 50 00:02:51,909 --> 00:02:53,395 technology to make their jobs easier. 51 00:02:53,395 --> 00:02:56,485 Startups, there was, sync vendors were starting to become a little 52 00:02:56,485 --> 00:02:59,205 bit of a thing then, and, you know, would these people even gel? 53 00:02:59,205 --> 00:03:00,715 And I think the answer was yes. 54 00:03:00,715 --> 00:03:02,755 The event was a huge success one day. 55 00:03:02,755 --> 00:03:04,675 Single track, everyone loved it. 56 00:03:05,035 --> 00:03:08,125 and it was a pretty obvious thing then the following year, 2025 to 57 00:03:08,185 --> 00:03:10,095 expand it, we had a venue for 350. 58 00:03:10,645 --> 00:03:14,746 We did two full days plus a community day, to more talks, more people. 59 00:03:14,746 --> 00:03:17,896 Everything was 2 to 3x the scale. 60 00:03:18,106 --> 00:03:21,242 And I think it really showed that we had this, community here and 61 00:03:21,242 --> 00:03:23,732 that, that there was a lot of cohesion and people wanted to come. 62 00:03:24,137 --> 00:03:27,467 Come back to it at the same time, I think, because sync and sync engines were 63 00:03:27,467 --> 00:03:31,187 having a pretty big moment last year that ended up being a, at least one of the 64 00:03:31,187 --> 00:03:35,147 day, was largely given over to that topic and we explored that pretty thoroughly. 65 00:03:36,017 --> 00:03:40,067 So now we find ourselves thinking about, okay, one, we just enjoy 66 00:03:40,067 --> 00:03:42,797 bringing this group of people together in this community of people together. 67 00:03:42,797 --> 00:03:46,097 We want to get our friends back together in Berlin in the spring or the 68 00:03:46,097 --> 00:03:48,527 summertime to just spend time together. 69 00:03:48,837 --> 00:03:51,987 But at the same time, we also know the industry's changing and we also 70 00:03:51,987 --> 00:03:55,677 know that it would be a little boring to just retread the same ground. 71 00:03:55,887 --> 00:03:58,407 So that's causing us to want to be a little more creative with the 72 00:03:58,407 --> 00:04:01,437 direction we're going this year, which is I think why we wanted to get 73 00:04:01,437 --> 00:04:02,917 together and talk about that today. 74 00:04:02,917 --> 00:04:03,197 Yeah, and I think if you're 75 00:04:05,436 --> 00:04:10,116 just now looking back over the last two to three years, I feel that the 76 00:04:10,116 --> 00:04:12,516 world is just such a different place. 77 00:04:12,765 --> 00:04:17,245 When we look back then, local-first was really like, a wake up call 78 00:04:17,288 --> 00:04:18,948 in this kinda like very almost. 79 00:04:19,230 --> 00:04:23,946 not by today's standard, almost like static world, where things have 80 00:04:24,066 --> 00:04:28,416 feel like seem pretty settled in terms of technologies, et cetera. 81 00:04:28,416 --> 00:04:30,336 And we have like our best practices. 82 00:04:30,696 --> 00:04:35,766 And local-first at that point was much more of sort of like a provocation, 83 00:04:35,766 --> 00:04:40,400 hey, we can think differently about this and like rethink, the 84 00:04:40,440 --> 00:04:42,420 typical like cloud model, et cetera. 85 00:04:42,420 --> 00:04:44,100 And like it stirred up quite a bit of dust. 86 00:04:44,100 --> 00:04:47,490 And I think this is what, what got people really interested about like 87 00:04:47,490 --> 00:04:51,510 insiders who've been thinking about those lines for like many years. 88 00:04:51,710 --> 00:04:55,400 Some even like, decades and have always built software this way. 89 00:04:55,910 --> 00:05:01,281 And, then last year's conference I think has really shown a huge step to the year 90 00:05:01,281 --> 00:05:06,981 before in terms of like how much further, how much mature the ecosystem has gotten. 91 00:05:07,731 --> 00:05:11,244 But, I think it would be, like this year is different. 92 00:05:11,507 --> 00:05:16,097 It's not just like, yes, the ecosystem has gotten a lot more mature and like 93 00:05:16,097 --> 00:05:20,057 all of those technologies are more production ready, there's like more 94 00:05:20,057 --> 00:05:24,647 players, et cetera, but the world around it is like, entirely different. 95 00:05:24,777 --> 00:05:26,997 AI is really shaking up everything. 96 00:05:27,324 --> 00:05:31,944 I think putting software as a entire, like industry a little bit in a different 97 00:05:31,944 --> 00:05:37,624 light now, and, I think this does not make local-first at all, less relevant. 98 00:05:37,894 --> 00:05:41,014 I think this is, as for most things, a huge tailwind. 99 00:05:41,434 --> 00:05:46,277 But given that, both of us have been really like, immersed in that new AI 100 00:05:46,277 --> 00:05:51,041 space as well, now, I think we can come with a perspective where we can speak 101 00:05:51,041 --> 00:05:56,531 as insiders from the local-first space, but also as insiders from the AI space. 102 00:05:56,831 --> 00:06:00,431 And with that fresh perspective, I think we wanted to give this year's 103 00:06:00,431 --> 00:06:02,611 Local-First Conf a new, code of paint. 104 00:06:03,204 --> 00:06:07,477 this is where, you've been this year leading the effort around the CFP. 105 00:06:08,047 --> 00:06:13,091 And yeah, so I'm excited to, talk through the different themes we're expecting 106 00:06:13,267 --> 00:06:18,157 for the conference where we have a set of curated speakers already, but 107 00:06:18,247 --> 00:06:23,071 we think, the best content is really brought up from the community, who is 108 00:06:23,071 --> 00:06:25,321 building things in the local-first space. 109 00:06:25,321 --> 00:06:29,304 And this is, like the people who are maybe listening to this podcast, or who 110 00:06:29,304 --> 00:06:31,911 have attended, the last few conferences. 111 00:06:32,271 --> 00:06:35,114 So yeah, this is, what I would love to talk to you about. 112 00:06:35,401 --> 00:06:38,491 Yeah, well, maybe we can just start with that call to action really upfront, which 113 00:06:38,491 --> 00:06:42,301 is we're broadening the umbrella for what kind of talks we want a lot this year. 114 00:06:42,601 --> 00:06:45,811 And so if you do nothing else, you can just stop listening to the podcast. 115 00:06:45,811 --> 00:06:47,101 Go read the CFP page. 116 00:06:47,251 --> 00:06:49,951 If you think there's any chance of what you're working on, which has something 117 00:06:49,951 --> 00:06:55,216 to do with empowering users or making, computers and software and the internet 118 00:06:55,516 --> 00:07:00,706 freer and more capable for enhancing human life, then you should submit 119 00:07:00,706 --> 00:07:02,426 a talk that's your CTA right there. 120 00:07:02,671 --> 00:07:04,078 To drill, one level deeper. 121 00:07:04,418 --> 00:07:07,448 you already mentioned the AI topic and obviously that feeds into this. 122 00:07:07,628 --> 00:07:11,288 There's also the desire to continue to build and expand on the themes of this 123 00:07:11,288 --> 00:07:15,038 community and respond to other things that are happening in the tech industry. 124 00:07:15,488 --> 00:07:19,448 On the AI side, obviously, where I believe, you know, creation of software 125 00:07:19,448 --> 00:07:22,838 is changing fundamentally with AI assisted coding, and I think people in our 126 00:07:22,838 --> 00:07:25,118 community have a very wide mix of views. 127 00:07:25,118 --> 00:07:28,935 There's people who come from a. very skeptical point of view, I probably 128 00:07:28,935 --> 00:07:32,578 lean more in that direction myself, especially up until recently, there's 129 00:07:32,578 --> 00:07:36,328 some bubble, it's gonna burst soon, it's overhyped, it's, you know, whatever. 130 00:07:36,625 --> 00:07:39,271 And on the other end, there's probably plenty of maximalist, people that, 131 00:07:39,571 --> 00:07:43,351 you know, I think it's not even such a bold thing now to say that 2026 might 132 00:07:43,351 --> 00:07:48,211 be the last year that humans do all that much direct coding in terms of 133 00:07:48,211 --> 00:07:50,491 like writing out programming syntax. 134 00:07:50,672 --> 00:07:54,525 and I suspect people listening fall in a wide range of, views there, and 135 00:07:54,525 --> 00:07:56,295 we wanna make room for all of that. 136 00:07:56,295 --> 00:08:00,075 But, you know, at the same time, we're acknowledging that this industry shift is 137 00:08:00,075 --> 00:08:04,365 happening while also staying grounded in our values and what we're all here for. 138 00:08:05,148 --> 00:08:08,971 For the record, October, 2025 was the last time I wrote a 139 00:08:08,971 --> 00:08:10,621 single line of code by myself. 140 00:08:10,981 --> 00:08:14,088 So, but I'm leaning into that, quite heavily. 141 00:08:14,358 --> 00:08:16,699 But I think as you say that's a wide spectrum. 142 00:08:16,939 --> 00:08:21,356 And nonetheless, I think that makes, even more clear that we need really 143 00:08:21,356 --> 00:08:26,101 good principles for the software we're creating, and that we also want to use, I 144 00:08:26,101 --> 00:08:31,846 think, software through AI, like through the terms like slop software, et cetera, 145 00:08:31,846 --> 00:08:36,450 I think has made it even more, even more important to really care about like what 146 00:08:36,450 --> 00:08:41,730 we're using, not just for the perspective of that "it's nice to use", but also for 147 00:08:41,730 --> 00:08:47,223 the perspective of like, data ownership, privacy, security, all of those. 148 00:08:47,443 --> 00:08:51,703 And I think local-first is still the guiding light for us. 149 00:08:52,000 --> 00:08:57,640 It's kinda like the gold standard for what software should feel like, what 150 00:08:57,640 --> 00:08:59,770 software should be architected around. 151 00:09:00,190 --> 00:09:04,840 And I think this is where we hopefully get the best of both worlds, where we 152 00:09:04,840 --> 00:09:09,790 can have the cake and eat it too in that regard that we can dream about 153 00:09:09,790 --> 00:09:12,640 software that follows those principles. 154 00:09:12,640 --> 00:09:17,470 But now it's a lot more feasible to build those and to build those, 155 00:09:17,470 --> 00:09:22,330 not just for the sake of like building software for potentially 156 00:09:22,570 --> 00:09:27,313 like a million target users, but for yourself, like personal scale software. 157 00:09:27,763 --> 00:09:32,233 And this is where we had this amazing closing keynote from Maggie 158 00:09:32,233 --> 00:09:36,163 Appleton at the first Local-First Conference about Homecooked software. 159 00:09:36,553 --> 00:09:38,353 And I think in a way this year, 160 00:09:38,413 --> 00:09:41,863 barefoot developers, I think was the term she, she used. 161 00:09:41,863 --> 00:09:42,073 That's right. 162 00:09:42,103 --> 00:09:44,743 And certainly what you could do with language model assisted coding 163 00:09:44,743 --> 00:09:47,323 back then was, you know, much. 164 00:09:47,643 --> 00:09:49,623 more limited than what you can do today. 165 00:09:49,623 --> 00:09:52,972 But even then, she predicted that, which I think is a good, a good illustration 166 00:09:52,972 --> 00:09:55,732 of the kinds of people that are in our community that, you know, were 167 00:09:55,732 --> 00:09:57,892 values driven, but very forward facing. 168 00:09:57,892 --> 00:10:00,642 She, she saw what was coming years before. 169 00:10:00,919 --> 00:10:01,399 Exactly. 170 00:10:01,399 --> 00:10:05,709 And like now, this year's conference is basically like, really like double 171 00:10:05,729 --> 00:10:10,762 clicking on her entire vision here where like everyone can dream about 172 00:10:10,762 --> 00:10:13,012 software and like now make it a reality. 173 00:10:13,372 --> 00:10:18,859 And local-first doesn't just make it, more secure and private, but like overall 174 00:10:18,859 --> 00:10:23,402 also like solves a whole bunch of other problems that is still hard with software. 175 00:10:23,462 --> 00:10:27,392 So I'm really excited how like those two or the, how the conference 176 00:10:27,452 --> 00:10:31,615 can be a place where those two different worlds can come together. 177 00:10:31,615 --> 00:10:37,585 All of like the relentless, accelerated AI community, who wants to, build things 178 00:10:37,585 --> 00:10:41,035 really quickly, but also the people who like really care about software 179 00:10:41,035 --> 00:10:43,015 quality and like those principles. 180 00:10:43,345 --> 00:10:48,879 And this is something where I think it can be, a really nice meeting space for folks. 181 00:10:48,939 --> 00:10:53,852 So, Adam, how about you wanna give us a broad overview of the different 182 00:10:53,852 --> 00:10:57,372 kinda like, themes and then we walk through them one at a time? 183 00:10:57,682 --> 00:10:58,192 Yeah, sure. 184 00:10:58,192 --> 00:11:01,451 So the overall theme for the conference this year is user empowerment. 185 00:11:01,866 --> 00:11:04,371 that comes from, actually a panel we had last year. 186 00:11:04,371 --> 00:11:07,751 I host a panel and one of the, panelists was Martin Kleppmann, who 187 00:11:07,751 --> 00:11:09,371 this year is a keynote speaker for us. 188 00:11:09,371 --> 00:11:12,984 Obviously he's really one of the, founding fathers of, local-first. 189 00:11:13,254 --> 00:11:17,454 But the way he described it was, we were talking here a little bit about the kind 190 00:11:17,454 --> 00:11:22,044 of the specific implementation details of sync engines and CRDTs versus the 191 00:11:22,344 --> 00:11:24,024 broader set of values and principles. 192 00:11:24,474 --> 00:11:26,004 And I think he put it really well in saying. 193 00:11:26,394 --> 00:11:27,654 It's all about user empowerment. 194 00:11:27,654 --> 00:11:33,144 If something gives users more empowerment, more control, more capabilities, more 195 00:11:33,144 --> 00:11:38,434 options, more freedom, then that to him falls under the umbrella of local-first. 196 00:11:38,554 --> 00:11:41,914 And if it's something that sort of takes agency away, which there are times 197 00:11:41,914 --> 00:11:45,154 where that makes sense, especially in consumer software or whatever, that 198 00:11:45,214 --> 00:11:49,018 people want less control over how they're using their computers, that's fine. 199 00:11:49,018 --> 00:11:51,898 But that, leaves it outside the realm of local-first. 200 00:11:51,918 --> 00:11:54,898 So we're borrowing that idea and expanding on it. 201 00:11:54,928 --> 00:11:58,438 User empowerment, greater agency, obviously a lot of that tends 202 00:11:58,438 --> 00:12:00,028 to flow through data ownership. 203 00:12:00,058 --> 00:12:03,598 'cause the data ultimately when you're creating documents on a 204 00:12:03,598 --> 00:12:07,108 computer, the data and feeling like it belongs to you is, tends to be 205 00:12:07,108 --> 00:12:09,149 core to that feeling of empowerment. 206 00:12:09,479 --> 00:12:13,717 So with that overall theme, took together with our, collaborator, Eileen Wagner, 207 00:12:13,717 --> 00:12:18,682 also a long time member of the community who is, helping us here with the CFP. 208 00:12:19,012 --> 00:12:21,702 But she basically sorted us into three categories here. 209 00:12:21,702 --> 00:12:25,482 We've got new territories, which of course includes plenty of AI stuff. 210 00:12:25,482 --> 00:12:29,232 We've got local-first maturity, which is trying to see, you 211 00:12:29,232 --> 00:12:30,642 know, now we're pretty far along. 212 00:12:30,642 --> 00:12:32,532 There's a lot of great sync engine vendors. 213 00:12:32,532 --> 00:12:35,922 There's some real world success stories, but there's also 214 00:12:35,982 --> 00:12:39,732 interesting new challenges to face user experience challenges, legal 215 00:12:39,732 --> 00:12:41,622 challenges, or legal frameworks. 216 00:12:41,955 --> 00:12:45,945 And then where we're really trying to expand here is the larger ecosystem. 217 00:12:45,945 --> 00:12:49,305 And there's a lot of exciting stuff happening in, whether it's in something 218 00:12:49,305 --> 00:12:53,142 like the area of game developments, but also something like social media. 219 00:12:53,142 --> 00:12:57,012 I think, you know, with ATproto and to a lesser extent sort of the masked 220 00:12:57,132 --> 00:13:00,925 on ActivityPub world of things is having their own moment and kind 221 00:13:00,925 --> 00:13:05,815 of causing people to ask questions about data ownership and freedom and 222 00:13:05,875 --> 00:13:09,985 decentralization in this realm that traditionally has been a very kind of 223 00:13:09,985 --> 00:13:14,875 centralized, closed, big tech, you know, inscrutable algorithm world of things. 224 00:13:15,105 --> 00:13:19,225 And to what extent are data ownership and local-first values applicable there. 225 00:13:19,769 --> 00:13:24,299 So those three frames, the kind of new territories, the maturity of local-first 226 00:13:24,599 --> 00:13:28,559 and the larger ecosystem are where we have kind of a few subtopics for each 227 00:13:28,559 --> 00:13:32,885 one, and we're trying to hope, hoping to get a good coverage on, CFP submissions. 228 00:13:33,392 --> 00:13:33,812 Awesome. 229 00:13:33,812 --> 00:13:36,812 Thanks so much for providing this great overview. 230 00:13:36,842 --> 00:13:41,102 I would actually love to double click on each of those and dig a 231 00:13:41,102 --> 00:13:46,219 little bit more in, since I think each topic has, so much potential. 232 00:13:46,219 --> 00:13:49,852 And I think we could, like, if we'd had a lot more time and space for the 233 00:13:49,852 --> 00:13:54,562 conference, we could almost like make each of those their explicit like theme day. 234 00:13:54,622 --> 00:13:59,212 And so now we only have, well, those two days for the conference. 235 00:13:59,443 --> 00:14:01,364 for the talks in one extra day. 236 00:14:01,554 --> 00:14:03,344 So we need to curate and pick. 237 00:14:03,624 --> 00:14:08,184 But I think this gives us like a really nice broad buffet of like different, 238 00:14:08,437 --> 00:14:10,427 topics and talks to choose from. 239 00:14:10,637 --> 00:14:16,234 And ideally, those, themes speak to people who want to submit a talk. 240 00:14:16,474 --> 00:14:18,544 So yeah, let's go through them. 241 00:14:18,549 --> 00:14:23,627 So maybe starting with the, local-first maturity, maybe that is, where we, can 242 00:14:23,627 --> 00:14:29,000 reconnect the most from where we've left off the first two years of the conference 243 00:14:29,000 --> 00:14:34,024 and also Sync Conf and, I think this is like local-first maturity is really like 244 00:14:34,024 --> 00:14:38,327 that that is the essence of what last year's conferences, were about, where we 245 00:14:38,327 --> 00:14:43,024 had local-first success stories, companies like Linear, et cetera come to mind. 246 00:14:43,254 --> 00:14:45,414 The UX challenges of local-first. 247 00:14:45,414 --> 00:14:47,124 This is, you've mentioned Eileen before. 248 00:14:47,304 --> 00:14:51,864 She's given an amazing talk about that last year legal framework. 249 00:14:51,894 --> 00:14:56,147 Maybe a little bit more top of mind this year around, let's start with 250 00:14:56,147 --> 00:14:58,067 those like local-first success stories. 251 00:14:58,067 --> 00:15:00,407 Anything that comes to mind for you in particular. 252 00:15:01,142 --> 00:15:05,275 Well, for sure the success stories should be, there's obviously people building 253 00:15:05,275 --> 00:15:08,635 libraries, you know, the sync engine vendors who are well funded and have a 254 00:15:08,635 --> 00:15:09,805 lot of customers and that sort of thing. 255 00:15:09,805 --> 00:15:13,075 But what I really wanna hear about is the end user applications. 256 00:15:13,165 --> 00:15:17,968 You mentioned Linear, Linear CTO Tuomas has been kind enough to give us 257 00:15:17,968 --> 00:15:21,897 excellent talks the last two years but I also think of on a smaller scale, 258 00:15:21,897 --> 00:15:24,413 there's something like, we had a talk last year that was about someone 259 00:15:24,413 --> 00:15:26,003 building a forestry application. 260 00:15:26,243 --> 00:15:30,228 So this is where someone walking around with a phone that needs to basically 261 00:15:30,228 --> 00:15:34,458 tag each tree in the forest with what's going on with it, whether it has a bug 262 00:15:34,518 --> 00:15:38,952 infestation, whether it needs whatever else and it's a pretty simple thing. 263 00:15:38,982 --> 00:15:42,387 And you know, I like that talk a lot because also they basically said, 264 00:15:42,387 --> 00:15:45,117 look, our data modeling is very simple. 265 00:15:45,297 --> 00:15:48,327 They built their own sync that needs to, of course, not only work 266 00:15:48,327 --> 00:15:50,007 offline, but for a big chunk of time. 267 00:15:50,007 --> 00:15:51,147 But it's a pretty simple queue. 268 00:15:51,447 --> 00:15:55,827 And unlike full text editing, that is a rich text editing that's very 269 00:15:55,827 --> 00:16:01,407 complicated you actually can accomplish all the goals of local-first with not a 270 00:16:01,407 --> 00:16:05,207 huge amount of engineering for some of these simpler, use cases and domains. 271 00:16:05,207 --> 00:16:07,774 So I think there's plenty of people in the community that have their 272 00:16:07,774 --> 00:16:11,644 hobby projects or whatever that they use together with their friends. 273 00:16:12,004 --> 00:16:16,444 But what I like to see we're far enough along now is what is at scale. 274 00:16:16,444 --> 00:16:20,231 I wanna see a Linear and Obsidian or some of these smaller scale things 275 00:16:20,231 --> 00:16:23,471 like the forestry app that have been doing this for a long time. 276 00:16:23,471 --> 00:16:26,561 You have real world data and you can talk about the challenges of debugging. 277 00:16:26,561 --> 00:16:29,681 You can talk about the challenges of accumulating data history. 278 00:16:29,681 --> 00:16:33,619 You can talk about what are the ugly and difficult parts of this 279 00:16:33,619 --> 00:16:36,709 that we're grappling with from using this frontier technology. 280 00:16:37,303 --> 00:16:38,323 Yeah, definitely. 281 00:16:38,353 --> 00:16:41,953 And I think there's, just thinking through like the software and new 282 00:16:41,953 --> 00:16:45,708 software that I've been using over the last like year or so, I've been 283 00:16:45,708 --> 00:16:50,298 using like Rayon for example, quite a bit, which is like a Figma like, 284 00:16:50,390 --> 00:16:54,562 software, but for more like architecture use cases as I'm like remodeling 285 00:16:55,065 --> 00:17:00,405 the house I grew up in and that's been like really nice to get sort of the 286 00:17:00,405 --> 00:17:05,055 benefits that I'm like spoiled with of like using production software 287 00:17:05,055 --> 00:17:07,305 like Figma now for more use cases. 288 00:17:07,305 --> 00:17:11,982 And I think that's just like one example for like, that seeping into more and 289 00:17:11,982 --> 00:17:14,502 more parts of the software we're using. 290 00:17:14,502 --> 00:17:18,025 And I think that will just, continue and like, particularly 291 00:17:18,025 --> 00:17:23,795 now where, AI is another driver of like personalized software. 292 00:17:24,335 --> 00:17:28,535 I think it's just like very clear that we want those benefits of 293 00:17:28,535 --> 00:17:32,015 like collaboration, that things should work offline, et cetera. 294 00:17:32,255 --> 00:17:37,175 Like if anything, those problems are now even more pressing, where when 295 00:17:37,175 --> 00:17:41,655 we have software that is like meant for individuals, and to have like 296 00:17:41,655 --> 00:17:46,896 that software, be shared, across like family members or, friends. 297 00:17:46,896 --> 00:17:50,646 This is where you don't like, you, probably don't build it out as like 298 00:17:50,646 --> 00:17:54,489 a SaaS architecture from the get go, but you're building this like 299 00:17:54,489 --> 00:17:56,109 in a different way to begin with. 300 00:17:56,399 --> 00:17:59,789 SQLite certainties having a bit of a moment again as well. 301 00:17:59,999 --> 00:18:02,609 So I think we're gonna see a lot more of that. 302 00:18:02,879 --> 00:18:05,429 Files are, are, are are the hot new thing again. 303 00:18:06,149 --> 00:18:06,509 Unix 304 00:18:06,509 --> 00:18:06,839 tools. 305 00:18:06,839 --> 00:18:06,959 Yeah. 306 00:18:07,049 --> 00:18:07,679 Honestly. 307 00:18:08,189 --> 00:18:12,706 Honestly, files could have almost been like another category, in itself. 308 00:18:12,706 --> 00:18:16,006 Like it should really also, and not to sound like a broken record, 309 00:18:16,006 --> 00:18:22,901 but like, yeah, with the AI hat on, files and folders are like the best 310 00:18:22,901 --> 00:18:25,328 foundation for agents to do good work. 311 00:18:25,748 --> 00:18:31,391 And it has, intentionally or not really, like, become, again, a main way of like 312 00:18:31,391 --> 00:18:33,821 how data collaboration works, et cetera. 313 00:18:34,241 --> 00:18:37,601 And so, yeah, I think there's a lot to be explored there. 314 00:18:37,711 --> 00:18:41,581 A lot of creative people doing weird, interesting things. 315 00:18:41,581 --> 00:18:46,128 So if you're, if you're like that, please submit a talk and, yeah, I 316 00:18:46,128 --> 00:18:50,164 think there's like whatever, you've been working on in regards to like 317 00:18:50,214 --> 00:18:54,931 traditional local-first in the sense of like the previous conferences, we would 318 00:18:54,931 --> 00:19:00,087 love to hear from you, but also around the UX challenges, of, local-first. 319 00:19:00,087 --> 00:19:04,167 I think we're still just scratching the surface of like trying to 320 00:19:04,227 --> 00:19:09,177 understand that, trying to like deal with the implications of that. 321 00:19:09,267 --> 00:19:13,977 I think we're allowing ourselves to dream a little bit bigger or a little 322 00:19:13,977 --> 00:19:16,257 bit more nichey with software now. 323 00:19:16,737 --> 00:19:22,887 And now that might require like new user experience paradigms and like, 324 00:19:22,887 --> 00:19:27,417 particularly also when it comes to extending collaboration just beyond 325 00:19:27,567 --> 00:19:33,760 people, but also like collaborating with like other agents, is another 326 00:19:33,760 --> 00:19:38,320 aspect that I think squarely fits into the realm of local-first as it's 327 00:19:38,320 --> 00:19:40,510 another like kind of trust boundary. 328 00:19:40,570 --> 00:19:43,120 You wanna understand what has happened. 329 00:19:43,150 --> 00:19:44,560 Do you want to collaborate? 330 00:19:44,590 --> 00:19:47,320 Do you wanna like detach, be by yourself? 331 00:19:47,650 --> 00:19:52,060 Do you want to like revoke authentication, authorization, et cetera. 332 00:19:52,537 --> 00:19:54,302 Maybe real time collaborate. 333 00:19:54,607 --> 00:19:59,193 I think there's a lot here, whether you want to look at us for the AI lens or 334 00:19:59,193 --> 00:20:03,723 not, but I think that area is like super underexplored and if you've been exploring 335 00:20:03,723 --> 00:20:05,427 that, we'd love to hear from you. 336 00:20:06,003 --> 00:20:06,183 Yeah. 337 00:20:06,183 --> 00:20:08,073 The UX challenges for me as part of the. 338 00:20:09,803 --> 00:20:13,360 I don't wanna say quite say fun part, but sort of it is because it's a chance 339 00:20:13,360 --> 00:20:15,550 to rethink those computing primitives. 340 00:20:15,550 --> 00:20:15,880 Right? 341 00:20:16,060 --> 00:20:19,510 And the files example is a good one to me, where files and the Unix 342 00:20:19,510 --> 00:20:25,270 methodology and everything as a tool and pipes and redirects was really kind 343 00:20:25,270 --> 00:20:27,040 of a first love for me in many ways. 344 00:20:27,040 --> 00:20:30,400 In computing, especially creative com, you know, computing that's 345 00:20:30,400 --> 00:20:31,750 designed for creating things. 346 00:20:32,395 --> 00:20:36,115 And then, you know, mobile basically abstracted files away. 347 00:20:36,115 --> 00:20:39,595 And that's kind of a win for most of the consumer use cases. 348 00:20:39,825 --> 00:20:43,845 but actually is kind of disempowering for creative work that you wanna do. 349 00:20:43,845 --> 00:20:46,925 And now again, the files are, somehow having a moment again. 350 00:20:46,955 --> 00:20:50,225 But now we live in this world of things like real-time collaboration, version 351 00:20:50,225 --> 00:20:54,612 control pull requests, yeah, live collaboration, async collaboration, and 352 00:20:54,612 --> 00:20:58,212 thinking about how all those people pieces fit together or how you can use those 353 00:20:58,212 --> 00:21:01,538 building blocks for the particular piece of software you happen to be building. 354 00:21:01,868 --> 00:21:03,758 That is very interesting to me. 355 00:21:04,442 --> 00:21:07,322 So another one is legal frameworks. 356 00:21:07,442 --> 00:21:11,102 So we are working with, like, we're building those systems, maybe just 357 00:21:11,102 --> 00:21:15,025 for us, maybe for others, but, like we're not building this in a vacuum. 358 00:21:15,055 --> 00:21:19,285 Sure, I can build some software that might just run like locally 359 00:21:19,315 --> 00:21:21,998 on my, phone or on my computer. 360 00:21:22,433 --> 00:21:25,343 Or like on like a little shared server in my home. 361 00:21:25,673 --> 00:21:28,203 But in that regard, no one cares. 362 00:21:28,263 --> 00:21:33,183 But like once I put it a little bit more out there for the world to see and to use, 363 00:21:33,390 --> 00:21:37,680 or possibly that it's like still meant to be used for a smaller group, but it's 364 00:21:37,680 --> 00:21:40,290 still accessible for others, possibly. 365 00:21:40,440 --> 00:21:42,300 So you need to think about the implications. 366 00:21:42,300 --> 00:21:43,950 You need to possibly lock it down. 367 00:21:43,950 --> 00:21:47,100 You need to constrain it, you need to set expectations. 368 00:21:47,370 --> 00:21:53,430 And one like way of setting expectations or being constrained is through like 369 00:21:53,430 --> 00:21:58,890 our legal system, whatever legislation we're falling into, like Germany, where 370 00:21:58,890 --> 00:22:00,690 we're doing this conference in like. 371 00:22:00,775 --> 00:22:06,165 Very particular set of legal constraints, also like European constraints, et cetera. 372 00:22:06,445 --> 00:22:08,365 But we're not building this in a vacuum. 373 00:22:08,365 --> 00:22:13,328 We need to like fit into the constraints, realities of the real world. 374 00:22:13,722 --> 00:22:16,272 One big question is like, where does data live? 375 00:22:16,572 --> 00:22:18,462 But also of like, other questions? 376 00:22:18,462 --> 00:22:22,782 Well, like what sort of rights do I as a user of software have? 377 00:22:22,782 --> 00:22:27,935 Like, GDPR is certainly something that really changed, the ecosystem 378 00:22:27,935 --> 00:22:30,025 quite a bit for better or for worse? 379 00:22:30,218 --> 00:22:35,118 I think actually one, funny aspect of GDPR, like maybe one intended or 380 00:22:35,118 --> 00:22:40,105 unintended consequence that I'm quite excited about is that GDPR has had a 381 00:22:40,105 --> 00:22:43,435 massive impact on data compatibility. 382 00:22:43,975 --> 00:22:47,215 Where you can now go to like pretty much every service and 383 00:22:47,215 --> 00:22:49,225 through the right to be forgotten. 384 00:22:49,465 --> 00:22:53,965 Like you can still export your data, but that exported data now actually 385 00:22:53,965 --> 00:22:56,065 gives you all the data for that service. 386 00:22:56,245 --> 00:22:57,940 So you can now like go to LinkedIn. 387 00:22:57,940 --> 00:23:02,680 It's like, hey, like give me all my data or like, go to x.com, give me all 388 00:23:02,680 --> 00:23:07,840 my data, and like now you can export that and like build tools with it. 389 00:23:08,050 --> 00:23:12,850 So I think that's like one interesting ways of like how a legal framework 390 00:23:12,850 --> 00:23:17,260 has unlocked data compatibility, even though that like, and not in 391 00:23:17,260 --> 00:23:20,380 a very rigorous way, but at least something better than nothing. 392 00:23:20,590 --> 00:23:23,290 And so that just like as one anecdote of like. 393 00:23:23,560 --> 00:23:27,520 How something has changed and unlocked new opportunities. 394 00:23:27,520 --> 00:23:29,950 And I think there will be a lot more like that. 395 00:23:30,400 --> 00:23:32,950 Like good developments, bad developments. 396 00:23:33,140 --> 00:23:37,160 Another real world analogy might be if you're going to Linear and you 397 00:23:37,160 --> 00:23:41,180 create a team, you need to set like, Hey, where should the data be stored? 398 00:23:41,300 --> 00:23:42,920 Do you want to have it stored in the US? 399 00:23:42,920 --> 00:23:44,690 You don't wanna have it stored in EU. 400 00:23:44,960 --> 00:23:47,120 Like, that's another way to like really model that. 401 00:23:47,650 --> 00:23:51,697 And I think that's just like, scratching the surface of how 402 00:23:51,697 --> 00:23:55,567 we need to bake those realities into the software we're building. 403 00:23:55,987 --> 00:24:01,133 And, I think there, this is an area that that's really worthwhile thinking through 404 00:24:01,523 --> 00:24:05,413 and be hopefully inspired by, new ideas. 405 00:24:06,210 --> 00:24:06,390 Yeah. 406 00:24:06,390 --> 00:24:09,570 I'm weirdly interested in the legal frameworks area and I hope we get 407 00:24:09,570 --> 00:24:11,370 some good submissions on that. 408 00:24:11,400 --> 00:24:15,605 I think it is natural for people in the tech world to dismiss or 409 00:24:15,605 --> 00:24:17,615 downplay that aspect of things. 410 00:24:17,615 --> 00:24:21,395 Especially maybe if you come from the Silicon Valley, American 411 00:24:21,395 --> 00:24:23,675 Maverick laws just get in my way. 412 00:24:23,675 --> 00:24:25,625 I'm gonna ignore them perspective. 413 00:24:25,825 --> 00:24:27,385 but yeah, we're, we're both here in Europe. 414 00:24:27,385 --> 00:24:30,802 I, two, two things that come to mind for me on that, that are 415 00:24:30,802 --> 00:24:32,057 interesting and of this moment. 416 00:24:32,898 --> 00:24:37,098 One is, yeah, you mentioned GDPR and much maligned in some ways. 417 00:24:37,098 --> 00:24:42,105 You know, the cookie banner thing is an unfortunate consequence of 418 00:24:42,105 --> 00:24:46,785 how that all played out, but it also has provided real front guidelines 419 00:24:46,785 --> 00:24:51,952 for handling of person identifying information that I think genuinely has 420 00:24:51,952 --> 00:24:54,322 made the internet a safer place, right? 421 00:24:54,322 --> 00:24:58,072 It's full of threat and attacks and that's only getting worse 422 00:24:58,072 --> 00:25:01,732 now that, that stuff can all be superpowered with language models. 423 00:25:01,732 --> 00:25:05,569 But, you know, when I was working on an AI powered browser called Dia 424 00:25:05,839 --> 00:25:10,166 last year, it was actually genuinely helpful to us to basically go and ask 425 00:25:10,166 --> 00:25:12,586 the question, what is GDPR recommended? 426 00:25:12,641 --> 00:25:17,021 And it basically comes down to, okay, we wanna be able to train on user data 427 00:25:17,021 --> 00:25:21,731 that's incredibly valuable to our users in the same way that, you know, Google's 428 00:25:21,941 --> 00:25:25,481 magic results and auto complete works so well because they're training it on 429 00:25:25,481 --> 00:25:26,621 what other people are searching for. 430 00:25:26,621 --> 00:25:28,481 And that's how it can seem to kind of read your mind. 431 00:25:28,901 --> 00:25:31,441 That's the same thing is true with, Cursor and the auto complete. 432 00:25:31,681 --> 00:25:35,856 So there's huge value in training on user data, even though there's, you know, but, 433 00:25:35,861 --> 00:25:38,854 but there's obviously some, things you need to be really cautious about there. 434 00:25:38,854 --> 00:25:41,104 And GDPR gives actually really good guidance. 435 00:25:41,104 --> 00:25:45,049 It basically says if you strip all information that can tie 436 00:25:45,049 --> 00:25:46,399 it back to a particular user. 437 00:25:46,399 --> 00:25:49,219 Certainly anything personally identifiable, like a name or an email, 438 00:25:49,309 --> 00:25:52,849 but also you're decoupling it from the user ID at the moment of collection 439 00:25:53,179 --> 00:25:57,709 and anonymizing there and you throw it away within 30 days, you're all good. 440 00:25:57,919 --> 00:26:00,919 That's a very simple framework and actually we could do a lot with that. 441 00:26:00,919 --> 00:26:04,069 We could really, go pretty far with training our own, 442 00:26:04,359 --> 00:26:06,519 models with data in that frame. 443 00:26:06,519 --> 00:26:08,859 And it was actually nice to have that, not just from a legal 444 00:26:08,859 --> 00:26:11,979 perspective, but from just a best practices guidance perspective. 445 00:26:12,479 --> 00:26:16,766 so GDPR has been a mixed bag in some ways, but has also I 446 00:26:16,766 --> 00:26:18,171 think provided genuine value. 447 00:26:18,376 --> 00:26:20,656 The other thing I would throw out that is also probably kind of 448 00:26:20,656 --> 00:26:24,706 from the Europe perspective that's top of mind now is sovereignty. 449 00:26:24,886 --> 00:26:27,076 Basically data sovereignty and software sovereignty. 450 00:26:27,376 --> 00:26:31,186 And I think a lot of the European Union as a whole, but also individual countries 451 00:26:31,186 --> 00:26:35,296 are thinking about how to decouple themselves a little bit from US providers. 452 00:26:35,596 --> 00:26:39,706 And I think we can put aside, or we don't wanna make a political conference 453 00:26:39,706 --> 00:26:43,756 or be too tied to whatever current event happens to be in the news at the moment. 454 00:26:43,936 --> 00:26:47,416 But I think in general, this is a good perspective for nations to have 455 00:26:47,416 --> 00:26:50,626 and it is just a scaled up version of local-first data ownership. 456 00:26:51,076 --> 00:26:55,306 It is just a version of the same reason I say it's good for me to have control 457 00:26:55,306 --> 00:26:59,356 over my data and have the ability to make choices about it, not only in 458 00:26:59,356 --> 00:27:00,706 the short term, but in the long term. 459 00:27:00,706 --> 00:27:01,696 Where do I back it up? 460 00:27:01,906 --> 00:27:04,696 Can I delete it, can I duplicate it, et cetera. 461 00:27:04,876 --> 00:27:06,496 There's a version of that for nations. 462 00:27:06,496 --> 00:27:09,792 And I think that's a little bit what, Europe is grappling with right now. 463 00:27:09,792 --> 00:27:13,662 And I think totally separate from anything happened in the current moment. 464 00:27:13,662 --> 00:27:15,387 That's a good thing to think about. 465 00:27:15,387 --> 00:27:19,591 That's a long-term thing that helps create greater agency and autonomy 466 00:27:19,591 --> 00:27:22,402 and security for your nation, just as it does for an individual. 467 00:27:23,454 --> 00:27:27,737 And I think there's also really like a window of opportunity, around that, where 468 00:27:27,947 --> 00:27:32,594 like, whether, what do you wanna call, like, big players like Monopoly or not? 469 00:27:32,827 --> 00:27:37,474 that aside, I think there are now real cases of precedent where I think certain 470 00:27:37,474 --> 00:27:42,274 German states, for example, have made the switch away from Microsoft to like 471 00:27:42,274 --> 00:27:44,637 some, open source software, et cetera. 472 00:27:44,637 --> 00:27:50,040 And like, I think that sets a, like me as someone who has like, very good 473 00:27:50,040 --> 00:27:53,850 feelings towards open source and wants to see open source software succeed, 474 00:27:54,240 --> 00:27:56,070 I think there's new momentum here. 475 00:27:56,340 --> 00:27:59,640 And I think local-first can really also through that perspective, 476 00:27:59,910 --> 00:28:01,890 play a more mainstream role. 477 00:28:02,125 --> 00:28:05,515 Since like you don't just, yes, you can have like your open source react 478 00:28:05,515 --> 00:28:09,055 components, but that does make like fully functional software where you 479 00:28:09,055 --> 00:28:10,585 can replace other software with it. 480 00:28:10,585 --> 00:28:10,765 Yeah. 481 00:28:10,765 --> 00:28:14,455 Like you need to think about data and if you're making that painful 482 00:28:14,455 --> 00:28:18,385 switch from one software stack, that might even like include all the way 483 00:28:18,385 --> 00:28:22,979 to the operating system, you probably like experience a lot of pain there. 484 00:28:23,249 --> 00:28:27,769 And ideally, you now choose a foundation that gives you more 485 00:28:27,769 --> 00:28:31,459 optionality in the future when you might want to do another switch. 486 00:28:31,729 --> 00:28:35,222 And this is where things like, data compatibility, et cetera, 487 00:28:35,282 --> 00:28:36,632 all like play into this. 488 00:28:37,267 --> 00:28:43,807 And where ideally, like legal frameworks set a good foundation and like encourage 489 00:28:43,837 --> 00:28:49,340 things like data compatibility, et cetera, but in a thoughtful way where it does 490 00:28:49,340 --> 00:28:55,390 not, introduce so much overhead and like bureaucracy that it grinds down like the 491 00:28:55,620 --> 00:28:57,640 velocity of like hurricane ship software. 492 00:28:57,640 --> 00:29:00,034 And I think this is like the best of all worlds that 493 00:29:00,034 --> 00:29:02,914 local-first wants to empower that. 494 00:29:03,060 --> 00:29:05,775 Where if you build with like those local-first data layers. 495 00:29:06,720 --> 00:29:11,010 The data layer takes care of that for you, and you can focus on the software. 496 00:29:11,190 --> 00:29:15,900 So I think this is all like a very rich area, to talk about and I think that 497 00:29:15,900 --> 00:29:20,910 what's makes it really interesting is to hear it through the lens of like one 498 00:29:20,910 --> 00:29:26,040 particular story, whether it is like a company having made that switch, whether 499 00:29:26,040 --> 00:29:30,888 you've built software and like you've made your software more compatible and 500 00:29:30,888 --> 00:29:36,348 interoperable in some way, or, yeah, there's like all of those stories that 501 00:29:36,528 --> 00:29:40,590 I've heard over the years and I think those deserve a broader, audience. 502 00:29:40,590 --> 00:29:44,377 And so if you have something to share in that regard, please definitely take 503 00:29:44,377 --> 00:29:46,550 the step and like, submit it to the CFP. 504 00:29:46,907 --> 00:29:47,477 That's a great point. 505 00:29:47,477 --> 00:29:51,947 It almost makes me wanna make data portability or something to that regard. 506 00:29:52,224 --> 00:29:53,634 a top level topic. 507 00:29:53,634 --> 00:29:56,004 But I think this comes back to the principle in local-first, 508 00:29:56,004 --> 00:29:57,204 which we call the Long Now. 509 00:29:57,729 --> 00:30:00,519 Which is, yeah, once you've been around for a while, you 510 00:30:00,519 --> 00:30:01,659 can see the gray in my beard. 511 00:30:01,659 --> 00:30:05,019 I've been working with computers for many decades now, and I've 512 00:30:05,019 --> 00:30:06,789 seen many products come and go. 513 00:30:06,789 --> 00:30:08,379 Software platforms come and go. 514 00:30:08,739 --> 00:30:11,919 But being able to preserve that data for a variety of reasons. 515 00:30:11,919 --> 00:30:13,089 'cause it's important for your company. 516 00:30:13,119 --> 00:30:14,409 'cause it's important for you personally. 517 00:30:14,409 --> 00:30:18,779 There's emotional meaning and saving your old family photos or the, master thesis 518 00:30:18,779 --> 00:30:20,369 he wrote in university or whatever. 519 00:30:20,769 --> 00:30:24,039 And then of course it is natural that especially the startup industry, 520 00:30:24,039 --> 00:30:26,769 which skews young and people are building products and moving fast 521 00:30:26,769 --> 00:30:30,399 and thinking about tomorrow and next quarter and not 10 years from now. 522 00:30:30,449 --> 00:30:31,729 That's, that's fine. 523 00:30:31,729 --> 00:30:34,479 and well and good, but maybe that is a role. 524 00:30:34,479 --> 00:30:37,735 Governments can play a little bit, but it's also a role that, yeah, maybe our 525 00:30:37,735 --> 00:30:41,065 community can play, which is encourage people to think a little bit about that 526 00:30:41,065 --> 00:30:46,102 long term, both as users, what's going to make my data be a little safer, give me 527 00:30:46,102 --> 00:30:47,962 that long term optionality, as you said. 528 00:30:48,194 --> 00:30:52,239 but also product makers to think in terms of like, okay, may maybe 529 00:30:52,239 --> 00:30:55,449 my product won't be hero, will have changed a lot in five or 10 years. 530 00:30:55,689 --> 00:30:59,485 Am I empowering my users, to take their data and go to where they need 531 00:30:59,485 --> 00:31:00,805 to go when the time comes for that? 532 00:31:01,604 --> 00:31:05,324 And I think it's just like a wonderful way to think about the world a bit more, like 533 00:31:05,324 --> 00:31:10,304 positive sum, where I think we've, like in some areas of software, we've already 534 00:31:10,484 --> 00:31:14,650 figured us out, like in a pretty good way when we have like languages and like 535 00:31:14,650 --> 00:31:16,720 ideally like typed languages, et cetera. 536 00:31:16,720 --> 00:31:19,990 Like you have an API that you can compatible with and then 537 00:31:19,990 --> 00:31:21,550 you can like reuse modules. 538 00:31:21,730 --> 00:31:24,850 We even got that all the way to like UI, et cetera. 539 00:31:25,180 --> 00:31:27,850 But for data, this is where we're struggling still the most. 540 00:31:27,850 --> 00:31:32,060 And I think this is like one of the core themes around local-first where 541 00:31:32,060 --> 00:31:35,780 there has been tremendous progress to, to make, progress in that regard. 542 00:31:36,220 --> 00:31:40,630 And in that theme of like user agency, et cetera, personal software, 543 00:31:40,840 --> 00:31:44,464 I think this will play an even bigger role, where we are gonna 544 00:31:44,464 --> 00:31:46,384 have like more software than ever. 545 00:31:46,414 --> 00:31:48,964 Whether it's like we've built this ourselves, whether it's agents 546 00:31:48,964 --> 00:31:52,654 have built it, and then those things can, can talk to each other. 547 00:31:53,074 --> 00:31:56,764 But I think it's also like changing in an interesting way where the 548 00:31:56,764 --> 00:32:01,354 contract between that does not have to be like absolutely perfect anymore. 549 00:32:01,684 --> 00:32:07,000 But like now AI can also like, shoulder a bit of that burden of like talking 550 00:32:07,000 --> 00:32:10,390 to each other, where now we can like get away with having fuzzy software. 551 00:32:11,195 --> 00:32:16,620 Where it has like CSV over here and like a JSON API over there, and the 552 00:32:16,800 --> 00:32:21,440 agent can like very easily either ad hoc one-off, like convert something or 553 00:32:21,440 --> 00:32:23,330 create a script that talks to each other. 554 00:32:23,540 --> 00:32:28,070 So I think we're like, we're getting closer and closer having like some 555 00:32:28,070 --> 00:32:33,140 bridges and ideally over time we'll get to like really having very reliable 556 00:32:33,140 --> 00:32:38,826 bridges that just, in our minds when we hear a story like over there and 557 00:32:38,826 --> 00:32:42,516 like we write it down over there, like we are a bridge and just works. 558 00:32:42,516 --> 00:32:45,490 And ideally software can be like that in the future. 559 00:32:45,983 --> 00:32:49,703 I like that a lot and I like that point that software is becoming 560 00:32:49,733 --> 00:32:53,693 cheaper to make, so therefore it will be more ephemeral, more throwaway. 561 00:32:53,693 --> 00:32:56,573 You'll be more likely to make something that's purpose-built for 562 00:32:56,578 --> 00:32:59,813 a, a shorter period of time, whether it's an internal tool at a company 563 00:32:59,813 --> 00:33:01,103 or a personal app or something. 564 00:33:01,463 --> 00:33:02,453 Something like that. 565 00:33:02,753 --> 00:33:05,963 And in the past we've really relied, or to date you could say we've 566 00:33:05,963 --> 00:33:10,673 really relied on, my Gmail is still readable 20 years ago because Gmail 567 00:33:10,673 --> 00:33:12,413 still exists and is well maintained. 568 00:33:12,736 --> 00:33:16,546 but instead, maybe if we have a system that is more based on, well, I can 569 00:33:16,546 --> 00:33:21,106 use an email tool that's a little bit bespoke and gets stops being relevant 570 00:33:21,106 --> 00:33:24,196 or stops being maintained two years from now, and it's no big deal because 571 00:33:24,196 --> 00:33:28,126 getting my email out or whatever data came from that tool and into the next 572 00:33:28,126 --> 00:33:32,026 thing that I'm going to use is pretty straightforward to do from through some 573 00:33:32,026 --> 00:33:34,306 combination of export data portability. 574 00:33:34,306 --> 00:33:38,896 But also, as you said, that importing and ingesting data and making data 575 00:33:38,896 --> 00:33:41,536 coherent with the help of AI is easier. 576 00:33:41,536 --> 00:33:43,396 So I like that idea of a lot. 577 00:33:43,396 --> 00:33:47,956 This, the software is kind of ephemeral and fits the need you have in the exact 578 00:33:47,956 --> 00:33:49,756 moment, and then you can move on from it. 579 00:33:49,786 --> 00:33:52,526 The data is something that has a longer lifespan. 580 00:33:52,526 --> 00:33:55,406 It can move with you from, product to product. 581 00:33:56,306 --> 00:33:56,936 Exactly. 582 00:33:57,476 --> 00:33:57,896 Great. 583 00:33:57,896 --> 00:34:01,505 So this was local-first maturity, and we've already teased a little 584 00:34:01,505 --> 00:34:05,865 bit of like some adjacent topics, but the, the other, big pillars here, 585 00:34:05,931 --> 00:34:10,775 as Elene put it, is like the new territories and the larger ecosystem. 586 00:34:10,775 --> 00:34:14,975 So maybe if we spend a bit of time on the new territories where we have like, 587 00:34:14,975 --> 00:34:20,281 designing for agency, then around like AI, local and open weight models, and 588 00:34:20,281 --> 00:34:23,191 then also AI assistance for local-first. 589 00:34:23,431 --> 00:34:26,101 Maybe starting with the first one, design for agency. 590 00:34:26,338 --> 00:34:27,568 What should people think about here? 591 00:34:27,748 --> 00:34:28,108 Yeah. 592 00:34:28,108 --> 00:34:34,888 For me, this would come to the topic of, I like computers as a tool for thought, 593 00:34:34,888 --> 00:34:39,208 a way to augment, thinking, a way to make humans more creative, more capable, 594 00:34:39,208 --> 00:34:41,098 able to accomplish our goals better. 595 00:34:41,578 --> 00:34:45,538 I sometimes think, especially as consumer products have have come to 596 00:34:45,538 --> 00:34:49,828 dominate, for example, just sort of usage and revenue in the computing world. 597 00:34:49,828 --> 00:34:52,408 There ends up being this weird race to the bottom of, hey, 598 00:34:52,678 --> 00:34:53,938 the computer will think for me. 599 00:34:54,188 --> 00:34:55,208 that's what we like about AI. 600 00:34:55,208 --> 00:34:55,448 Good. 601 00:34:55,448 --> 00:34:56,318 I don't need to think anymore. 602 00:34:56,318 --> 00:34:57,368 The computer can do it for me. 603 00:34:57,418 --> 00:35:01,208 And that to me is a very sad perspective. 604 00:35:01,328 --> 00:35:05,121 And rather I would think, wanna think that, AI or other capabilities 605 00:35:05,178 --> 00:35:08,884 that we continue to develop as, we continue to explore the computing 606 00:35:08,884 --> 00:35:14,104 space are things that help us do more and better the things we enjoy 607 00:35:14,104 --> 00:35:15,844 doing, being creative and thinking. 608 00:35:16,181 --> 00:35:20,471 and so I think the design for agency topic is the idea of, okay. 609 00:35:20,831 --> 00:35:24,311 So if you're doing something with language models, with computer vision, 610 00:35:24,311 --> 00:35:30,458 with agents, whatever it is, what can be part of your design, in the sense of how 611 00:35:30,458 --> 00:35:36,068 your product works, that really engages and creates more agency and more deeper 612 00:35:36,068 --> 00:35:37,963 understanding from your users, right? 613 00:35:38,143 --> 00:35:41,623 The negative scenario or the people often quote is that Star Trek 614 00:35:41,623 --> 00:35:43,033 next generation episode, right? 615 00:35:43,033 --> 00:35:46,813 Where people have this, these omniscient computers that do everything and 616 00:35:46,813 --> 00:35:50,113 then they lose all the capability to understand not only just how to 617 00:35:50,113 --> 00:35:52,843 fix the machines, but just really how the world around them works. 618 00:35:53,113 --> 00:35:55,873 And then when those computers start to fail, they're suddenly 619 00:35:55,993 --> 00:35:57,103 like helpless children. 620 00:35:57,463 --> 00:36:02,233 And I think for me, good computing tools of all sorts, and that includes everything 621 00:36:02,233 --> 00:36:06,103 to do with AI and language models deepens our understanding of the world, deepens 622 00:36:06,103 --> 00:36:10,303 our understanding of the technology by freeing us from some of the fussy details 623 00:36:10,303 --> 00:36:15,593 of where exactly does the semicolon go, and allowing me to spend more time on the 624 00:36:15,883 --> 00:36:20,503 mental models and the frameworks and the principles and how things fit together and 625 00:36:20,503 --> 00:36:23,203 how we use those to accomplish our goals. 626 00:36:23,823 --> 00:36:24,183 Yeah. 627 00:36:24,183 --> 00:36:29,133 And I think this is, I mean, for, for me, AI has been like one of the biggest 628 00:36:29,193 --> 00:36:32,613 unlocks to like just learn new things. 629 00:36:32,613 --> 00:36:36,513 I've gotten into so many new areas, like I've gotten deeper on like 630 00:36:36,513 --> 00:36:38,073 some hardware related things. 631 00:36:38,073 --> 00:36:41,523 I've gotten deeper on understanding how wifi works, 632 00:36:41,616 --> 00:36:43,956 like band channels, et et cetera. 633 00:36:43,986 --> 00:36:48,026 Like gave me a deeper and a broader understanding of the world, like 634 00:36:48,026 --> 00:36:52,460 I have been on an accelerated pace of, learning and understanding. 635 00:36:52,760 --> 00:36:56,510 and I think that's like the positive sum, version of that or like the 636 00:36:56,520 --> 00:37:02,676 flip side or the other, scenario, of an like the utopia maybe of that 637 00:37:02,676 --> 00:37:04,963 Star Trek, anecdote you shared. 638 00:37:05,273 --> 00:37:10,786 So, one other much more concrete topic that I think has been, a very 639 00:37:10,846 --> 00:37:15,720 new one that's come, about over the last few years as, like off the 640 00:37:15,720 --> 00:37:18,060 shelf models became available at all. 641 00:37:18,690 --> 00:37:22,906 And now we also have like, open weight models and they're even small enough 642 00:37:23,236 --> 00:37:29,483 that you can run them on your own device, whether it's on a very beefy, desktop 643 00:37:29,483 --> 00:37:34,624 grade machine or whether it's like, scaled down models that might even work in the 644 00:37:34,624 --> 00:37:39,664 browser, that might even work on like a phone or other much lower end devices. 645 00:37:40,054 --> 00:37:47,134 I think that opens a whole new category of software and use cases, et cetera, whether 646 00:37:47,134 --> 00:37:52,321 it is, like that you built a little system that relies on like detecting 647 00:37:52,321 --> 00:37:54,961 certain shapes to do a certain thing. 648 00:37:54,961 --> 00:37:58,501 Maybe you're building, like maybe you have chicken at home and you wanna like 649 00:37:58,501 --> 00:38:03,691 automate a little gate that like, opens when a chicken walks through or whether 650 00:38:03,691 --> 00:38:05,431 you have like some other use cases. 651 00:38:05,431 --> 00:38:10,298 But this is you ideally wanna do that like reliably, locally, efficiently, et cetera. 652 00:38:10,298 --> 00:38:14,471 And there's so many, like we can, we can just like dream so much bigger now 653 00:38:14,471 --> 00:38:20,141 of like what we can build in a way that does not rely on like Anthropic currently 654 00:38:20,141 --> 00:38:25,031 not having an outage and us like paying whatever, a lot of money each time. 655 00:38:25,031 --> 00:38:29,328 We are like using a bit of like API quota and, I think this is 656 00:38:29,328 --> 00:38:31,824 gonna be like right now, this is. 657 00:38:31,924 --> 00:38:35,944 Probably similar to the early days of like computers being room 658 00:38:35,944 --> 00:38:38,308 sized, and not very capable yet. 659 00:38:38,338 --> 00:38:42,528 And I think that's just gonna, like at some point our little like devices like 660 00:38:42,528 --> 00:38:48,424 this, will be able to do like really, really impressive, inference capabilities. 661 00:38:48,864 --> 00:38:52,748 And I think this is like, now is the time to start thinking about 662 00:38:52,748 --> 00:38:56,558 that and start already embracing that in all sorts of different 663 00:38:56,558 --> 00:38:58,748 scenarios, form factors, et cetera. 664 00:38:59,108 --> 00:39:02,618 So I think this area could be about like if you're building 665 00:39:02,948 --> 00:39:06,788 infrastructure, building tool, like maybe you're the company behind some 666 00:39:06,788 --> 00:39:10,838 of those open models and you wanna show people what's already possible. 667 00:39:10,838 --> 00:39:11,918 I think that's really great. 668 00:39:12,241 --> 00:39:16,201 if you are building applications, systems, et cetera, that already 669 00:39:16,201 --> 00:39:18,788 use, open models, local models. 670 00:39:19,458 --> 00:39:21,528 that is like a story we'd love to hear. 671 00:39:21,738 --> 00:39:26,908 We had last year, Thomas from Google already share, some earlier versions 672 00:39:26,938 --> 00:39:32,244 of like how you can run, Gemini Nano in Chrome and showed some really cool demos. 673 00:39:32,574 --> 00:39:37,688 And all of this has come a really long way and, I think across all sorts of 674 00:39:37,688 --> 00:39:40,058 different devices, platforms, et cetera. 675 00:39:40,328 --> 00:39:44,708 So if you're doing cool things in that regard, please definitely apply to speak. 676 00:39:45,318 --> 00:39:49,188 Yeah, I think so much attention is on the frontier models and the 677 00:39:49,382 --> 00:39:53,012 sort of one giant chat bot that can do everything, multimodal models, 678 00:39:53,022 --> 00:39:54,872 ChatGPT, and Claude and so forth. 679 00:39:55,311 --> 00:39:59,929 And those are of course impressive and very useful and there's a lot to that. 680 00:40:00,291 --> 00:40:02,797 But I really have, opened up to the value of things like 681 00:40:02,797 --> 00:40:04,387 lightweight classifier models. 682 00:40:05,197 --> 00:40:08,694 Yeah, embeddings, including, there's plenty of text embedding, models that 683 00:40:08,694 --> 00:40:12,084 you can just run in a Python server with PyTorch that are great, really 684 00:40:12,204 --> 00:40:16,074 almost as good as basically as good as what you can get in the big providers. 685 00:40:16,584 --> 00:40:20,274 Things like computer vision and OCR have actually been good for a pretty long time. 686 00:40:20,514 --> 00:40:23,664 I think in some ways, you know, maybe myself, I almost feel a little, 687 00:40:23,914 --> 00:40:28,084 surprised, you know, something about the AI hype, you know, around chatbots 688 00:40:28,084 --> 00:40:31,534 and large language models then caused me to take a closer look at a lot of 689 00:40:31,534 --> 00:40:34,894 these other ML technologies that have been getting good for a long time. 690 00:40:35,194 --> 00:40:38,134 A lot of the text embedding stuff and semantic search and things like 691 00:40:38,134 --> 00:40:41,884 that have been, you know, really had their breakthroughs 10 years ago. 692 00:40:42,094 --> 00:40:44,734 Computer vision has also been very good for a long time. 693 00:40:44,734 --> 00:40:47,344 We know that because captures keep getting harder and harder. 694 00:40:47,684 --> 00:40:51,254 And you can do a lot with these, in those very specialized cases, like what you 695 00:40:51,254 --> 00:40:54,964 were describing with, yeah, whether it's a robot in your house or, something like 696 00:40:54,964 --> 00:40:58,984 that, but also software where you can have the model run on each keystroke, right? 697 00:40:58,984 --> 00:41:00,514 You can do all kinds of things. 698 00:41:00,989 --> 00:41:04,139 They're very different versus needing to wait several seconds for 699 00:41:04,289 --> 00:41:07,642 a heavy inference to come back from the big language model provider. 700 00:41:07,732 --> 00:41:11,362 And just broadly, the story of computing is one of taking these 701 00:41:11,392 --> 00:41:15,532 computing things that are originally are heavy and slow and as you said, 702 00:41:15,532 --> 00:41:19,792 require a room size computer and making them smaller and cheaper and faster. 703 00:41:19,797 --> 00:41:22,312 And that just opens up all kinds of new use cases. 704 00:41:22,522 --> 00:41:24,892 And there's really great open models out there from Quinn 705 00:41:24,892 --> 00:41:26,692 to Llama to many, many others. 706 00:41:26,992 --> 00:41:30,532 And yeah, I think if you're working either on open weight models, or open 707 00:41:30,532 --> 00:41:34,432 models of any kind or the libraries behind it, like Transformer.js or 708 00:41:34,432 --> 00:41:36,712 PyTorch or something like that, we'd love to hear a talk from you. 709 00:41:37,585 --> 00:41:38,125 Awesome. 710 00:41:38,185 --> 00:41:45,165 So the next topic in the new territories, area is AI assistance for local-first. 711 00:41:45,555 --> 00:41:50,205 This is, I guess more around like the application of those new materials 712 00:41:50,235 --> 00:41:52,005 to those new building blocks we got. 713 00:41:52,395 --> 00:41:56,595 And I think that can be really, like wide ranging, can be going in 714 00:41:56,595 --> 00:41:58,215 all sorts of different directions. 715 00:41:58,215 --> 00:42:01,515 Like, and I think the exciting part is actually the stuff that we can't 716 00:42:01,515 --> 00:42:05,805 even, like, think about right now, but that maybe some of you are, 717 00:42:05,842 --> 00:42:09,892 actively exploring and have built like surprising things on top of this. 718 00:42:10,232 --> 00:42:13,802 But just some things that came, to mind for me as I've been like, 719 00:42:14,018 --> 00:42:19,242 experimenting myself or thinking this through, like just to, generate ideas. 720 00:42:19,332 --> 00:42:24,762 How can AI address like some of those open problems in the local-first 721 00:42:24,762 --> 00:42:29,082 space or like some challenges that are so like kind of hard to crack 722 00:42:29,382 --> 00:42:30,942 and if we're allowing ourselves. 723 00:42:31,712 --> 00:42:36,602 To go beyond the realm of like deterministic, traditional software. 724 00:42:36,602 --> 00:42:39,349 And like, we're using those fuzzy materials a bit more. 725 00:42:39,349 --> 00:42:42,655 And like we're thinking of like, hey, if we're actually, embedding 726 00:42:42,655 --> 00:42:47,169 them into our local software stack, what becomes possible through this? 727 00:42:47,169 --> 00:42:52,142 So like, just one way, for example is let's say you're building a power tool and 728 00:42:52,142 --> 00:42:57,542 like you have a very, very sophisticated set of like system settings, et cetera. 729 00:42:57,542 --> 00:43:01,322 Like for me, for example, the Apple system settings come to mind and 730 00:43:01,322 --> 00:43:04,612 they're really like, I don't think they're the pinnacle of modern software. 731 00:43:05,032 --> 00:43:07,255 And, I always need to use search. 732 00:43:07,255 --> 00:43:11,175 And like five different things come up and then it's like the 15th one 733 00:43:11,175 --> 00:43:12,835 that I was actually looking for. 734 00:43:13,165 --> 00:43:15,625 And like, it's just like very painful. 735 00:43:16,405 --> 00:43:22,885 Now where we are so spoiled by like using things like ChatGPT et cetera, 736 00:43:22,885 --> 00:43:27,512 like it really feels like worlds a part of like how good something could be. 737 00:43:27,512 --> 00:43:34,612 So why not use something like a, ChatGPT chat interface to interact with your 738 00:43:34,612 --> 00:43:39,412 own app that like, you can, instead of have to click the clock settings, 739 00:43:39,615 --> 00:43:44,950 like why not allow your entire app to begin configurable through AI models. 740 00:43:45,340 --> 00:43:48,910 And so to make that actually happen now you need to bring in like that 741 00:43:49,120 --> 00:43:52,270 fuzzy LM piece into your software. 742 00:43:52,270 --> 00:43:56,800 But like exploring that, like figuring out best practices, figuring out 743 00:43:56,800 --> 00:44:01,920 like how can you actually, constrain that, that you don't, give the, users 744 00:44:01,920 --> 00:44:06,537 so much, power that they can like accidentally tear apart the app and 745 00:44:06,537 --> 00:44:08,247 it's no longer functional, et cetera. 746 00:44:08,247 --> 00:44:10,757 I think there's a lot to be explored there. 747 00:44:11,137 --> 00:44:12,737 So just as one area. 748 00:44:13,057 --> 00:44:18,667 Another one is to give even more power to the user or build systems in a way 749 00:44:18,667 --> 00:44:20,497 that wasn't really feasible before. 750 00:44:20,497 --> 00:44:25,027 We talked about the data interoperability before data exports, et cetera. 751 00:44:25,447 --> 00:44:29,017 Now we can actually build like fuzzy bridges between different 752 00:44:29,017 --> 00:44:32,797 software that were never meant to be talking to each other. 753 00:44:32,867 --> 00:44:36,437 Whether it's like in real time or whether it's just like making 754 00:44:36,437 --> 00:44:41,270 sort of, export interfaces a little bit more ergonomic to use. 755 00:44:41,360 --> 00:44:46,744 And now you could imagine whatever, like using your, local supermarket 756 00:44:46,834 --> 00:44:51,440 that might have like a certain, online shopping, like not even API, 757 00:44:51,440 --> 00:44:56,787 but just like a maybe connecting that with your own home build to-do app. 758 00:44:57,147 --> 00:45:02,030 And, there is not really like the end all be all integration API, yet 759 00:45:02,090 --> 00:45:03,860 you can make those things happen now. 760 00:45:03,860 --> 00:45:08,870 So this is just like one of many, many ideas where hopefully you all have 761 00:45:08,870 --> 00:45:12,803 been exploring some interesting use cases that we'd love to hear about. 762 00:45:12,803 --> 00:45:16,774 So this is what comes to mind for me about AI assistance for local-first. 763 00:45:16,774 --> 00:45:21,019 Yeah, using these new technologies to make the world more local-first, 764 00:45:21,019 --> 00:45:23,239 especially maybe parts that have been resistant to that. 765 00:45:23,239 --> 00:45:27,229 I think breaking down data silos is one of the biggest ones that's been such a long 766 00:45:27,229 --> 00:45:31,789 running topic at you could switch with projects like Cambria, but almost always 767 00:45:31,789 --> 00:45:36,319 it involves kind of a convince everyone to build their software on a completely 768 00:45:36,319 --> 00:45:40,069 different stack, which sort of Yeah. 769 00:45:40,069 --> 00:45:41,899 Is very much an uphill battle. 770 00:45:41,899 --> 00:45:42,229 Right. 771 00:45:42,531 --> 00:45:45,405 And as you point out, there's a great agency in our personal lives. 772 00:45:45,435 --> 00:45:49,440 I've experienced this already with, yeah, being able to pull out, I don't know. 773 00:45:49,440 --> 00:45:52,740 You know, there's some website that has a list of things and I want 774 00:45:52,740 --> 00:45:55,650 to put that into a spreadsheet, but it doesn't offer a CSV export. 775 00:45:56,040 --> 00:45:56,550 Okay. 776 00:45:56,700 --> 00:45:59,910 I basically could do that with a screenshot or a PDF and let the language 777 00:45:59,910 --> 00:46:03,810 model do the hard work of, you know, turning it into something coherent. 778 00:46:04,037 --> 00:46:06,047 Similar idea there is just downloading. 779 00:46:06,047 --> 00:46:10,017 I like to download my transactions from my bank account on a quarterly basis, so I 780 00:46:10,027 --> 00:46:14,837 have a record of that that's separate from whatever clunky interface the bank has. 781 00:46:15,197 --> 00:46:19,277 But inevitably, if you have a couple of accounts and then they're all in slightly 782 00:46:19,277 --> 00:46:22,577 different formats and trying to like bring them all together into any kind 783 00:46:22,577 --> 00:46:24,827 of unified ledger is nearly impossible. 784 00:46:24,827 --> 00:46:28,007 It's just not worth the bother for an individual or whatever. 785 00:46:28,007 --> 00:46:32,267 But actually now with language model, it's pretty easy to do an ad hoc thing where I 786 00:46:32,267 --> 00:46:36,917 can go do the download effort myself once a quarter, but then pull that all into a 787 00:46:36,917 --> 00:46:39,137 spreadsheet with a normalized data set. 788 00:46:39,137 --> 00:46:40,517 That feels very empowering for me. 789 00:46:40,517 --> 00:46:44,627 It's like, okay, now I have all my transactions in a place that I control 790 00:46:44,837 --> 00:46:48,797 in a format that I want that isn't dependent on, you know, my bank letting 791 00:46:48,797 --> 00:46:52,787 me, you know, I go to login to download something and it tells me, oh, sorry, 792 00:46:52,787 --> 00:46:57,467 we only support transactions, you know, back six months or whatever thing, right? 793 00:46:57,467 --> 00:47:02,207 Very disempowering user experience that instead it's something that data 794 00:47:02,207 --> 00:47:06,077 that feels like it belongs to me, my banking and transaction data, and I 795 00:47:06,077 --> 00:47:09,707 have it in a place and a format that I can control for the long term. 796 00:47:10,512 --> 00:47:11,082 Exactly. 797 00:47:11,235 --> 00:47:16,407 it's just so cool to see like how far people take, like how much they embrace 798 00:47:16,407 --> 00:47:20,727 the new forms of like, agencies that they, I guess they, they've been always like 799 00:47:20,727 --> 00:47:25,727 longing for, but now, like it becomes feasible to, like really go for it. 800 00:47:25,727 --> 00:47:29,537 And I think it's so, also so liberating that you can build software for 801 00:47:29,537 --> 00:47:34,337 yourself unburdened by like making it enterprise grade, et cetera. 802 00:47:34,667 --> 00:47:38,027 And so you can like, you know, I need to hold it this way and it's 803 00:47:38,027 --> 00:47:39,734 gonna work, otherwise it falls apart. 804 00:47:39,734 --> 00:47:40,394 But that's fine. 805 00:47:40,874 --> 00:47:45,067 And I think that's like the, beauty of like building tools for yourself 806 00:47:45,127 --> 00:47:49,957 and like local-first gives us the glue that they actually can bear like 807 00:47:49,957 --> 00:47:52,344 real weight and like work together. 808 00:47:52,854 --> 00:47:58,084 So that is the new territories, which I think is very related to the larger 809 00:47:58,134 --> 00:48:02,594 ecosystem where we have, a set of like other, very interesting topics. 810 00:48:02,594 --> 00:48:06,942 So like some of them we've teased before, but just to run through like on a high 811 00:48:06,942 --> 00:48:10,182 level and then we can go into them like again, we have like self-sovereign 812 00:48:10,212 --> 00:48:14,892 identity, we have social media and data ownership, encrypted messaging, 813 00:48:15,192 --> 00:48:19,912 game development with local-first and malleable software in the age of LLMs. 814 00:48:19,962 --> 00:48:23,759 So maybe starting with the self-sovereign identity, what 815 00:48:23,819 --> 00:48:25,319 should people think about here? 816 00:48:25,559 --> 00:48:30,249 Which sort of stories might we be interested in having at the conference? 817 00:48:30,959 --> 00:48:31,109 Yeah. 818 00:48:31,109 --> 00:48:34,808 Well in general, this category of the larger ecosystem is one I'm excited 819 00:48:34,808 --> 00:48:37,628 about because I think there are so many adjacent communities that are 820 00:48:37,628 --> 00:48:41,708 interested in, yeah, for example, things like freedom, you know, the open source 821 00:48:41,708 --> 00:48:43,448 community, obviously the Linux world. 822 00:48:44,168 --> 00:48:48,868 The Open Web and open standards, as well as things like the end-to-end encryption, 823 00:48:48,898 --> 00:48:51,205 kind of privacy, nerds out there. 824 00:48:51,205 --> 00:48:53,815 You know, there's, that's not exactly what local-first is 825 00:48:53,815 --> 00:48:55,255 about, but it's very adjacent. 826 00:48:55,255 --> 00:48:58,135 There's a lot of shared technologies and things, and this is an invitation 827 00:48:58,135 --> 00:48:59,905 for folks in those communities. 828 00:48:59,905 --> 00:49:04,375 And I'm even sending out some emails and so forth to folks to try to invite them in 829 00:49:04,375 --> 00:49:06,355 to say, look, let's see where we overlap. 830 00:49:06,355 --> 00:49:08,522 Let's, find the edges there. 831 00:49:08,868 --> 00:49:09,005 yeah. 832 00:49:09,005 --> 00:49:11,765 The identity topic though, this is a big one for me. 833 00:49:11,765 --> 00:49:14,135 It's something I've been fussing about my whole career. 834 00:49:14,135 --> 00:49:17,498 I have whole, article about it in my making computers better 835 00:49:17,738 --> 00:49:19,195 essay, from a few years ago. 836 00:49:19,195 --> 00:49:22,165 But it basically boils down to, I think the way we identify ourselves to 837 00:49:22,165 --> 00:49:26,485 computers is kind of an embarrassment, usernames and passwords, identity theft. 838 00:49:27,085 --> 00:49:30,625 Now you've got, I don't know, magic login links and login with Google 839 00:49:30,725 --> 00:49:35,455 passkeys and whatever, and it's just endlessly confusing for most 840 00:49:35,815 --> 00:49:38,185 users and even for power users. 841 00:49:38,185 --> 00:49:41,365 Yeah, the amount of time in my day, I end up spite typing in two factor 842 00:49:41,365 --> 00:49:44,995 codes and digging magic links out of emails and just trying to convince 843 00:49:44,995 --> 00:49:46,795 the computer that I am who I say I am. 844 00:49:47,185 --> 00:49:48,295 And I know why it's that way. 845 00:49:48,295 --> 00:49:51,445 Of course, as a professional that's worked in this field and particularly 846 00:49:51,445 --> 00:49:54,722 in the security, space, on and off, over the course of my career. 847 00:49:54,722 --> 00:49:56,841 But it really is a shame to me. 848 00:49:57,201 --> 00:50:00,411 And then in the meantime, you have this more and more control and 849 00:50:00,411 --> 00:50:04,341 more and more ownership over our identities that go to the companies. 850 00:50:04,341 --> 00:50:07,341 And basically it's been a kind of a good deal to say like, look, I'll 851 00:50:07,341 --> 00:50:10,791 just let Google manage my identity and then I can click the login with Google 852 00:50:10,791 --> 00:50:14,565 button everywhere, because it is such a hard problem and I can just kind 853 00:50:14,565 --> 00:50:17,115 of outsource that to them, sort of. 854 00:50:17,115 --> 00:50:19,815 But there, it's kind of like a little bit of a deal with the devil there 855 00:50:19,905 --> 00:50:21,225 comes with a lot of trade offs. 856 00:50:21,573 --> 00:50:25,400 and local-first is a chance to, rethink that, reinvent that. 857 00:50:25,434 --> 00:50:27,546 and of course there are also a lot of adjacent communities. 858 00:50:27,546 --> 00:50:29,286 Things like Open ID for example. 859 00:50:29,286 --> 00:50:33,276 We'll talk a little bit more about, app proto and ActivityPub in a moment. 860 00:50:33,276 --> 00:50:36,656 But they have their own concepts of identity and how you give people ownership 861 00:50:36,656 --> 00:50:40,396 over your online handle, in a way that's sort of secure and safe, but doesn't 862 00:50:40,666 --> 00:50:44,356 put the user in the position of needing to manage a bunch of cryptography keys, 863 00:50:44,356 --> 00:50:45,986 which they're inevitably going to lose. 864 00:50:46,216 --> 00:50:49,906 so I think anyone working on anything in this space, whether it feels 865 00:50:50,046 --> 00:50:52,636 local-firsty or not, if you're working with passkeys, if you're working 866 00:50:52,636 --> 00:50:56,083 with, any kind of identity system, yeah, we'd love to hear from you. 867 00:50:56,993 --> 00:50:59,403 Yeah, I think this is a highly. 868 00:50:59,690 --> 00:51:00,863 unsolved area. 869 00:51:00,893 --> 00:51:02,663 I wouldn't say under explored. 870 00:51:02,663 --> 00:51:04,736 I think, people are really exploring it. 871 00:51:04,736 --> 00:51:06,656 It's a hard problem in really interesting ways. 872 00:51:06,956 --> 00:51:11,396 It is a hard problem and I think we've, we've also come a long way already. 873 00:51:11,396 --> 00:51:14,676 Like for example, like if you're depending on, which pact with the 874 00:51:14,676 --> 00:51:18,726 devil you've made, you can already have a slightly more convenient life. 875 00:51:18,756 --> 00:51:24,320 So me being, in pact with the Apple ecosystem, I am very thankful for things 876 00:51:24,320 --> 00:51:29,090 like Touch Id, et cetera, that are then like nicely working together with things 877 00:51:29,090 --> 00:51:31,381 like 1Password, Passkeys, et cetera. 878 00:51:31,513 --> 00:51:36,678 and like being able to understand how those things work and like thinking 879 00:51:36,678 --> 00:51:38,328 through the implications, et cetera. 880 00:51:38,328 --> 00:51:41,825 Like let's me be for, for this point in time, in a relative sweet 881 00:51:41,825 --> 00:51:46,545 spot between, like living a secure life, but also having convenience. 882 00:51:46,635 --> 00:51:50,325 But there's still like great levels of like lock in, et cetera. 883 00:51:50,355 --> 00:51:55,065 And I think this is also like all of that aside, like we're now also, 884 00:51:55,131 --> 00:51:58,821 having this like phase shift in regards to software where we are no 885 00:51:58,821 --> 00:52:03,745 longer the only ones who are doing things on our behalf, with computers. 886 00:52:03,745 --> 00:52:08,551 Like if we're, now whatever, like maybe ordering, like some food or 887 00:52:08,577 --> 00:52:13,603 we're doing some other things, maybe we're, like scheduling something on, 888 00:52:13,723 --> 00:52:15,913 the calendar or whatever it might be. 889 00:52:16,213 --> 00:52:19,903 It's no longer just us, but there's like another thing that needs 890 00:52:20,113 --> 00:52:22,453 identity and even more fine grains. 891 00:52:22,453 --> 00:52:26,550 And I think one aspect here as well is like if you're going from beyond 892 00:52:26,550 --> 00:52:32,596 authentication to authorization, so far we have like, all off scopes, et cetera, 893 00:52:32,596 --> 00:52:34,936 but like it's very all or nothing. 894 00:52:34,936 --> 00:52:38,226 And it's all like, okay, you, now get access forever. 895 00:52:38,466 --> 00:52:43,426 But maybe for like those agentic use cases, you might want to think about it 896 00:52:43,436 --> 00:52:48,156 like, okay, you get to do this like one time, or you get to do this like for five 897 00:52:48,156 --> 00:52:52,706 minutes, or you get to do this like only for this email address or for this scope. 898 00:52:52,736 --> 00:52:57,776 So I think we want to make this like much more fine-grained, but in a way 899 00:52:57,836 --> 00:53:02,546 where you like, it's, it's frankly like going through like the permission 900 00:53:02,546 --> 00:53:06,716 granting on like a Google OAuth flow is already overwhelming for most people. 901 00:53:07,016 --> 00:53:09,686 And now making this even more fine granular 902 00:53:09,776 --> 00:53:09,956 Yeah, 903 00:53:09,986 --> 00:53:11,156 I think is like really 904 00:53:11,246 --> 00:53:11,346 tough challenge. 905 00:53:11,346 --> 00:53:11,531 Thinking of a 906 00:53:11,696 --> 00:53:12,056 giant 907 00:53:12,056 --> 00:53:12,441 array of check 908 00:53:12,446 --> 00:53:12,656 check 909 00:53:12,656 --> 00:53:16,270 boxes on the GitHub, token generation and there's a reason for it. 910 00:53:16,270 --> 00:53:22,233 But then yes, again, that trade off of, convenience and security is ever present. 911 00:53:22,633 --> 00:53:26,173 I think it's that intersection of like user experience 912 00:53:26,173 --> 00:53:28,810 design and also cryptography. 913 00:53:28,810 --> 00:53:30,580 Like how do you make it actually secure? 914 00:53:30,610 --> 00:53:34,886 and I think that to, come together in the best of both worlds. 915 00:53:35,216 --> 00:53:39,326 I think that is like a area that's like highly deserving of like 916 00:53:39,326 --> 00:53:43,196 more resources, more investment, more smart people working on it. 917 00:53:43,496 --> 00:53:46,166 So if you're working on that, we'd love to hear from you. 918 00:53:46,560 --> 00:53:50,040 ideally where you cover quite a bit of that intersection. 919 00:53:50,280 --> 00:53:53,390 I think this is probably not the best, conference to share 920 00:53:53,390 --> 00:53:56,540 like your new breakthrough of like some cryptographic nuance. 921 00:53:57,273 --> 00:54:01,703 But I think where we can like bring those two things together where you can also 922 00:54:01,703 --> 00:54:06,063 with the focus of like, or one aspect can be the solve sovereign, aspect. 923 00:54:06,123 --> 00:54:09,813 I think it doesn't have to be a prerequisite, but I think if 924 00:54:09,813 --> 00:54:12,063 something points in that direction. 925 00:54:12,333 --> 00:54:17,953 And also I think another, tricky part here is like that we need to kind of 926 00:54:18,043 --> 00:54:24,291 create a bridge from like today, where people are like, defacto using things 927 00:54:24,291 --> 00:54:29,138 like login with Google, et cetera, because it strikes A, acceptable balance 928 00:54:29,138 --> 00:54:30,878 of like convenience and security. 929 00:54:31,268 --> 00:54:36,768 And if there's like a much to esoteric, version out there that doesn't have a 930 00:54:36,768 --> 00:54:40,265 credible path to becoming mainstream, I think that also is something that 931 00:54:40,265 --> 00:54:42,335 needs to be taken into consideration. 932 00:54:42,335 --> 00:54:46,245 So this is a very nuanced topic and I think as one of 933 00:54:46,245 --> 00:54:48,335 the harder nuts to crack here. 934 00:54:48,955 --> 00:54:51,415 but if you're working on that, we'd love to hear from you. 935 00:54:51,835 --> 00:54:56,405 And I think with that it is like a natural transition to, things like 936 00:54:56,675 --> 00:55:01,735 BlueSky ATproto where we talk about, social media and data ownership. 937 00:55:01,745 --> 00:55:03,675 So you've already mentioned initially. 938 00:55:04,410 --> 00:55:10,160 That Martin Kleppmann is affiliated with the ATproto ecosystem, and I think that's 939 00:55:10,160 --> 00:55:16,127 just like one example for a broader, set of design challenges and like goals 940 00:55:16,187 --> 00:55:20,627 that I think is like nicely adjacent or partially overlapping with local-first. 941 00:55:20,627 --> 00:55:24,087 So do you want to give a little bit more detail on this topic? 942 00:55:24,647 --> 00:55:25,247 Yeah, for sure. 943 00:55:25,247 --> 00:55:28,537 I think social media and data ownership is so interesting because social 944 00:55:28,537 --> 00:55:34,373 media is the classic centralization and the classic opaque appliance 945 00:55:34,373 --> 00:55:38,603 with a mysterious algorithm that you can't and don't know how it works. 946 00:55:38,633 --> 00:55:44,273 And in general, I would say we could, obviously could, and indeed I have 947 00:55:44,273 --> 00:55:47,003 in the past done whole podcasts on social media and the both the 948 00:55:47,003 --> 00:55:48,283 positive and negatives that that has. 949 00:55:48,923 --> 00:55:53,753 Impacts that has had on society and us as individuals, but narrowing in 950 00:55:53,813 --> 00:55:57,390 a little more on this kind of, yeah, user empowerment around, and data 951 00:55:57,390 --> 00:55:59,610 ownership in our computing lives. 952 00:56:00,050 --> 00:56:04,563 It is the nature of the classics like an Instagram or a Facebook or even a 953 00:56:04,563 --> 00:56:07,653 Twitter back in its heyday that, you know, it, it is a very centralized 954 00:56:07,653 --> 00:56:09,453 closed, you have little ownership. 955 00:56:09,663 --> 00:56:13,490 You can just participate in this stream of, content. 956 00:56:13,903 --> 00:56:17,647 and now we have this new crop of, creations. 957 00:56:17,858 --> 00:56:21,537 I think Mastodon and ActivityPub was, kind of early out the gate, 958 00:56:21,537 --> 00:56:24,680 but now there's some really good energy around, BlueSky and ATproto. 959 00:56:24,810 --> 00:56:27,900 Actually, I think even as right as we record this, the ATmosphere 960 00:56:27,900 --> 00:56:29,610 Conferences is happening. 961 00:56:29,644 --> 00:56:33,447 which yeah, just a lot of good energy in there and it's 962 00:56:33,447 --> 00:56:34,977 a different set of problems. 963 00:56:34,977 --> 00:56:39,612 I think the classic local-first app is something more like Linear 964 00:56:39,612 --> 00:56:43,572 or Notion or Figma, which is a document editor, heavy creative work. 965 00:56:43,572 --> 00:56:45,312 You're collaborating with a small team. 966 00:56:45,312 --> 00:56:47,742 You really care a lot about the output of that data. 967 00:56:48,282 --> 00:56:51,612 Social media where it's sort of global town square, little fragments of 968 00:56:51,612 --> 00:56:56,262 data, everyone gets merged into a big stream is on the surface, a completely 969 00:56:56,352 --> 00:57:01,285 different, both technical and kind of design or user experience, challenge. 970 00:57:01,285 --> 00:57:04,435 But at the same time, when you look at what ActivityPub and ATproto have 971 00:57:04,435 --> 00:57:07,589 been grappling with, including things like personal data servers and how 972 00:57:07,589 --> 00:57:10,919 you have ownership over your identity, how you make your handle portable, 973 00:57:10,919 --> 00:57:14,669 but still kind of globally unique and especially verifiable, trustable, right? 974 00:57:14,809 --> 00:57:17,359 not easy to impersonate that you're the New York Times or a 975 00:57:17,359 --> 00:57:18,859 politician or something like that. 976 00:57:19,189 --> 00:57:22,279 So there's some really interesting technical and design work that has 977 00:57:22,279 --> 00:57:24,409 been done in both of those communities. 978 00:57:24,559 --> 00:57:28,022 I think we already have, some folks lined up to speak on some aspects of those. 979 00:57:28,022 --> 00:57:31,622 But I think there's, that's such a rich area that if you're working on something 980 00:57:31,862 --> 00:57:35,582 in this space with either one of those communities or otherwise in the kind 981 00:57:35,582 --> 00:57:39,962 of broader, how do we make social media something more empowering for users, 982 00:57:40,242 --> 00:57:42,889 yeah, we'd love to hear that, we'd love to see a, talk submission from you. 983 00:57:43,467 --> 00:57:43,857 Yeah. 984 00:57:43,857 --> 00:57:49,377 In particular also like using those protocols, those ideas to go beyond 985 00:57:49,377 --> 00:57:53,037 social media where you can basically like to, draw that bridge to the 986 00:57:53,037 --> 00:57:55,067 previous, like, identity question. 987 00:57:55,067 --> 00:57:58,949 Like right now I think this is, we we're blind to it at this point, that 988 00:57:58,949 --> 00:58:02,849 when there's new software, like, okay, of course I need to create an account. 989 00:58:03,209 --> 00:58:07,571 And like, depending on, which sort of patterns you use, maybe 990 00:58:07,631 --> 00:58:11,073 you automatically like, look for that sign in with Google button. 991 00:58:11,103 --> 00:58:15,476 Or if it's a developer oriented thing, like sign in with, GitHub, I'm more in 992 00:58:15,476 --> 00:58:19,649 the camp of like, where I always try to like just, create a standalone account 993 00:58:19,709 --> 00:58:22,044 just in case, something were to happen. 994 00:58:22,364 --> 00:58:24,651 but does it have to be this way? 995 00:58:24,711 --> 00:58:28,881 Are there other ways and like, and not just for. 996 00:58:29,131 --> 00:58:34,481 social media use cases, but for something that spans a much, wider range of 997 00:58:34,481 --> 00:58:39,271 scenarios, whether that is like something, I don't know, like maybe there's something 998 00:58:39,271 --> 00:58:44,801 in between where it's something in between Figma, Linear and, and social 999 00:58:44,801 --> 00:58:49,964 media, whether it's like a, smaller scale community that you are part of where 1000 00:58:49,964 --> 00:58:51,764 you do some very interesting things. 1001 00:58:51,764 --> 00:58:57,434 Maybe a community for a school that you volunteer in, or you have a much 1002 00:58:57,434 --> 00:59:02,017 richer set of different, kinds of media you deal with, whether it's documents, 1003 00:59:02,017 --> 00:59:07,084 whether it's chats, whether it's like some, posts, some forum like things. 1004 00:59:07,354 --> 00:59:09,634 Then you also have like things like identity. 1005 00:59:09,784 --> 00:59:12,964 And I think this is just like an example for thinking about 1006 00:59:12,964 --> 00:59:14,344 this a little bit more broadly. 1007 00:59:14,614 --> 00:59:18,884 So if you're using, some of those technologies or exploring that. 1008 00:59:19,604 --> 00:59:21,194 And see how we can bridge those worlds. 1009 00:59:21,264 --> 00:59:26,537 we'd love to hear from you and, going from one, one, tricky technical, 1010 00:59:26,804 --> 00:59:28,544 nuanced topic to another one. 1011 00:59:28,944 --> 00:59:33,414 Another like evergreen in local-first and beyond is like 1012 00:59:33,414 --> 00:59:35,964 encryption and encrypted messaging. 1013 00:59:36,234 --> 00:59:40,714 So, I think highly related to the areas that we talked about before. 1014 00:59:41,104 --> 00:59:44,651 And I think is another, aspect of software. 1015 00:59:44,651 --> 00:59:49,121 I think for a lot of software, kinda like still an optional part because it's hard, 1016 00:59:49,361 --> 00:59:54,634 but ideally over time it becomes just like a no-brainer to have to have encryption, 1017 00:59:54,874 --> 00:59:57,171 for your, data, for your applications. 1018 00:59:57,471 --> 01:00:02,448 It is, I think also dependent on the kind of like app you're working with. 1019 01:00:02,478 --> 01:00:05,688 So for example, if you're working in the finance space or the 1020 01:00:05,688 --> 01:00:09,731 healthcare space, I think it's much more, table stakes already. 1021 01:00:09,761 --> 01:00:13,211 And then there are other areas where it's much less common. 1022 01:00:13,604 --> 01:00:18,834 but ideally this is a quality that does not, quadruple the effort it takes 1023 01:00:18,834 --> 01:00:23,154 to build the app, but it's some just something, a data library, for example, 1024 01:00:23,154 --> 01:00:25,644 like Automerge takes care of for you. 1025 01:00:26,094 --> 01:00:29,324 And so yeah, those are like some preliminary thoughts 1026 01:00:29,344 --> 01:00:31,098 on this topic from, my end. 1027 01:00:31,128 --> 01:00:33,254 But yeah, curious to hear yours, Adam. 1028 01:00:33,914 --> 01:00:39,394 I mean, end to end encryption is first of all just such a powerful technology. 1029 01:00:39,454 --> 01:00:44,644 You know, I think wars have been one with, it is one frame you could use yet remains 1030 01:00:45,001 --> 01:00:49,673 hard to put into practice in a way that is approachable for users because, you 1031 01:00:49,673 --> 01:00:53,453 might think of it, everything boils down to a key management problem, but managing 1032 01:00:53,453 --> 01:00:58,223 keys is really difficult and incredibly abstract, even for very power users. 1033 01:00:58,446 --> 01:01:01,699 And certainly for a more mainstream computer user. 1034 01:01:02,179 --> 01:01:05,149 Yet you see all these places where encryption, when it has been brought to 1035 01:01:05,149 --> 01:01:08,809 bear effectively in the computing world, it is a complete game changer, I think 1036 01:01:08,809 --> 01:01:12,799 of like SSH for connecting to servers versus Telnet, which came before it. 1037 01:01:13,129 --> 01:01:17,599 I think of HTTPS for website encryption has, I think, literally brought trillions 1038 01:01:17,599 --> 01:01:21,979 about dollars of value into the economy through enabling things like e-commerce 1039 01:01:21,979 --> 01:01:24,793 and online, stock brokerages and so on. 1040 01:01:25,543 --> 01:01:29,923 Something like password managers and how they can provide a secure, setup, 1041 01:01:30,023 --> 01:01:32,531 based on their, encryption arrangement. 1042 01:01:32,531 --> 01:01:36,281 And then more recently you've got, for example, encrypted messaging like Signal. 1043 01:01:36,611 --> 01:01:41,411 And I think when we find ways to bring those to bear in a way that 1044 01:01:41,411 --> 01:01:45,851 is approachable for end users, it can just unlock so much value. 1045 01:01:45,851 --> 01:01:49,511 But it remains the case that there are a lot of areas where we haven't really 1046 01:01:49,691 --> 01:01:51,701 effectively brought it to bear fully. 1047 01:01:51,851 --> 01:01:55,114 And yeah, if you're working on any of these, whether it's something 1048 01:01:55,114 --> 01:01:58,384 relatively well established like encrypted messaging or something a 1049 01:01:58,384 --> 01:01:59,924 little more, on the cutting edge. 1050 01:01:59,924 --> 01:02:03,481 Like last year we heard a, an amazing talk from Brooklyn Zelenka about 1051 01:02:03,491 --> 01:02:07,771 Keyhive, which is a kind of adding an identity encryption layer to Automerge. 1052 01:02:08,438 --> 01:02:13,511 This just remains an area that is very hard, but I think again, just has the 1053 01:02:13,511 --> 01:02:15,071 potential to provide so much value. 1054 01:02:15,941 --> 01:02:16,481 Awesome. 1055 01:02:16,611 --> 01:02:21,691 so another adjacent topic, a little bit more lighthearted, and yet probably one 1056 01:02:21,691 --> 01:02:27,214 of the most, OG adopters of local-first, and they've always built software 1057 01:02:27,214 --> 01:02:29,524 this way, which is game development. 1058 01:02:30,088 --> 01:02:37,318 I think most games, particularly in the era of like internet not 1059 01:02:37,318 --> 01:02:42,118 being as ubiquitous as it is today, have like, out of necessity. 1060 01:02:42,156 --> 01:02:44,014 They needed to be local-first. 1061 01:02:44,014 --> 01:02:49,414 So you can play single player campaigns or whatever, but also like 1062 01:02:49,414 --> 01:02:54,221 in that gold, era you had, a lot of games also having multiplayer modes. 1063 01:02:54,311 --> 01:02:56,951 And this is where you had like LAN parties, et cetera. 1064 01:02:56,951 --> 01:03:01,527 And like, software was just like built in a different way where like internet 1065 01:03:01,691 --> 01:03:05,261 a lot of games, even didn't like support it or like was one, feature 1066 01:03:05,261 --> 01:03:07,151 that was added a little bit later. 1067 01:03:07,601 --> 01:03:08,796 And it's just like a. 1068 01:03:09,191 --> 01:03:14,951 I think an interesting case study for how like a different parallel universe evolved 1069 01:03:15,161 --> 01:03:19,691 and like did very complex data management with a lot of like similar challenges 1070 01:03:19,691 --> 01:03:23,651 where you needed to have like, reactivity you needed to be like, very efficient. 1071 01:03:23,651 --> 01:03:26,801 You needed to like care about persistence. 1072 01:03:27,007 --> 01:03:30,570 so I think there's like a lot of parallel universe lessons learned 1073 01:03:30,780 --> 01:03:35,654 that maybe never really, fully percolated into the web ecosystem. 1074 01:03:35,874 --> 01:03:37,667 Probably sometimes for, the worst. 1075 01:03:37,667 --> 01:03:44,397 I think I always feel, I can learn a lot from the game ecosystem overall, as 1076 01:03:44,487 --> 01:03:48,587 a lot of like lessons have been already much, earlier been learned there. 1077 01:03:48,917 --> 01:03:52,517 And I think now it's like that, that is still true today. 1078 01:03:52,847 --> 01:03:57,462 And video games have always been like the, one of the most demanding use cases 1079 01:03:57,762 --> 01:04:02,679 and where like latency really matters, where like all of those things, matter. 1080 01:04:02,769 --> 01:04:07,505 And, I think it's just about time that we bridge the world from game 1081 01:04:07,505 --> 01:04:09,935 development to the local-first ecosystem. 1082 01:04:10,235 --> 01:04:15,642 So if you're, doing things in the game development space and you have 1083 01:04:15,642 --> 01:04:19,452 some interesting stories to share about how your game has always 1084 01:04:19,452 --> 01:04:23,955 been local-first, then, please definitely consider submitting a talk. 1085 01:04:23,955 --> 01:04:26,085 So that, that's kind of my, my perspective on it. 1086 01:04:26,475 --> 01:04:29,265 But I think you have even more nuanced takes on this, Adam. 1087 01:04:29,689 --> 01:04:30,445 Well, yeah. 1088 01:04:30,445 --> 01:04:32,425 I don't know about nuance, but it's definitely close to my heart. 1089 01:04:32,425 --> 01:04:34,315 I actually began my career in video games. 1090 01:04:34,315 --> 01:04:38,125 I probably like a lot of little kids, I basically got inspired to 1091 01:04:38,125 --> 01:04:39,625 learn to program, to make games. 1092 01:04:40,050 --> 01:04:41,540 And, began my career there. 1093 01:04:41,540 --> 01:04:44,420 I worked for four or five years in the game industry, including on 1094 01:04:44,420 --> 01:04:46,610 PlayStation 2 games and Dreamcast games. 1095 01:04:46,610 --> 01:04:48,520 And that was kind of the era then. 1096 01:04:48,964 --> 01:04:53,934 And it's an incredible world of technology that's very separate from the productivity 1097 01:04:53,934 --> 01:04:58,977 software, web operating system, thing where you're just drawing 60 frames or 1098 01:04:58,977 --> 01:05:01,977 120 frames per second on a blank canvas. 1099 01:05:02,277 --> 01:05:06,064 The APIs to the underlying system are just a very, very different thing 1100 01:05:06,064 --> 01:05:08,884 than perhaps what we're used to, in some ways easier to work with, in some 1101 01:05:08,884 --> 01:05:13,834 ways harder, but yeah, the performance and being able to use it wherever. 1102 01:05:14,187 --> 01:05:17,607 even certainly in the age before internet connection was ubiquitous, 1103 01:05:17,607 --> 01:05:21,057 but even nowadays, yeah, you expect if you have a game on your device, that 1104 01:05:21,057 --> 01:05:23,367 you should be able to play it on the plane and not be worried about whether 1105 01:05:23,367 --> 01:05:25,347 you have internet in that moment. 1106 01:05:25,347 --> 01:05:27,897 Even think something like, you know, even once you start to add some 1107 01:05:27,897 --> 01:05:31,767 internet connectivity, Hey, I wanna post my high score to a global thing, 1108 01:05:31,767 --> 01:05:34,377 but you don't expect the game will just stop working the moment that the 1109 01:05:34,377 --> 01:05:37,367 high score API is not available, so. 1110 01:05:37,699 --> 01:05:41,504 And then you have so much going on in the world of these different engines like 1111 01:05:41,504 --> 01:05:47,207 Unreal and Unity and Gadot engine slash game development tools and yeah, I just 1112 01:05:47,207 --> 01:05:51,124 feel like there's, maybe a lot these communities can learn from each other 1113 01:05:51,124 --> 01:05:54,154 and yeah, if you're someone in that, tough space, we'd like to hear from you. 1114 01:05:54,802 --> 01:05:59,185 Yeah, I think there are just so many amazing stories waiting to be heard in 1115 01:05:59,185 --> 01:06:04,349 the local-first ecosystem, whether it's someone who like struggled to scale 1116 01:06:04,349 --> 01:06:09,352 World of Warcraft, to like massively multiplayer scenarios where I mean, 1117 01:06:09,352 --> 01:06:11,702 it's already hard to build multiplayer. 1118 01:06:12,007 --> 01:06:17,127 like systems for 10 users, but now, like, think about this, like in the, tens of 1119 01:06:17,127 --> 01:06:22,267 thousands of users, in like, in much harsher, real time scenarios, et cetera. 1120 01:06:22,267 --> 01:06:27,071 So, or whether you're building like, more modern games that, they're currently 1121 01:06:27,101 --> 01:06:29,141 like scaling things even further. 1122 01:06:29,447 --> 01:06:30,917 so yeah, we'd love to hear from you. 1123 01:06:31,071 --> 01:06:35,674 if you think this is interesting for the local-first community and as the 1124 01:06:35,674 --> 01:06:41,331 last topic, in the list, we, we probably like the, kinda like the key topic that 1125 01:06:41,471 --> 01:06:46,501 is the, center for us this year is like malleable software in the age of LLMs or. 1126 01:06:47,281 --> 01:06:49,561 malleable software like at all. 1127 01:06:49,651 --> 01:06:52,201 But then also in the age of LLMs, that's new. 1128 01:06:52,201 --> 01:06:56,731 That was not really like the, we got like the early glimpses of that last 1129 01:06:56,731 --> 01:07:00,991 year, but at last year's conference, but I think it was still so early, I don't 1130 01:07:00,991 --> 01:07:03,061 think like Claude Code was released yet. 1131 01:07:03,306 --> 01:07:07,271 ChatGPT was out there, of course, but I think it was not yet at a 1132 01:07:07,271 --> 01:07:12,264 point where it like really clicked for everyone that this is real and 1133 01:07:12,264 --> 01:07:13,734 not just like a niche use case. 1134 01:07:14,214 --> 01:07:19,801 And I think now it's like, already so ubiquitously available that we can 1135 01:07:19,981 --> 01:07:22,171 make this a part of our application. 1136 01:07:22,501 --> 01:07:25,831 And I think there's like really one of the broadest topics here, almost like a 1137 01:07:25,831 --> 01:07:30,580 catchall for a lot of, different, talks that could fit into the conference. 1138 01:07:30,903 --> 01:07:34,023 the anchoring point here is like, uh, what we've talked about before with 1139 01:07:34,023 --> 01:07:42,303 like Maggie's, closing keynote from 2024 was it, with like barefoot developers. 1140 01:07:42,843 --> 01:07:46,146 And I think this can, cast a really wide net. 1141 01:07:46,196 --> 01:07:48,920 And, what are your thoughts on, this topic? 1142 01:07:49,190 --> 01:07:51,860 Yeah, we'll probably first just defining malleable software. 1143 01:07:51,860 --> 01:07:56,390 For those not familiar, this is the idea that software systems are ones 1144 01:07:56,390 --> 01:08:01,370 that could be modified by their users as we're using it or to adapt to our needs. 1145 01:08:01,800 --> 01:08:06,460 and of course, Ink & Switch has a, I dunno, 8,000 word essay up on the website 1146 01:08:06,730 --> 01:08:10,481 if you search for malleable software, that can dive into detail on that. 1147 01:08:10,781 --> 01:08:14,741 But, you know, how does language model assisted coding it seems to enable so 1148 01:08:14,741 --> 01:08:18,671 much the ability to create new software and especially for people who are on the, 1149 01:08:18,761 --> 01:08:22,281 not necessarily professional software developers, even if they're technical and 1150 01:08:22,651 --> 01:08:24,401 good at thinking about software systems. 1151 01:08:24,821 --> 01:08:29,006 But when you think of the classic, let's call it vibe coded app, it's 1152 01:08:29,006 --> 01:08:32,636 let me make a one-off website that I can post, but that's very different 1153 01:08:32,636 --> 01:08:37,136 from being able to modify the tools that I'm relying on in my daily life. 1154 01:08:37,166 --> 01:08:40,766 My text editor, my email client, my web browser, et cetera. 1155 01:08:41,243 --> 01:08:45,743 so I think on one hand, language model coding should, in theory, 1156 01:08:45,983 --> 01:08:50,903 really open up some major new possibilities for malleable software. 1157 01:08:51,083 --> 01:08:53,963 On the other hand, I think we have some of the same old problems we have always had 1158 01:08:53,963 --> 01:08:59,703 around data siloing around kind of closed ecosystems, closed appliance heavily. 1159 01:09:00,033 --> 01:09:04,703 You know, the control, the locus of control lies completely with the software 1160 01:09:04,703 --> 01:09:07,163 vendors and very little with the users. 1161 01:09:07,453 --> 01:09:11,983 and how has this world evolved or what new possibilities are opened by the 1162 01:09:11,983 --> 01:09:16,543 existence of language model assisted coding or other AI things or other things 1163 01:09:16,543 --> 01:09:20,803 that have changed in the technology world since we last convened in 2025? 1164 01:09:21,553 --> 01:09:26,413 Yeah, I think this is, I mean, this topic in itself has been sort of like a core 1165 01:09:26,413 --> 01:09:31,273 pillar for this year's conference, and I think we've like added quite a lot other 1166 01:09:31,273 --> 01:09:36,580 ones just because, it felt like the right moment to, to broaden the net a bit. 1167 01:09:37,090 --> 01:09:41,353 But I think here, like malleable software, this was, a theme that I 1168 01:09:41,353 --> 01:09:45,239 mean, is is one of the key tracks of, the Ink & Switch, family. 1169 01:09:45,479 --> 01:09:50,009 But I think now with AI models, et cetera, it's really something 1170 01:09:50,009 --> 01:09:51,419 that's becoming mainstream. 1171 01:09:51,809 --> 01:09:56,189 And where before I think it was like very something that developers 1172 01:09:56,189 --> 01:09:58,199 maybe had romantic thoughts about. 1173 01:09:58,199 --> 01:10:02,986 Like, oh yeah, this was like, also on my list of side projects I wanted to build. 1174 01:10:03,316 --> 01:10:06,556 And now you can actually build it for yourself. 1175 01:10:06,556 --> 01:10:10,966 Or maybe you built software for like someone in your family, or it's 1176 01:10:10,966 --> 01:10:14,776 just like a, almost as an act of kindness of like supporting someone 1177 01:10:14,916 --> 01:10:16,646 like this is through your abilities. 1178 01:10:16,646 --> 01:10:18,556 And now it becomes so easy. 1179 01:10:19,461 --> 01:10:25,458 I think now that will lead to so much, creation of, new software that will also 1180 01:10:25,458 --> 01:10:27,738 have like interesting new consequences. 1181 01:10:28,104 --> 01:10:31,401 whether it's that we need to wrestle with, the quality of the software, 1182 01:10:31,451 --> 01:10:33,051 or the interoperability of it. 1183 01:10:33,471 --> 01:10:37,881 And also like, yeah, how much control do we give to, users? 1184 01:10:37,881 --> 01:10:42,124 How much control, do we, like where do we draw the line between, 1185 01:10:42,404 --> 01:10:47,114 a developer building the software versus a user using the software? 1186 01:10:47,334 --> 01:10:50,551 Like a lot of power tools are very customizable. 1187 01:10:50,551 --> 01:10:53,161 But now I think that line gets more and more blurry. 1188 01:10:53,464 --> 01:10:57,375 We'd love to hear from people who have been wrestling with, those 1189 01:10:57,375 --> 01:11:02,275 blurry lines and trying out different ideas and sharing your learnings. 1190 01:11:02,464 --> 01:11:05,301 What sort of software you've been building as a barefoot developer. 1191 01:11:05,498 --> 01:11:06,068 Yeah. 1192 01:11:06,098 --> 01:11:10,617 I think this is, quite the portfolio of, different themes and 1193 01:11:10,617 --> 01:11:15,121 different topics that we're gonna have for, this year's conference. 1194 01:11:15,421 --> 01:11:15,841 That's right. 1195 01:11:15,841 --> 01:11:19,711 We maybe even run the risk of being too scattershot or too wide 1196 01:11:19,716 --> 01:11:21,511 ranging, but in some ways I'd rather. 1197 01:11:21,776 --> 01:11:24,676 take the risk of having too many ideas, too many interesting, 1198 01:11:24,849 --> 01:11:28,596 intellectual, avenues to pursue, then get stuck in a rut. 1199 01:11:28,596 --> 01:11:32,186 And now seems like a great time to do it with the changes happening, in the 1200 01:11:32,186 --> 01:11:36,506 technology world, but in general because of the foundation of success we've had 1201 01:11:36,506 --> 01:11:40,166 in the event to date and the community we've built around, that just seems 1202 01:11:40,166 --> 01:11:43,686 like the perfect time to, broaden and bring new people in and fresh ideas. 1203 01:11:44,217 --> 01:11:49,257 And like all of that is not to say that we ran out of ideas with like the, 1204 01:11:49,497 --> 01:11:53,911 traditional way how we did the Local-First Conference, the last two years before. 1205 01:11:54,271 --> 01:11:58,657 But I think what it's always been all about for me and for I'm sure for you as 1206 01:11:58,657 --> 01:12:02,697 well is like the people who actually come and like the conversations we're having. 1207 01:12:03,147 --> 01:12:08,817 And it is like no one in that ecosystem has been there only because they 1208 01:12:08,907 --> 01:12:10,617 only care about the data sync engine. 1209 01:12:10,617 --> 01:12:16,497 It's always been in service of like some broader vision and like some specific 1210 01:12:16,497 --> 01:12:21,647 use case that really like, had real impact and real meaning for their life. 1211 01:12:22,067 --> 01:12:24,767 And like I think all of those different themes we now talked 1212 01:12:24,767 --> 01:12:29,634 about is like one manifestation, one direction of that and local-first. 1213 01:12:30,129 --> 01:12:31,719 It's like the foundation of that. 1214 01:12:31,719 --> 01:12:33,639 It's kinda like what brings us all together. 1215 01:12:34,029 --> 01:12:37,779 And I think what all of those people who have attended and have contributed to 1216 01:12:37,779 --> 01:12:42,260 the conferences have, in common is that they have a very, deep understanding of 1217 01:12:42,260 --> 01:12:47,712 the, technical world and of the world more broadly, but, also a broad interest 1218 01:12:47,977 --> 01:12:53,244 and, are always keen to be inspired to think a little bit different about, 1219 01:12:53,574 --> 01:12:56,094 like be beyond their current horizon. 1220 01:12:56,154 --> 01:13:01,592 And I think this is, what we, try to, also, facilitate through the conference 1221 01:13:01,622 --> 01:13:06,598 where those, different talks are hopefully little points of inspiration that can 1222 01:13:06,598 --> 01:13:11,392 lead to further exploration by everyone who's attending and paying interest. 1223 01:13:11,975 --> 01:13:12,635 Well said. 1224 01:13:13,145 --> 01:13:16,535 Well, I think we hammered in the, you know, submit a talk call to action. 1225 01:13:16,796 --> 01:13:17,895 we'll say it one more time. 1226 01:13:17,895 --> 01:13:20,898 The CFPs open till I think, end of April. 1227 01:13:20,928 --> 01:13:23,478 So you have a little bit of time, but not too much time. 1228 01:13:23,808 --> 01:13:26,508 But also if, even if you don't wanna give a talk, buy a ticket, 1229 01:13:26,568 --> 01:13:28,488 just come listen to all the ideas. 1230 01:13:28,845 --> 01:13:31,155 we got sponsorship booths available as well. 1231 01:13:31,155 --> 01:13:35,415 If, the folks at the, an event like this are interesting to, talk 1232 01:13:35,415 --> 01:13:36,885 to for your, for your company. 1233 01:13:37,095 --> 01:13:40,935 So we'd love to have you come participate in any of those three forms. 1234 01:13:42,150 --> 01:13:42,690 Awesome. 1235 01:13:42,810 --> 01:13:46,090 Well, Adam, thanks so much for talking this story together. 1236 01:13:46,160 --> 01:13:49,470 I feel like this is also already Yeah. 1237 01:13:49,470 --> 01:13:54,507 Given me so much excitement for, what we're gonna hear, at the conference 1238 01:13:54,547 --> 01:13:58,557 and like in a way where we still don't fully know which stories we'll hear, 1239 01:13:58,557 --> 01:14:01,110 but we'll know about the broad topics. 1240 01:14:01,310 --> 01:14:07,050 So that've got me very excited already and I can't wait for, the conference and 1241 01:14:07,050 --> 01:14:09,420 to meet a lot of familiar and new faces. 1242 01:14:10,350 --> 01:14:10,980 Same here. 1243 01:14:11,040 --> 01:14:12,660 I'm excited for the new themes. 1244 01:14:12,660 --> 01:14:16,590 I'm excited to see old friends, and I'm excited to enjoy beautiful, 1245 01:14:16,860 --> 01:14:18,900 beautiful Berlin in the summertime. 1246 01:14:21,270 --> 01:14:21,750 All righty. 1247 01:14:21,870 --> 01:14:22,620 Thanks so much. 1248 01:14:22,620 --> 01:14:23,220 Take care. 1249 01:14:23,225 --> 01:14:23,405 Bye. 1250 01:14:23,785 --> 01:14:24,005 Bye.