บทถอดความฉบับเต็มของ Ep #17: Frank Bach (Anthropic): Building Your Own Design Tool, Beginning Anywhere, and a Swiss Army Knife Career

in the life of Frank at Anthropic look like? 00:02:02.440 --> 00:02:02.700 and, 00:02:02.806 --> 00:02:03.096 Frank Bach: Great. 00:02:03.120 --> 00:02:07.010 Jonny: know, how does the design team function and, and, I'd love to just hear 00:02:07.010 --> 00:02:08.410 how you, how you guys design over there. 00:02:09.624 --> 00:02:10.454 Frank Bach: Yeah, absolutely. 00:02:11.964 --> 00:02:15.184 I wish there was a, a perfect answer to that question, but it 00:02:15.194 --> 00:02:17.204 really definitely changes every day. 00:02:18.114 --> 00:02:20.334 what is a day of designing at Anthropic like? 00:02:20.334 --> 00:02:25.124 I think the, the speed at which the technology is moving, and it's probably 00:02:25.154 --> 00:02:29.184 similar, a similar vibe at a lot of these kind of AI lab companies, 00:02:29.654 --> 00:02:31.264 is it's all moving so quickly. 00:02:31.284 --> 00:02:36.144 There's, you know, there's regulation, there's, there's new technologies, 00:02:36.144 --> 00:02:37.794 there's new data centers being built. 00:02:37.804 --> 00:02:39.624 Like, all this stuff is happening kind of… 00:02:40.054 --> 00:02:40.764 What's that movie? 00:02:40.764 --> 00:02:42.964 Like, everything everywhere all at once kind of thing. 00:02:42.964 --> 00:02:44.534 Like, that's sort of the feeling. 00:02:45.454 --> 00:02:51.064 the pace is very intense, and I would say it's not unusual, that you will, 00:02:51.744 --> 00:02:54.844 know, like a, like a 12-month roadmap probably doesn't make any sense. 00:02:54.914 --> 00:02:57.204 Like, you're sort of reprioritizing things on the day. 00:02:58.344 --> 00:03:02.474 so for people who really thrive in, like, a super structured sort of like, 00:03:02.474 --> 00:03:05.884 "I need to know what I'm doing over the, over the next several months," it would 00:03:05.884 --> 00:03:07.464 probably be a bit of a tough place to be. 00:03:08.274 --> 00:03:10.734 people who thrive on, you know, if you have maybe have a bit of an 00:03:10.744 --> 00:03:12.484 entrepreneurial mindset and you get… 00:03:12.504 --> 00:03:17.894 You know, you're energized by the speed of things, I think that it, it becomes a very 00:03:17.894 --> 00:03:22.324 fun place to work, of like, you know, you don't really know what's coming tomorrow. 00:03:23.024 --> 00:03:25.344 if you can live with that unpredictability, then awesome. 00:03:26.174 --> 00:03:30.924 so that's-- I, I'm feeling, like, super, super energized by it all. 00:03:30.924 --> 00:03:33.764 And I was, you know, just before this job, I was at DoorDash, and 00:03:33.764 --> 00:03:35.334 I, I quite enjoyed my time there. 00:03:36.104 --> 00:03:37.654 and then before that, Instagram. 00:03:37.684 --> 00:03:42.534 Both those companies were very much, sort of reset the roadmap every six 00:03:42.534 --> 00:03:45.344 months kind of situations, where you would, you know, you'd build these 00:03:45.344 --> 00:03:49.354 long, elaborate presentations that would sort of show what is my big bet 00:03:49.664 --> 00:03:53.374 for the next three to six months, and, like, how have we found conviction 00:03:53.374 --> 00:03:56.504 and what do we believe in and what's our North Star and all these things. 00:03:56.504 --> 00:04:00.064 And, like, those are great skills to have, and you become kind of a good storyteller. 00:04:00.734 --> 00:04:03.774 but when it comes to, like, what stuff actually got built at the end of the 00:04:03.774 --> 00:04:05.524 year, it's a totally different story. 00:04:06.704 --> 00:04:09.884 so being here and, like, you know, boots on the ground, sort of like 00:04:09.914 --> 00:04:14.684 builder mode, we're making things every single day, I'm absolutely loving it. 00:04:14.694 --> 00:04:17.324 You know, I think it, you know, the hours can be long and the work's 00:04:17.324 --> 00:04:20.824 intense, but I finish every day feeling like, "Holy shit, that was a day. 00:04:20.874 --> 00:04:22.684 That was like, that was fun." 00:04:22.774 --> 00:04:23.024 Like… 00:04:24.084 --> 00:04:24.484 Jonny: Awesome. 00:04:24.974 --> 00:04:28.024 you posted recently, and this, this is the post that caught my eye and 00:04:28.344 --> 00:04:28.694 Frank Bach: Yeah 00:04:28.714 --> 00:04:32.884 Jonny: I decided I just had to, had to drop you a, drop you a DM. 00:04:32.884 --> 00:04:36.524 But, you described building your own design tool, 00:04:37.240 --> 00:04:37.710 Frank Bach: Yes 00:04:38.464 --> 00:04:40.104 Jonny: and, that was what? 00:04:40.114 --> 00:04:41.164 A month ago maybe. 00:04:42.474 --> 00:04:44.364 first of all, can you describe it to us? 00:04:44.364 --> 00:04:47.814 You know, it's-- it was, slightly mysterious in your LinkedIn post that 00:04:47.814 --> 00:04:48.884 we'll put in the show notes, but, 00:04:49.088 --> 00:04:49.438 Frank Bach: I know 00:04:49.474 --> 00:04:51.904 Jonny: can you describe a little bit more how it works and, and you know, 00:04:51.928 --> 00:04:52.028 Frank Bach: For 00:04:52.184 --> 00:04:53.144 Jonny: where is it up to now? 00:04:53.144 --> 00:04:54.634 Is it, has it ca- carried on evolving? 00:04:55.224 --> 00:04:55.664 Frank Bach: For sure. 00:04:55.714 --> 00:04:57.894 So the, yeah, that LinkedIn post kind of took off. 00:04:57.894 --> 00:05:01.384 I wasn't expecting that, and it's led to a lot of, interesting DMs of 00:05:01.394 --> 00:05:02.974 people wanting me to show the tool. 00:05:03.424 --> 00:05:05.204 Unfortunately, I cannot show the tool. 00:05:05.204 --> 00:05:05.914 I wish I could. 00:05:05.944 --> 00:05:09.704 Maybe one day it will be a case study, of mine, and I'll have to 00:05:09.704 --> 00:05:13.694 sort of strip away all the, all the, the sort of inside information. 00:05:13.704 --> 00:05:15.794 But I'll do my best to explain it. 00:05:16.394 --> 00:05:17.514 I started at Anthropic. 00:05:17.584 --> 00:05:19.444 People told me, "This place is crazy. 00:05:19.444 --> 00:05:20.234 It's intense. 00:05:22.254 --> 00:05:26.534 onboarding weeks will be the,

you know, the lightest sort of weeks of the job." 00:05:26.534 --> 00:05:29.724 You know, pretty typical of any, any job before you really hit your first project. 00:05:30.710 --> 00:05:35.230 And I think when I was at DoorDash, we were getting really into the, the 00:05:35.250 --> 00:05:40.590 AI tooling, but it was sort of in this like half Figma, half, you know, 00:05:40.600 --> 00:05:42.890 dabbling in Cursor, Claude, et cetera. 00:05:42.890 --> 00:05:46.870 And you sort of like don't really know where, to start, where to, you know, 00:05:47.030 --> 00:05:48.530 how do I start a project these days? 00:05:48.530 --> 00:05:51.580 You know, is it a new page in Figma or like what do I even do? 00:05:52.430 --> 00:05:53.520 And I thought, you know what? 00:05:53.550 --> 00:05:57.080 There's this, there's this vibe right now that-- And I don't know if it's, 00:05:57.090 --> 00:06:01.510 it's right or wrong, but the people who tend to stand out right now are 00:06:01.510 --> 00:06:03.270 the people who are not using Figma. 00:06:04.280 --> 00:06:06.790 I'm sure it's just a trend, a moment we're kind of going through 00:06:06.790 --> 00:06:07.620 and I thought, "You know what? 00:06:07.880 --> 00:06:08.620 I'm right here. 00:06:08.660 --> 00:06:10.260 I'm in this AI company. 00:06:10.480 --> 00:06:13.670 I should probably make the best of the situation and kind of like 00:06:13.940 --> 00:06:17.950 during my chill, chill onboarding week, see what I can get up to." 00:06:17.970 --> 00:06:18.240 You know? 00:06:18.240 --> 00:06:22.060 Like I, I've got, you know, I've got a enterprise Claude account, lots of 00:06:22.060 --> 00:06:23.870 tokens, and I can make cool stuff. 00:06:24.120 --> 00:06:25.510 So where, you know, what the hell? 00:06:25.560 --> 00:06:26.060 Why not? 00:06:26.550 --> 00:06:32.060 And then so I, I got my first project, that was like, you know, sort of most 00:06:32.060 --> 00:06:35.270 jobs you'll start your first, your first project is kind of like a softball 00:06:35.300 --> 00:06:38.720 thing just to kind of get into the design system and, you know, kind of get 00:06:38.720 --> 00:06:40.240 your, get your feet wet, so to speak. 00:06:40.240 --> 00:06:46.150 And, I, I kick this project off and I'm, I'm kind of using Claude 00:06:46.160 --> 00:06:50.790 to, generate HTML prototypes of, of what I'm trying to build. 00:06:50.790 --> 00:06:54.840 And Claude was already hooked up to the design system, so 00:06:54.840 --> 00:06:55.830 it's all coming out, you know. 00:06:56.170 --> 00:06:59.860 Thankfully, designing with Claude looks, it kind of looks like Claude all the time. 00:06:59.860 --> 00:07:02.100 So when you are building Claude, it's great. 00:07:02.430 --> 00:07:02.800 You know? 00:07:02.850 --> 00:07:05.010 When, when you're trying to build something else that kind 00:07:05.010 --> 00:07:07.360 of looks like Claude, it can be a bit of a different situation. 00:07:08.720 --> 00:07:11.560 so I'm, I'm building this prototype

and I'm, you know, I'm feeling, 00:07:11.620 --> 00:07:12.930 man, this is-- I'm having fun. 00:07:12.930 --> 00:07:14.290 I'm just prompting, typing. 00:07:14.600 --> 00:07:18.530 It's kind of I'm generating stuff at a speed that I've never quite done before. 00:07:18.540 --> 00:07:21.830 You know, gone are, I think gone are the days of spending three 00:07:21.830 --> 00:07:25.850 to four weeks, know, finessing the settings screen or whatever. 00:07:25.850 --> 00:07:27.640 You know, like we're not, we're not doing that anymore. 00:07:27.660 --> 00:07:27.930 We're… 00:07:28.710 --> 00:07:32.170 And I'm building this thing and I'm g- I'm getting excited about this. 00:07:32.200 --> 00:07:32.900 Okay, cool, great. 00:07:32.940 --> 00:07:35.800 I've, I've like I, I sort of hit this moment I should share 00:07:35.800 --> 00:07:37.210 this with my product manager. 00:07:37.410 --> 00:07:37.700 Okay. 00:07:37.700 --> 00:07:39.360 How do I even share it with my product manager? 00:07:39.490 --> 00:07:41.430 I go, "Shoot, maybe I should have been in Figma. 00:07:42.350 --> 00:07:45.450 I've taken a really bad s- you know, a big misstep here and 00:07:45.450 --> 00:07:46.730 I've totally wasted my time. 00:07:47.070 --> 00:07:50.420 So how would I rebuild what I've done in this prototype in Figma?" 00:07:50.730 --> 00:07:51.590 Then I go, "Wait a second." 00:07:52.400 --> 00:07:53.350 I prompt it, you know. 00:07:53.360 --> 00:07:54.830 What if I make a share function? 00:07:55.330 --> 00:07:56.720 Okay, now we got a share function. 00:07:56.790 --> 00:07:57.380 Okay, cool. 00:07:57.380 --> 00:07:58.910 This is starting to become something. 00:07:59.260 --> 00:08:00.320 I can share the project. 00:08:00.992 --> 00:08:04.122 And then my, my product manager was away that day or something 00:08:04.122 --> 00:08:04.952 and I go, "You know what? 00:08:04.952 --> 00:08:07.852 Maybe I should record a Loom video of like walking through it." 00:08:07.942 --> 00:08:11.352 And then I just prompt it, "Hey, generate me, make me a Loom feature." 00:08:11.742 --> 00:08:12.632 And then boom, it's there. 00:08:12.632 --> 00:08:14.782 And it's like it's not quite perfect, but it's sort of there. 00:08:14.782 --> 00:08:20.122 I'm like, "Holy gee-" Like, in like three prompts I've got my f- my camera 00:08:20.282 --> 00:08:24.552 facing like in, you know, feed in the f- in the HTML prototype and a share 00:08:24.552 --> 00:08:28.612 link and I'm like, I, I'm just in this like insane workflow that I've never, 00:08:28.632 --> 00:08:32.242 you know, I haven't felt in years, 'cause we've gotten so used to just 00:08:32.242 --> 00:08:34.132 the classic Figma way of doing things. 00:08:34.182 --> 00:08:36.162 And this is not a, a diss at Figma in any way. 00:08:36.162 --> 00:08:37.212 I think they're fantastic. 00:08:38.152 --> 00:08:40.442 but I was just like, I'm like, "Oh my God," I'm-- it's like 00:08:40.442 --> 00:08:41.852 riding a bike for the first time. 00:08:41.852 --> 00:08:42.902 I'm like, "I'm really doing it. 00:08:42.912 --> 00:08:43.682 I'm really building." 00:08:43.692 --> 00:08:45.552 You know, it's like something I've not done before. 00:08:46.212 --> 00:08:50.542 And, you know, like just feature by feature I'm sort of building the thing 00:08:50.542 --> 00:08:56.062 I need in the moment, on the side of building the project, you know? 00:08:56.062 --> 00:09:00.442 So I'm like of like not even aware that I'm doing this, but eventually I've got 00:09:00.442 --> 00:09:04.542 this fully functioning like mini tool that's perfect for what I needed that day. 00:09:05.120 --> 00:09:05.430 Jonny: Right 00:09:06.002 --> 00:09:09.102 Frank Bach: I sort of hit the next wall which was, that 00:09:09.102 --> 00:09:11.252 project is kind of complete. 00:09:11.332 --> 00:09:12.872 Now here's your next project." 00:09:13.572 --> 00:09:15.052 "Okay, where do I begin for this one?" 00:09:15.062 --> 00:09:19.982 And I go, "Okay, what if I sort of back up from this, this project and build 00:09:19.982 --> 00:09:22.182 a container for multiple projects?" 00:09:22.192 --> 00:09:25.172 So now I'm kind of like rebuilding, you know, the Figma sort of 00:09:25.172 --> 00:09:27.402 recent project screen in a way. 00:09:28.072 --> 00:09:29.962 And then I start doing all these ridiculous things. 00:09:29.962 --> 00:09:31.182 I'm like, "Oh, I'm having so much fun. 00:09:31.182 --> 00:09:35.592 What if my, What if I make like some shortcut keys? 00:09:35.592 --> 00:09:38.412 If I hit the R button on my keyboard it refreshes. 00:09:38.602 --> 00:09:41.202 What if I add like a glimmer over the screen? 00:09:41.202 --> 00:09:46.702 What if, what if every time I refresh I get, a wallpaper background of 00:09:46.702 --> 00:09:50.032 like a piece of artwork from a museum so I'm like inspired by 00:09:50.052 --> 00:09:51.342 seeing new things all the time?" 00:09:51.342 --> 00:09:55.022 And I'm like, "You would never ship this in a enterprise tool to the world." 00:09:55.032 --> 00:09:59.322 But for me and what I needed that day and that week was like, "This is awesome." 00:09:59.322 --> 00:10:01.442 And then I'm, you know, people got pretty excited about it. 00:10:01.452 --> 00:10:04.372 My manager told me, "You should record a video using the tool and share it 00:10:04.372 --> 00:10:08.102 with the team," and like, you know, hopefully being able to inspire 00:10:08.102 --> 00:10:10.022 some of the Claude Design stuff. 00:10:10.052 --> 00:10:14.772 And then like in some kind of backwards way I sort of rebuilt a like Frank's 00:10:14.772 --> 00:10:16.512 crazy version of Claude Design. 00:10:17.512 --> 00:10:17.872 you know. 00:10:17.992 --> 00:10:20.172 And so yeah, that, that's kind of the story there. 00:10:20.552 --> 00:10:23.612 And, at this point now I'm still using that tool. 00:10:23.612 --> 00:10:25.932 I've definitely like leaned more into the Figma usage. 00:10:27.342 --> 00:10:32.532 but yeah, it was, it was a, it was a

really fun, exciting time of, yeah, just, 00:10:32.542 --> 00:10:37.092 just seeing what is really possible with, you know, access to all of these tools. 00:10:38.670 --> 00:10:39.880 Jonny: It's a really good example of like, 00:10:42.180 --> 00:10:45.540 think and act and, and design just sort of emerging out of you. 00:10:45.710 --> 00:10:48.050 whether you would call it design or it probably is, you know, 00:10:48.060 --> 00:10:50.410 y- your-- I, I need this thing. 00:10:50.480 --> 00:10:52.500 This is the vibe right now, depending on my mood 00:10:52.502 --> 00:10:52.892 Frank Bach: Right. 00:10:52.950 --> 00:10:55.900 Jonny: at this moment in time, I'm gonna add something superfluous or 00:10:56.252 --> 00:10:56.622 Frank Bach: Right. 00:10:57.200 --> 00:10:58.610 Jonny: and I don't need to justify it to anyone. 00:10:59.180 --> 00:11:02.000 And then you look back to it and you're like, "This is a mess," but, 00:11:02.422 --> 00:11:02.782 Frank Bach: Yeah. 00:11:02.850 --> 00:11:04.520 Jonny: you know, it w- it might mean something. 00:11:04.520 --> 00:11:05.920 It-- There might be some value in this, 00:11:06.616 --> 00:11:07.026 Frank Bach: Totally, 00:11:07.230 --> 00:11:07.710 Jonny: the fact 00:11:07.766 --> 00:11:11.006 Frank Bach: it, it led to work that was built so that, you know, it doesn't 00:11:11.016 --> 00:11:14.216 really matter what you made it in, it just, it, did it ship or did it not? 00:11:14.256 --> 00:11:14.876 And that's really 00:11:15.088 --> 00:11:15.158 Jonny: Yeah 00:11:15.556 --> 00:11:16.996 Frank Bach: the most important thing. 00:11:17.416 --> 00:11:22.186 I think we've, we've all spent probably a lot of time, you know, refining 00:11:22.656 --> 00:11:26.146 design components and things like this, and then you look at the, the actual 00:11:26.156 --> 00:11:29.696 production component library, and it looks nothing like the system you're 00:11:29.696 --> 00:11:32.646 working with in Figma, and, like, just trying to bridge that gap and 00:11:32.646 --> 00:11:34.346 get those two things more connected. 00:11:35.656 --> 00:11:38.016 yeah, it's just, it's totally changed the way I not just work, 00:11:38.016 --> 00:11:39.676 but think about the work, I guess. 00:11:40.826 --> 00:11:44.206 when I look at some of my sort of side projects and things that I've done 00:11:44.216 --> 00:11:49.056 for fun, it kinda was the way I would work anyway, but for some strange 00:11:49.056 --> 00:11:53.966 reason, I, I sort of separated my, the process of building my side projects 00:11:53.976 --> 00:11:57.216 was, like, very different from the way I would operate as a product designer 00:11:57.216 --> 00:11:58.936 within a company, if that makes sense. 00:11:59.826 --> 00:12:05.156 'cause I've, like, I have an e-comm store called Sunshine Shop, that was, you 00:12:05.156 --> 00:12:09.346 know, basically built on a W- WordPress WooCommerce template that I just hacked 00:12:09.346 --> 00:12:13.526 together and connected bits and pieces and print on demand and connected a 00:12:13.526 --> 00:12:16.926 shipping calculator and just sort of like, there's no Figma for that project. 00:12:17.480 --> 00:12:17.810 Jonny: Yeah 00:12:17.866 --> 00:12:18.526 Frank Bach: on the web. 00:12:18.556 --> 00:12:20.956 It was, like, straight to FTP upload, you know? 00:12:20.956 --> 00:12:23.556 Just like total cowboy mode, know? 00:12:24.918 --> 00:12:28.858 Tom: you mentioned at the end there, Frank, how you have recently been 00:12:28.938 --> 00:12:30.648 going back to Figma a bit, and 00:12:30.665 --> 00:12:30.955 Frank Bach: Yeah. 00:12:31.265 --> 00:12:31.555 Yeah 00:12:31.978 --> 00:12:36.128 Tom: Superfrank tool that you've built is something that you see yourself 00:12:36.128 --> 00:12:37.928 using in six months or 12 months. 00:12:37.928 --> 00:12:41.598 Is it something you're gonna continually evolve, or do you think it's more of 00:12:41.598 --> 00:12:45.988 a certain shape for a certain type of project, but actually it's just one tool 00:12:46.148 --> 00:12:49.798 and Figma is another tool and you'll use go back and forth between them regularly? 00:12:50.729 --> 00:12:55.119 Frank Bach: Yeah, I think I'll use my Frank tool as sort of an HTML prototype, 00:12:57.259 --> 00:13:01.309 tool, and then there's probably some ways to better connect 00:13:01.349 --> 00:13:03.899 the Figma files to that tool. 00:13:03.909 --> 00:13:05.349 I just haven't quite figured it out yet. 00:13:05.359 --> 00:13:07.019 They're kind of two separate paths. 00:13:08.129 --> 00:13:10.899 but for the nature of the work I'm doing, I'm on a growth team, so, you know, a 00:13:10.899 --> 00:13:13.919 lot of sort of A/B testing and like that. 00:13:14.649 --> 00:13:18.579 the Frank tool seems to be very, very good for, like, growth-type design. 00:13:19.719 --> 00:13:21.869 so yeah, I think I'll, I'll come back to it for sure. 00:13:22.719 --> 00:13:25.649 it's just, yeah, when, when those projects land, I'll sort of have to 00:13:25.649 --> 00:13:27.469 pick like, you know, path A or path B 00:13:28.790 --> 00:13:29.170 Tom: Yeah, totally. 00:13:30.700 --> 00:13:34.200 I think there's something else in the post that you shared that specifically 00:13:34.220 --> 00:13:37.780 talked about your mental model and having a tool that, that mapped 00:13:37.780 --> 00:13:39.640 to your mental model of things. 00:13:40.110 --> 00:13:44.480 was there anything about this tool that was very unusual compared to 00:13:44.480 --> 00:13:45.600 other tools in the market, or was 00:13:45.673 --> 00:13:45.923 Frank Bach: Yeah 00:13:46.100 --> 00:13:48.960 Tom: the amalgamation of features that was the, good thing? 00:13:49.833 --> 00:13:53.733 Frank Bach: I think the most unusual thing that I built, and I, I hope 00:13:53.743 --> 00:13:57.153 Figma steals this idea and somebody, you know, or somebody builds something 00:13:58.153 --> 00:14:02.023 this, was that I'm often, you know, if you're working in growth type

00:14:02.023 --> 00:14:05.813 projects, you've, you've got your, your control and then your, you know, 00:14:05.813 --> 00:14:09.223 your variants A, B, C, D, however many, you know, you might be run-- you could 00:14:09.223 --> 00:14:10.873 run two tests or 12 tests, who knows? 00:14:10.873 --> 00:14:11.803 Depends on the project. 00:14:13.943 --> 00:14:17.783 built a function where I could load up two prototypes at once, sort of 00:14:17.783 --> 00:14:22.803 like side by side split screen, and as I'm clicking either one of them, 00:14:22.813 --> 00:14:25.653 they're both progressing through the screens at the same time. 00:14:25.733 --> 00:14:29.383 And that became really helpful for sharing to design leadership of 00:14:29.383 --> 00:14:31.083 like, what's the current state? 00:14:31.093 --> 00:14:33.993 'Cause most people don't even know the current state of most products they're 00:14:33.993 --> 00:14:37.713 working on, 'cause we, you know, even when I worked at Headspace, like I was working 00:14:37.713 --> 00:14:41.443 on, on upsell conversion type screens, but everyone who works at Headspace 00:14:41.483 --> 00:14:44.533 doesn't have to pay for the app, so they haven't seen those screens in years. 00:14:44.923 --> 00:14:45.203 You know? 00:14:45.203 --> 00:14:47.783 So sort of like there's always this question of like, what is 00:14:47.783 --> 00:14:49.563 the current state of things? 00:14:49.613 --> 00:14:54.543 And so, being able to click through two prototypes sort of with one click and 00:14:54.543 --> 00:14:58.683 sort of give you this live view of like, here's what's changing on this screen, 00:14:58.683 --> 00:15:01.723 here's what's changing on this screen, here, you know, and, and give a feel and 00:15:01.723 --> 00:15:05.303 build up that confidence for people to agree, like, you know, give you the green 00:15:05.303 --> 00:15:08.003 light for that project felt very powerful. 00:15:08.603 --> 00:15:11.703 and that's probably one way that I'll keep, definitely keep using this tool. 00:15:13.023 --> 00:15:16.293 but yeah, incredible-- Like it was almost something that Claude did by accident. 00:15:16.323 --> 00:15:20.253 I, I, I sort of explained what I was roughly looking for and I was like, 00:15:20.503 --> 00:15:23.683 "Oh damn, this is like, this is even better than I thought it would be," 00:15:24.103 --> 00:15:27.703 which I think a lot of people are, are feeling, when they're using these, these 00:15:28.033 --> 00:15:28.213 Jonny: Mm. 00:15:29.034 --> 00:15:30.264 Tom: Yeah, absolutely. 00:15:30.333 --> 00:15:31.273 Jonny: accidents, yeah 00:15:31.470 --> 00:15:32.110 Frank Bach: Yeah, yeah 00:15:32.384 --> 00:15:36.354 Tom: one, one thing, thing we found with the guests we've been speaking 00:15:36.354 --> 00:15:40.004 to recently, everyone has a different interpretation of where the line is 00:15:40.004 --> 00:15:43.264 for what they want to retain as a human versus what they're happy to 00:15:43.753 --> 00:15:44.063 Frank Bach: Yeah 00:15:44.094 --> 00:15:44.794 Tom: the AI. 00:15:44.844 --> 00:15:48.564 but when I think about Figma, I think of it being very much a, like here are all 00:15:48.584 --> 00:15:50.024 the things at my disposal and I'm just 00:15:50.267 --> 00:15:50.697 Frank Bach: I see 00:15:50.824 --> 00:15:53.354 Tom: in the tool they need to craft the output that I want. 00:15:53.764 --> 00:15:57.834 But the world you're describing feels a bit more dynamic and interactive and 00:15:57.834 --> 00:16:03.694 fluid, and I could imagine that having potential to incorporate, I don't 00:16:03.694 --> 00:16:06.814 know, dynamic aspects of how customers are using the product right now. 00:16:07.044 --> 00:16:07.174 and 00:16:07.219 --> 00:16:07.549 Frank Bach: Yeah 00:16:07.704 --> 00:16:13.284 Tom: more about, yeah, that sort of user research side of, the sort 00:16:13.304 --> 00:16:15.824 of a multivariate testing side of, 00:16:15.981 --> 00:16:16.301 Frank Bach: Right 00:16:17.474 --> 00:16:21.094 Tom: so this is less about what your plans are specific with the tool, but more 00:16:21.094 --> 00:16:24.524 about you see the limits of this tool. 00:16:24.554 --> 00:16:25.664 Is that something that would 00:16:25.779 --> 00:16:26.079 Frank Bach: Yes 00:16:26.454 --> 00:16:30.284 Tom: in the r- world of possibility, or are you more on the like, I want to 00:16:30.284 --> 00:16:33.714 control the things and this is just a much nicer surface to control those things in? 00:16:34.583 --> 00:16:38.953 Frank Bach: I think there's a future where, like, just have these sort of mega 00:16:38.953 --> 00:16:42.823 tools that can do almost anything and you're, I don't know, you-- like, on a 00:16:42.833 --> 00:16:47.553 per user basis, you could sort of toggle on and off, what you're looking for. 00:16:47.553 --> 00:16:53.203 I wonder sometimes, like, you know, like, like Illustrator or Photoshop, kind of 00:16:53.523 --> 00:16:56.923 have been doing this for many years, where you've got different workspaces and 00:16:56.923 --> 00:16:59.913 they'll sort of pull up different features that you might need in that moment. 00:17:01.203 --> 00:17:07.023 but yeah, I, I could totally see a future tool that just has a live command line, 00:17:07.473 --> 00:17:10.843 you know, maybe along the bottom that you can just constantly type into and 00:17:10.843 --> 00:17:12.883 you're just reshaping the tool as you go. 00:17:12.883 --> 00:17:16.473 I don't know, maybe there's a, there's a big reset button at some point when 00:17:16.473 --> 00:17:20.683 you've gone too far and it's you've, you've created a, a, a monster. 00:17:21.173 --> 00:17:26.473 But, yeah, I do like this idea that, like, I can continuously evolve the tool 00:17:26.473 --> 00:17:33.443 for myself, almost like a shortcut into Claude, if that makes sense, the tool. 00:17:33.483 --> 00:17:34.983 I think that would be very, very fun. 00:17:35.643 --> 00:17:37.103 does everybody want to do that? 00:17:37.133 -->

00:17:37.743 I don't know. 00:17:37.753 --> 00:17:41.213 I mean, maybe it takes some special, like, unique people who wanna work in 00:17:41.213 --> 00:17:46.777 that way, but I also, like- Yeah, I think it, it opens up this big sort of 00:17:46.777 --> 00:17:53.527 existential question or for designers of, know, do we, sort of make, sort of 00:17:53.587 --> 00:17:57.367 just continue the, the path of making our own tools and what works for us and 00:17:57.367 --> 00:17:59.427 does that scale across design teams? 00:17:59.457 --> 00:18:00.007 I don't know. 00:18:00.357 --> 00:18:04.767 Does-- Or are we better off to just kind of continue, with these more 00:18:04.767 --> 00:18:08.317 enterprise tools that, that are, like, more one-size-fits-all type things? 00:18:09.607 --> 00:18:12.407 again, I, I-- Before working on Claude, I'd never worked… 00:18:12.477 --> 00:18:13.847 I'd mostly worked on mobile apps. 00:18:13.847 --> 00:18:18.467 I'd never worked on, software in this way, so I feel like I'm getting a crazy crash 00:18:18.467 --> 00:18:23.247 course in, like, not just designing tools, but, like, all I think about is tools now. 00:18:24.457 --> 00:18:29.217 and I think that's, like, a, a theme going back to my many different jobs 00:18:29.217 --> 00:18:33.147 before this was I think I, I, I've taken something and learned, like, learned 00:18:33.147 --> 00:18:37.787 something new that was very unique to that specific job with every role, and 00:18:37.797 --> 00:18:41.877 it's kind of like a, something that's important to me as I've been working 00:18:41.877 --> 00:18:45.667 through my career that I don't want to be a one-size-fits-all designer. 00:18:45.817 --> 00:18:49.287 Like, I've, I've met, you know, I've met folks who are, like, absolutely 00:18:49.287 --> 00:18:51.297 incredible at B2B dashboards. 00:18:51.297 --> 00:18:52.327 Like, that is their jam. 00:18:52.377 --> 00:18:53.287 They do it very well. 00:18:53.287 --> 00:18:54.327 They do this all day. 00:18:54.347 --> 00:18:55.377 It's all they want to do. 00:18:55.717 --> 00:18:56.457 That's amazing. 00:18:56.457 --> 00:18:57.297 I love that for them. 00:18:57.297 --> 00:19:03.427 But for me, I was like, I, I really like, variety in experiences, and so that's kind 00:19:03.427 --> 00:19:05.337 of why I went, you know, founded a studio. 00:19:05.807 --> 00:19:07.007 I went to Red Bull. 00:19:08.187 --> 00:19:10.827 when I worked at, Edenspiekermann, our main client was Red Bull. 00:19:10.837 --> 00:19:12.627 We were very much an integrated team. 00:19:13.617 --> 00:19:15.667 and then, you know, energy drinks to meditation. 00:19:15.727 --> 00:19:19.487 Doesn't make any sense, but, like, that was very exciting to me of like, okay, I 00:19:19.487 --> 00:19:23.747 want, like, whatever this is is fun, but I think I want the opposite now, you know? 00:19:24.387 --> 00:19:30.247 so I go meditation, and then, working on a meditation app, Headspace, we were 00:19:30.247 --> 00:19:34.487 focused a lot on kind of getting you off your phone and into the present moment. 00:19:34.517 --> 00:19:36.737 And then I went to social media, which you could argue is the 00:19:36.737 --> 00:19:39.257 total opposite of, like, how do we get you more in your phone? 00:19:40.437 --> 00:19:40.877 so I don't know. 00:19:40.877 --> 00:19:46.187 I j- I think I, I, I've become a bit of, like, a Swiss Army knife of, of design. 00:19:46.227 --> 00:19:48.817 Not to toot my own horn too much, but I, I really… 00:19:49.267 --> 00:19:50.557 I, I did this all on purpose. 00:19:50.557 --> 00:19:55.337 This was all very, intentional that I wanted to come out of this, I guess, 00:19:55.807 --> 00:20:01.051 out of this career having had a lot of different experiences and just, like I 00:20:01.091 --> 00:20:08.261 think I have a ge- a genuine interest in business and how, the various, like 00:20:08.261 --> 00:20:10.991 business structures and how companies make money, whether that's through 00:20:10.991 --> 00:20:15.931 subscription, through, you know, one-time fees, through advertising revenue. 00:20:15.941 --> 00:20:20.981 Like just learning, all of these different ways that, that businesses operate and, 00:20:21.031 --> 00:20:23.411 and you know, that you can earn a living. 00:20:23.441 --> 00:20:27.441 I think, you know, I, I don't know that this is the, end of my career, but 00:20:27.441 --> 00:20:31.931 I could see myself, compiling all of that information into like, you know, 00:20:31.931 --> 00:20:33.821 into whatever it is I do next, which I 00:20:33.961 --> 00:20:33.971 Jonny: Mm-hmm. 00:20:34.001 --> 00:20:34.571 Frank Bach: know what it is. 00:20:35.031 --> 00:20:36.631 I hope it's Claude for a very long time. 00:20:37.801 --> 00:20:42.751 but you know, I, I, I think that I could give advice to younger designers, 00:20:42.911 --> 00:20:47.251 folks who are maybe in school trying to sort of chart their path forward, I 00:20:47.251 --> 00:20:51.011 would, I would go for like variety and experiences 'cause I think it makes you, 00:20:51.801 --> 00:20:55.861 much more interesting, you're like kind of in those interview panels and you 00:20:55.861 --> 00:20:57.591 really stand out against other people 00:20:58.409 --> 00:20:58.769 Jonny: Mm-hmm. 00:21:00.389 --> 00:21:04.019 That's really interesting and, and, I recognize, I recognize 00:21:04.019 --> 00:21:08.449 that feeling of, I'm not sure it's getting bored easily, but it's, it 00:21:08.449 --> 00:21:10.149 is wanting to find something new. 00:21:10.170 --> 00:21:10.360 Frank Bach: Yeah 00:21:10.449 --> 00:21:14.309 Jonny: And often I'm envious of the people that are willing to just compound 00:21:14.309 --> 00:21:17.459 one thing over and over again and just, like, become a master at it. 00:21:17.669 --> 00:21:21.389 I do feel like, I don't know if it's zero sum, but, this sort of mastery 00:21:21.389 --> 00:21:26.329 versus invention paradigm, like either I want to spend a bit of time mastering

00:21:26.329 --> 00:21:29.779 something and then I actually just want to mess it all up and, like, 00:21:29.883 --> 00:21:30.443 Frank Bach: Right. 00:21:30.559 --> 00:21:32.159 Jonny: difficulty for myself once again. 00:21:32.163 --> 00:21:32.173 Frank Bach: Yeah 00:21:33.589 --> 00:21:37.699 Jonny: and y- you talking about, the, the Frank Mega tool or whatever 00:21:37.709 --> 00:21:41.229 we're calling it on this podcast, does it have a name or is it 00:21:41.687 --> 00:21:43.847 Frank Bach: I think it's called Frank Space. 00:21:44.729 --> 00:21:45.119 Jonny: space. 00:21:45.189 --> 00:21:45.609 Love it. 00:21:45.659 --> 00:21:46.069 Okay. 00:21:46.177 --> 00:21:46.377 Frank Bach: one. 00:21:46.467 --> 00:21:47.987 It's, it's gone through different… 00:21:48.057 --> 00:21:49.997 People at, at work call it different things. 00:21:50.027 --> 00:21:53.817 Then yeah, Frank Space, Bach Space, the Frank Mega Tool has definitely been thrown 00:21:53.961 --> 00:21:54.761 Jonny: Frank Megatool. 00:21:54.861 --> 00:21:55.261 Yeah. 00:21:56.031 --> 00:22:01.191 well Fr- Frank Megatool sort of, I, I, I wonder if that is the 00:22:01.191 --> 00:22:02.471 same thing on a smaller level. 00:22:02.471 --> 00:22:06.471 Like, you want to create something ephemeral that's, like, suits one 00:22:06.981 --> 00:22:10.101 problem, and then maybe a certain willingness to throw it, throw 00:22:10.171 --> 00:22:12.101 it away and, and not master it. 00:22:12.811 --> 00:22:15.351 and then create something new and, and find a new challenge and 00:22:15.371 --> 00:22:16.391 build something different for it. 00:22:17.267 --> 00:22:17.607 Frank Bach: Yeah 00:22:18.971 --> 00:22:22.811 Jonny: I love what you're saying about tools and, you know, how much do we 00:22:22.811 --> 00:22:27.911 master the tool given to us versus how much do we invent, a tool for ourselves. 00:22:29.691 --> 00:22:33.041 and so, yeah, definitely music to my ears. 00:22:33.151 --> 00:22:33.431 And 00:22:33.811 --> 00:22:34.121 Frank Bach: Yeah 00:22:34.381 --> 00:22:37.291 Jonny: are-- I think what we've learnt on this podcast over the 00:22:37.761 --> 00:22:41.021 16 episodes prior to this is that we're on the fence about Figma. 00:22:41.878 --> 00:22:41.908 Frank Bach: Yeah. 00:22:42.101 --> 00:22:42.441 Jonny: We, we, 00:22:43.148 --> 00:22:43.268 Frank Bach: Yeah 00:22:44.611 --> 00:22:48.861 Jonny: you know, that this is-- it's been at the center of, designer's 00:22:48.891 --> 00:22:53.301 world for, well, probably a decade now, not far off a decade actually. 00:22:53.301 --> 00:22:55.991 And then before that, there was, there was Sketch, and before 00:22:55.991 --> 00:22:57.041 that, there was Illustrator. 00:22:57.341 --> 00:22:59.321 You know, you go back and you-- There's, there's a, there's, 00:22:59.811 --> 00:23:00.911 someone sits on the throne. 00:23:01.883 --> 00:23:02.023 Frank Bach: Mm-hmm 00:23:02.441 --> 00:23:07.861 Jonny: and so to explore the possibility of there being more

than one tool and 00:23:07.861 --> 00:23:13.071 for us to be able to create tools for ourselves or for our teams, is exciting. 00:23:13.231 --> 00:23:18.161 I did want to ask you, as you think about the design industry in general, I know 00:23:18.161 --> 00:23:20.311 you, you, you run this, meetup in LA. 00:23:20.311 --> 00:23:23.731 You, y- you know probably more designers than, all of us put 00:23:23.731 --> 00:23:25.851 together, and, and speak to them. 00:23:26.671 --> 00:23:31.171 How much do you get a sense of this last couple of years changing, 00:23:32.981 --> 00:23:36.701 for everyone versus you joining Anthropic was a different culture, and 00:23:36.721 --> 00:23:38.901 therefore you work differently there? 00:23:39.161 --> 00:23:42.791 Like, are you seeing other people doing similar things elsewhere? 00:23:43.699 --> 00:23:44.009 Frank Bach: Yeah. 00:23:44.009 --> 00:23:49.969 So the-- will say even during my time at, at DoorDash, we kind of 00:23:49.999 --> 00:23:52.339 entered the transition into AI tools. 00:23:52.339 --> 00:23:56.149 The company had this big mission, you know, like, desi- they called 00:23:56.149 --> 00:23:58.149 it, I think Design 2028 or something. 00:23:58.149 --> 00:24:03.819 As by 2028, everyone who's a designer at DoorDash is, is 00:24:03.819 --> 00:24:05.169 basically a design engineer. 00:24:05.199 --> 00:24:08.739 You know, we need to lean into this, and the design leadership there was amazing 00:24:09.369 --> 00:24:13.189 to, you know, to give us the time and space to do that, 'cause otherwise I 00:24:13.189 --> 00:24:16.359 would, I would've had to learn it, you know, not on the company dime, I guess. 00:24:16.389 --> 00:24:16.629 You know? 00:24:16.629 --> 00:24:18.099 So th-thank goodness for that. 00:24:18.099 --> 00:24:18.869 I'm very grateful. 00:24:20.359 --> 00:24:23.709 and yeah, at, at first I definitely complained about it. 00:24:23.709 --> 00:24:26.789 I was, "Ah, I don't feel like, you know, I don't wanna learn a new tool. 00:24:26.809 --> 00:24:27.599 Learning new things is 00:24:27.631 --> 00:24:27.951 Jonny: Mm. 00:24:28.189 --> 00:24:32.239 Frank Bach: You know, kind of the, the human sort of complacency that, 00:24:32.239 --> 00:24:35.589 you know, s-sometimes it's okay that a job is just a job and it 00:24:35.589 --> 00:24:39.139 doesn't have to be the most exciting, incredible thing you're doing that day. 00:24:40.219 --> 00:24:42.639 so I definitely like leaned into that, like a little bit of a 00:24:42.659 --> 00:24:46.059 negative mindset around it, and I was, you know, was venting to 00:24:46.059 --> 00:24:47.859 other coworkers about this too and 00:24:49.969 --> 00:24:51.709 organizing these meetups in LA. 00:24:51.719 --> 00:24:57.809 We do them every quarter and, the kind of backstory around this is, we can talk a 00:24:57.809 --> 00:25:03.219 little bit more about it in a second, but I see it as a good forcing of function for 00:25:03.219 --> 00:25:07.909 myself to get out of

the house and talk to people for like sort of work-ish reasons. 00:25:08.819 --> 00:25:11.559 you know, I'm a, I'm a father of a six-year-old, I have a wife 00:25:11.609 --> 00:25:14.689 and, you know, you know, stay very busy with family stuff. 00:25:15.349 --> 00:25:19.229 And so to have this thing that doesn't feel like just an irresponsible, you 00:25:19.229 --> 00:25:23.249 know, adventure to the bar by myself sort of thing, you know, I, I sort of 00:25:23.279 --> 00:25:25.269 have this work-related reason to do this. 00:25:25.279 --> 00:25:29.849 So very s- very selfishly, set up these meetups every quarter and thankfully 00:25:29.849 --> 00:25:31.509 people keep showing up to them. 00:25:32.419 --> 00:25:37.809 but I've definitely caught that trend of there's sort of the, sort of a 50/50 00:25:37.809 --> 00:25:40.839 split of the people who are really excited and they can't believe that 00:25:40.839 --> 00:25:43.959 we're like really doing this, and then there's the people who are kind of 00:25:43.969 --> 00:25:47.229 leaning more sort of classic UX-y type 00:25:47.471 --> 00:25:47.991 Jonny: Mm-hmm. 00:25:48.119 --> 00:25:49.309 Frank Bach: are, is very valuable. 00:25:50.389 --> 00:25:54.649 I try not to spend a lot of time on the UX design Reddit that feels 00:25:54.649 --> 00:25:56.909 like a place where there's a lot of complaining about the way our 00:25:56.909 --> 00:26:01.639 industry is headed, and I don't think it's, an all or nothing situation. 00:26:01.659 --> 00:26:06.899 I think there are ways you can, you can, you know, amplify and, 00:26:06.919 --> 00:26:10.739 and speed up the research process and the iteration process. 00:26:11.569 --> 00:26:15.759 not everything needs to be a like one-shotted, you know, 00:26:15.759 --> 00:26:17.519 very vanilla bland prototype. 00:26:17.539 --> 00:26:21.769 I think that you can make some really incredible things and actually, Yeah, I 00:26:21.779 --> 00:26:27.299 think there's this like, this storyline around like using AI tools means you're 00:26:27.299 --> 00:26:30.689 not doing research or something, or you're like, you're, you're skipping a lot of 00:26:30.689 --> 00:26:32.489 the steps that you're supposed to take. 00:26:34.039 --> 00:26:37.759 yeah, I don't, I don't, I don't believe in that, I guess is, is where I'm headed. 00:26:38.909 --> 00:26:42.339 but I've definitely had a shift in the way I thought about these tools and about 00:26:42.339 --> 00:26:45.129 this process and now the, like the… 00:26:46.059 --> 00:26:50.519 And I, and I'm seeing that in the meetups, and also just, yeah, forcing myself to 00:26:50.519 --> 00:26:54.709 talk to people and get, sort of read the room and get a pulse of the industry and 00:26:54.709 --> 00:26:56.159 where are people's heads at with this. 00:26:56.159 --> 00:27:01.779 And I mean, I don't know if it is, if people have just been forced into 00:27:01.789 --> 00:27:03.889 having to adopt the tools at this point. 00:27:04.929 --> 00:27:08.989 I think there is some truth to like if you're not adopting these tools, 00:27:09.619 --> 00:27:13.549 it will be really hard for you to be employed in the f- in the future. 00:27:14.699 --> 00:27:16.249 and that is not a, 00:27:18.707 --> 00:27:21.567 Yeah, I, I, I think that's just the reality of the situation. 00:27:21.567 --> 00:27:22.677 It's not design only. 00:27:22.717 --> 00:27:25.307 I think, like I, I have friends who are accountants and lawyers 00:27:25.307 --> 00:27:30.337 and doctors and, they are all using AI tools to, you know, to speed up 00:27:30.337 --> 00:27:32.267 their processes or, or save time. 00:27:32.267 --> 00:27:36.697 And, you know, personally, I did my-- Last year I did my, my income tax 00:27:36.747 --> 00:27:39.157 with, with an AI, AI tool, you know. 00:27:39.157 --> 00:27:43.157 And like I used to spend, you know, two-- I used to spend a whole weekend 00:27:43.207 --> 00:27:46.907 organizing my receipts and my, you know, making sure all my filings 00:27:46.907 --> 00:27:48.757 were, were proper and stuff like that. 00:27:48.757 --> 00:27:52.137 And now I can just send an AI tool to filter through my Dropbox and 00:27:52.137 --> 00:27:53.977 just like, just see it go to work. 00:27:53.977 --> 00:27:57.827 And I'm just like, "My goodne-" Like I, I, I trust that it did the right thing. 00:27:57.827 --> 00:27:58.517 I hope it's right. 00:27:58.967 --> 00:28:00.157 I'll find out later if it wasn't. 00:28:00.627 --> 00:28:04.347 But, probably did a better job than I would've myself, you know. 00:28:04.347 --> 00:28:09.307 So I'm like, do I really wanna spend two days on a Saturday and Sunday doing 00:28:09.307 --> 00:28:14.177 my taxes when I could probably do it in 45 minutes just prompting an AI tool? 00:28:14.227 --> 00:28:19.237 So again, that's like long story around this, but, I think design is 00:28:19.237 --> 00:28:22.727 kind of going through that right now, and sounds like you guys are, are 00:28:22.727 --> 00:28:24.327 kind of hearing and seeing similar. 00:28:25.447 --> 00:28:25.707 Yeah. 00:28:26.688 --> 00:28:27.578 Tom: Yeah, very much so. 00:28:28.638 --> 00:28:36.598 it seems like it does-- There is this, not divide, but there are companies who 00:28:36.598 --> 00:28:40.098 are very AI, AI-driven, AI first, and 00:28:40.243 --> 00:28:40.533 Frank Bach: Yeah 00:28:40.818 --> 00:28:44.178 Tom: there are a bunch of companies that, that is not, maybe they're 00:28:44.178 --> 00:28:45.448 playing catch up or they're not 00:28:45.587 --> 00:28:45.987 Frank Bach: Yes 00:28:46.298 --> 00:28:48.318 Tom: enough to be investing as much there. 00:28:49.238 --> 00:28:52.558 So I think it just, I guess it leads to a spectrum of how far 00:28:52.558 --> 00:28:54.138 along people are individually. 00:28:54.528 --> 00:28:58.588 we run a, a meetup here in London, monthly-ish meetup for 00:28:58.779 --> 00:28:59.099 Frank Bach: Nice 00:28:59.528 --> 00:29:02.978 Tom: specifically focused around AI as sort of

an extension 00:29:03.098 --> 00:29:04.508 of this podcast in some way. 00:29:05.728 --> 00:29:07.878 and so yeah, that has been quite, quite 00:29:08.083 --> 00:29:08.413 Frank Bach: Yeah. 00:29:08.458 --> 00:29:09.998 Tom: get a glimpse into how people's 00:29:10.243 --> 00:29:10.543 Frank Bach: Yeah 00:29:10.608 --> 00:29:14.118 Tom: changing and how, how the tools that they're all, everyone's working with. 00:29:14.838 --> 00:29:18.718 I did actually have a question with you, for you about, mentioned, 00:29:18.718 --> 00:29:22.148 I think in another post about weird design processes and how, 00:29:22.195 --> 00:29:22.515 Frank Bach: yeah 00:29:22.578 --> 00:29:27.258 Tom: how processes should be, should be weirder or that that's an interesting 00:29:27.258 --> 00:29:28.348 topic of conversation for you. 00:29:28.348 --> 00:29:32.498 I'm curious, what is the, the weirdest design process you've come across? 00:29:32.518 --> 00:29:37.678 And whether that's someone at the meetup, someone at Anthropic, yourself, yeah, 00:29:37.718 --> 00:29:39.288 curious if that sparks anything for you. 00:29:40.307 --> 00:29:43.537 Frank Bach: Yeah, I think there's so many different ways to do this 00:29:43.567 --> 00:29:46.117 now, and there-- I don't think that there's a right or wrong way. 00:29:46.457 --> 00:29:49.737 I mean, I've met designers who are, they're just-- 00:29:49.917 --> 00:29:51.037 they're basically engineers. 00:29:51.037 --> 00:29:54.407 They're just straight in working it, you know, working on staging and 00:29:54.407 --> 00:29:57.767 pushing it to production, and there's no Figma file that even exists. 00:29:58.217 --> 00:30:01.697 They are just doing the thing that, that engineers have been doing for years. 00:30:03.287 --> 00:30:07.497 then there's folks who are like I, I've-- I'm hearing a lot of the, you know, 00:30:07.497 --> 00:30:10.917 kind of the Whisper tool, where you're, you're spending time basically just 00:30:11.247 --> 00:30:15.487 ch- talking into a microphone, telling, telling the tool what to do all day long. 00:30:15.857 --> 00:30:17.467 Sounds pretty tiring on the voice. 00:30:17.467 --> 00:30:19.087 I don't know if I'd want to do that all day long. 00:30:19.097 --> 00:30:23.787 But, I'm also seeing that, like, because of prompting, my, like, 00:30:23.787 --> 00:30:27.627 my, you know, we'll probably end up with carpal tunnel syndrome at some 00:30:27.627 --> 00:30:30.147 point 'cause we're typing probably a lot more than we used to, right? 00:30:31.167 --> 00:30:33.897 so that's, like, very, you know, there's, like, some, some 00:30:33.927 --> 00:30:37.717 ergonomics, thing going on there. 00:30:39.637 --> 00:30:46.127 I think the, the strangest or most interesting, process I've seen is, people 00:30:46.127 --> 00:30:54.407 having, sort of having the, the tools run research or run, try to build themselves, 00:30:55.197 --> 00:30:57.307 you know, in the off hours of, you know. 00:30:57.307 --> 00:31:01.787 So you're,

you sort of wake up in the morning to that your AI tool has, has 00:31:01.807 --> 00:31:05.657 given you of, like, suggestions for the day on what you might do today. 00:31:05.917 --> 00:31:09.057 And, like, at what point does that loop fully close, where you're just sort of a 00:31:09.057 --> 00:31:12.747 human observing the thing building, and you're like, you know, you just give it 00:31:12.747 --> 00:31:14.277 the thumbs up or thumbs down, you know? 00:31:14.277 --> 00:31:18.007 I think it's great to always have a person in control of these things, but 00:31:18.407 --> 00:31:20.357 it is getting pretty weird pretty fast. 00:31:20.357 --> 00:31:23.577 And, like, I, I'm discovering new things every day working with folks of, like, 00:31:24.347 --> 00:31:28.417 a product manager of mine who, you know, who's, who's running agents, doing all 00:31:28.427 --> 00:31:31.967 these other things and, like, come, comes back with a ton of research and 00:31:31.967 --> 00:31:35.797 insights that, like, you didn't-- It would have normally been a two-week 00:31:35.797 --> 00:31:37.727 sprint to even get to that point, right? 00:31:37.727 --> 00:31:41.947 And, like, so you're just sort of-- I'm finding myself doing a lot of, like, 00:31:42.037 --> 00:31:47.347 future predicting the work that I need to do before I even get to the work. 00:31:47.367 --> 00:31:51.807 So I'm sort of, like, getting my tools to-- that once I get to the work, like 00:31:51.807 --> 00:31:57.017 let's say, we're, a Tuesday today, like, by-- I'm gonna, I'm gonna do 00:31:57.017 --> 00:32:00.827 some pre-work so that by Monday, that next project is kind of queued up 00:32:01.127 --> 00:32:02.597 and I'm not starting from scratch. 00:32:03.897 --> 00:32:07.117 even if, like, half the stuff is, is throwaway, at least I've got, 00:32:07.117 --> 00:32:08.837 like, a good foundation to begin on. 00:32:10.097 --> 00:32:11.067 and there's, there's this, 00:32:11.068 --> 00:32:11.358 Tom: thing 00:32:12.317 --> 00:32:12.957 Frank Bach: what's the quote? 00:32:12.977 --> 00:32:18.583 There's a, There's a quote by John Cage, j- it's simple. 00:32:18.583 --> 00:32:19.913 It's just, "Begin anywhere." 00:32:20.353 --> 00:32:24.073 And I think that, like, I, that used to be sort of a, I don't know, a little bit 00:32:24.073 --> 00:32:27.443 of, like, a mindfulness reminder that I would give myself, and it's definitely 00:32:27.473 --> 00:32:31.513 coming, coming true more and more in my design process of, like, it actually 00:32:31.513 --> 00:32:32.813 doesn't really matter where you start. 00:32:32.863 --> 00:32:33.253 It's so 00:32:33.361 --> 00:32:33.841 Jonny: Right 00:32:33.883 --> 00:32:38.103 Frank Bach: to build, to build things and, and give something to react to, to see, to 00:32:38.103 --> 00:32:44.563 click, to feel, that just start anywhere and, figure out the details as you go. 00:32:44.563 --> 00:32:49.993 And I think I've been a, a big proponent of, like, you can do the 00:32:49.993 --> 00:32:55.113 research upfront and sort of de-risk the design, and that's, that's fine. 00:32:55.933 --> 00:33:00.033 you can also build the thing and then learn from the results or, 00:33:00.503 --> 00:33:03.453 or lack of results that you maybe built the wrong thing, or maybe 00:33:03.453 --> 00:33:05.033 there's it- iteration you can do. 00:33:05.033 --> 00:33:06.813 I think just don't stand still. 00:33:06.833 --> 00:33:08.263 Just start somewhere and do something. 00:33:08.923 --> 00:33:12.783 And, you know, builders, like the f- you know, cliche maybe, but I 00:33:12.793 --> 00:33:16.753 think the future is, of design is, is for builders and people who make 00:33:16.753 --> 00:33:21.283 stuff, stuff that lives and breathes in the real world, not just sits in 00:33:21.283 --> 00:33:24.133 our drafts, and PRDs for, you know. 00:33:24.853 --> 00:33:28.523 yeah, that's, that's my, my soapbox for the day there. 00:33:29.537 --> 00:33:30.447 Jonny: It's, it's a good one. 00:33:30.747 --> 00:33:31.357 It's a good one. 00:33:31.827 --> 00:33:34.977 W- I think we, we often preach something similar, which is, 00:33:35.857 --> 00:33:37.387 which is just to, to get started. 00:33:37.397 --> 00:33:40.547 The only way to understand how to do this stuff is to just do it. 00:33:41.137 --> 00:33:41.467 so, 00:33:43.557 --> 00:33:47.277 And, and the cost to delete anything is, like, I-- To me, the one that, that gets 00:33:47.277 --> 00:33:52.557 me is, like, it's cheap to create, but also because it was so cheap to create, 00:33:52.617 --> 00:33:54.327 I f- I feel nothing for it, you know? 00:33:54.817 --> 00:33:58.527 Whereas one of the, the classic designer tropes is falling in love with your design 00:33:58.527 --> 00:34:03.757 and then defending it and being blind to-- i-in the face of evidence to the contrary. 00:34:04.157 --> 00:34:04.467 s- 00:34:05.193 --> 00:34:05.913 Frank Bach: Absolutely. 00:34:06.133 --> 00:34:07.373 I, I'll, I'll add to that. 00:34:07.403 --> 00:34:12.103 I had a moment, I won't name the product, but something that I worked on years ago. 00:34:13.183 --> 00:34:15.823 they launched a new feature, and I was, I was pretty vocal 00:34:15.823 --> 00:34:17.973 in, in critiquing this feature. 00:34:18.423 --> 00:34:21.093 I think I felt a little bit passionate by it because it was something 00:34:21.093 --> 00:34:25.653 that I, I wished we had built during my time at this company. 00:34:26.423 --> 00:34:30.813 And then I heard from a, an old colleague who worked on the feature who wasn't very 00:34:30.813 --> 00:34:32.993 happy that I was so critical of the work. 00:34:33.393 --> 00:34:36.713 And, you know, sort of my-- maybe I've, I'm a little bit too hardcore in 00:34:36.713 --> 00:34:39.423 my sort of business sense these days. 00:34:40.343 --> 00:34:43.903 but I, you know, if it doesn't produce results, it doesn't really matter. 00:34:45.063 --> 00:34:47.283 you know, if we're gonna go har- like sup- you know, of course I'm

00:34:47.283 --> 00:34:49.913 an ar- also a creative and artistic person, and I think it's valuable 00:34:49.913 --> 00:34:53.493 to build things for fun and joy and, and just for the sake of building. 00:34:53.493 --> 00:34:58.273 But if we're here and we're measuring success and we have metrics to, 00:34:58.303 --> 00:35:02.873 to, to measure against, then if the thing you built didn't deliver 00:35:02.873 --> 00:35:07.253 that, it was kind of a huge waste of time, and I don't know, not saying, 00:35:07.483 --> 00:35:08.403 you should be ashamed of your… 00:35:08.433 --> 00:35:08.903 I'm just kidding. 00:35:09.763 --> 00:35:13.743 But it's like, it's like it didn't do the thing you wanted it to do, so you should 00:35:13.753 --> 00:35:17.813 be-- we sh- we're all adults here and we can talk about the failures and successes. 00:35:18.283 --> 00:35:21.683 Anyway, this colleague of, this colleague of mine back and said, 00:35:21.683 --> 00:35:24.403 "I didn't really appreciate you were so vocal of, about this. 00:35:24.723 --> 00:35:27.603 A lot of people worked really hard on this." 00:35:27.753 --> 00:35:32.073 And it was sort of like, yes, I'm a human and I'm empathetic, and I understand that. 00:35:32.083 --> 00:35:35.433 That is, that, that makes a lot of sense and, you know, sort of maybe I was a 00:35:35.433 --> 00:35:37.093 little too harsh with my criticism. 00:35:38.403 --> 00:35:41.263 just the fact that a lot of people put time and effort into something 00:35:41.263 --> 00:35:45.243 doesn't mean we should just praise it, it, know, because they did. 00:35:45.293 --> 00:35:45.523 Like, 00:35:45.735 --> 00:35:46.055 Jonny: Right 00:35:46.093 --> 00:35:47.373 Frank Bach: doesn't really make sense. 00:35:48.773 --> 00:35:50.633 so again, I, I sort of took it back. 00:35:50.633 --> 00:35:53.523 I was a little harsh, but, you guys get what I'm saying. 00:35:53.623 --> 00:35:53.713 It's 00:35:53.911 --> 00:35:54.311 Jonny: Yeah 00:35:54.413 --> 00:35:56.703 Frank Bach: didn't do, it didn't do what it needed to do, and we're all 00:35:56.703 --> 00:36:01.363 working in tech and startups, and many of us have equity in the companies we 00:36:01.363 --> 00:36:04.543 worked on in the past, and we wanna see them be successful, and we wanna, 00:36:04.863 --> 00:36:09.493 you know, we wanna see them have their liquidity event or whatever that might be. 00:36:10.483 --> 00:36:14.983 and like, yeah, if we're-- lot of designers talk big game about, you know, 00:36:16.383 --> 00:36:19.603 product management and doing right by the business and things, and it's like, 00:36:19.603 --> 00:36:22.973 well, if you're gonna talk like that, let's like, let's talk business, and 00:36:22.973 --> 00:36:25.943 there's not, unfortunately, there's not a lot of emotion in business. 00:36:26.665 --> 00:36:26.975 Jonny: Yeah. 00:36:27.403 --> 00:36:28.123 Frank Bach: but yeah, so 00:36:28.375 --> 00:36:31.025 Jonny: Yeah, and the, the, the market doesn't care basically.

00:36:31.265 --> 00:36:33.275 The market doesn't care about your feelings, yeah. 00:36:33.415 --> 00:36:34.775 Frank Bach: it doesn't, it doesn't. 00:36:34.855 --> 00:36:36.925 And there's a lot of other places where we can care about 00:36:36.925 --> 00:36:38.155 feelings, and that's wonderful. 00:36:38.445 --> 00:36:38.715 But 00:36:38.923 --> 00:36:38.973 Jonny: Yeah 00:36:39.135 --> 00:36:41.325 Frank Bach: just gonna be talking strictly business, that's the… 00:36:41.325 --> 00:36:43.005 I, I think it's, it's… 00:36:43.205 --> 00:36:47.625 It shows the sense of maturity, I think, in a designer who can, you know, who 00:36:47.625 --> 00:36:49.875 can kind of be blunt about that stuff 00:36:52.092 --> 00:36:55.502 Jonny: quite active posting jobs, on LinkedIn. 00:36:55.752 --> 00:37:03.572 for folks that find themselves in a non, AI-pilled organization or a non, 00:37:03.592 --> 00:37:09.692 you know, unlimited tokens, you go for it situation or a, a situation where 00:37:09.972 --> 00:37:13.042 maybe their manager and their manager's manager aren't that interested in 00:37:13.042 --> 00:37:14.502 this and there's not much guidance. 00:37:16.202 --> 00:37:19.852 I, I suppose I'm just interested in like what, what advice you'd give or 00:37:19.852 --> 00:37:23.692 guidance around just, making progress. 00:37:23.712 --> 00:37:26.262 As you said before, you know, if, if DoorDash hadn't 00:37:26.262 --> 00:37:26.962 paid for-- Was it DoorDash? 00:37:27.742 --> 00:37:28.002 I can't remember. 00:37:28.142 --> 00:37:29.982 If DoorDash hadn't paid for your tokens, you'd have had to 00:37:29.982 --> 00:37:31.092 have paid for them yourself. 00:37:31.092 --> 00:37:31.312 But, 00:37:31.684 --> 00:37:32.214 Frank Bach: Yes. 00:37:32.314 --> 00:37:32.634 Yes 00:37:32.932 --> 00:37:35.532 Jonny: that's non-obvious maybe to some people that that might 00:37:35.532 --> 00:37:37.672 be-- This might be existential. 00:37:37.682 --> 00:37:41.112 So, just yeah, any, any thoughts for people listening who 00:37:41.112 --> 00:37:42.152 might be in that situation? 00:37:43.172 --> 00:37:46.102 Frank Bach: Yeah, I think-- I don't think this is a very, like an… 00:37:46.162 --> 00:37:50.972 I think the tools have changed, but I think the concept of, folks who can find 00:37:50.972 --> 00:37:55.372 the time to create like passion side projects or solve their own problems, 00:37:56.022 --> 00:38:00.792 think that is usually a very standout project in your sort of, you know, 00:38:01.042 --> 00:38:03.512 sort of case study portfolio share out. 00:38:04.552 --> 00:38:07.442 you know, typical design interview loops, you'd probably 00:38:07.442 --> 00:38:09.002 share two major case studies. 00:38:09.422 --> 00:38:14.642 But I, I, I, I would really recommend folks who are, you know, who are 00:38:14.642 --> 00:38:19.842 interested in, in AI tools to make a thing of their own, solve their own problem. 00:38:21.292 --> 00:38:25.362 you know, it could be as simple as a grocery, you know, grocery list 00:38:25.362 --> 00:38:30.342 application or something, or I don't know, maybe there's, you're, if 00:38:30.342 --> 00:38:34.232 you're passionate about music, maybe you can make, you know, work on some 00:38:34.232 --> 00:38:39.602 like recording software or tool or a, a plugin for, you know, like a 00:38:39.612 --> 00:38:41.742 digital guitar amplifier or something. 00:38:41.742 --> 00:38:41.892 I don't know. 00:38:41.902 --> 00:38:45.662 There's just so many things you can build so quickly, even if it's imperfect. 00:38:46.842 --> 00:38:52.282 and that just shows that you have, an understanding, a curiosity, 00:38:52.342 --> 00:38:54.032 willingness to sort of build. 00:38:54.042 --> 00:39:00.492 Your-- You maybe haven't, implemented the AI process into your, your existing 00:39:00.492 --> 00:39:04.572 case studies, but you've shown that you have, you know, you've got that interest 00:39:04.572 --> 00:39:10.572 and that you, you are excited about this and that, yeah, you will be in whatever 00:39:10.622 --> 00:39:14.542 new team or new company you join, you will be kind of a net positive sort 00:39:14.542 --> 00:39:16.942 of AI influencer within the company. 00:39:18.402 --> 00:39:21.782 so I, I would, I would, I guess, yeah, take, sort of take it into your own 00:39:21.782 --> 00:39:26.412 hands and don't just wait for your manager to tell you, how to do things 00:39:26.452 --> 00:39:30.712 and that, you know, it might, I don't know, it-- depending on the job and 00:39:30.712 --> 00:39:35.932 the situation, trying to push these tools internally might, you know, 00:39:35.972 --> 00:39:40.002 I know s- some companies maybe are, are, are pushing back against it. 00:39:40.062 --> 00:39:40.572 I'm not sure. 00:39:40.572 --> 00:39:45.062 I, I don't hear too much about that these days, but, there's definitely some 00:39:45.062 --> 00:39:50.752 friction, I think, with, you know, w- within companies of, of how hard, they're 00:39:50.752 --> 00:39:53.892 feeling the push to, adopt these tools. 00:39:54.562 --> 00:39:55.512 but I think it's natural. 00:39:55.512 --> 00:39:58.832 I think it's just, it's happening so quickly and, you know, when you, when 00:39:58.832 --> 00:40:03.232 people feel like they have to learn something new, they feel, they, they 00:40:03.232 --> 00:40:06.792 don't quite understand, that's a really normal response to have, you know. 00:40:07.642 --> 00:40:07.962 Jonny: Yeah. 00:40:08.212 --> 00:40:08.522 Yeah. 00:40:09.188 --> 00:40:09.498 Frank Bach: Yeah 00:40:09.642 --> 00:40:12.362 Jonny: Yeah, so i-i- there's a bit of fear. 00:40:12.362 --> 00:40:14.432 I think there's, there's a bit of fear going around. 00:40:15.982 --> 00:40:19.902 but it also feels like we had sort of, I don't know, Tom, if you'd agree with 00:40:19.902 --> 00:40:25.432 this from, from work we've been doing, but we had like this time last year, peak 00:40:26.082 --> 00:40:31.682 of excitement at the possibility of, of what

AI might, might do for, for us in… 00:40:32.672 --> 00:40:37.852 And then, maybe a bit of doomerism, I don't know, like Claude Design comes out. 00:40:37.862 --> 00:40:41.532 This is obviously before you were at Anthropic maybe, but, but suddenly 00:40:41.532 --> 00:40:43.532 it's like, "Oh my God, are we doomed?" 00:40:44.312 --> 00:40:47.272 and then it feels like now we're in this sort of slightly more nuanced 00:40:47.282 --> 00:40:50.642 place where we're spotting the places where, where it works and we're 00:40:50.642 --> 00:40:54.452 spotting where it's useless and maybe where it might always be useless. 00:40:55.642 --> 00:40:59.672 or it-- there might always be a role for us humans, not just as assistants 00:40:59.672 --> 00:41:03.992 to, you know, get it some more tokens, but, to actually p-play a role in, 00:41:04.062 --> 00:41:04.302 Frank Bach: Yeah. 00:41:04.572 --> 00:41:05.182 Jonny: in design. 00:41:05.482 --> 00:41:05.682 Frank Bach: No. 00:41:06.742 --> 00:41:07.022 Yeah. 00:41:07.872 --> 00:41:10.522 Yeah, I'm definitely-- I, I'm opt- I'm optimistic about that. 00:41:10.522 --> 00:41:17.212 I think, I think, of, to the, the individual choice of how deep do you wanna 00:41:17.222 --> 00:41:20.182 go with this, and where do you see it fit? 00:41:20.272 --> 00:41:24.872 And, something that really struck me was, I think there was an article that 00:41:24.872 --> 00:41:29.867 came out, and it, it wasn't-- It, it was, I think, a, a different tool But 00:41:29.867 --> 00:41:35.132 it was a bunch of, of, mathematicians who were quite upset an AI tool was 00:41:35.132 --> 00:41:36.702 being used to solve a math problem. 00:41:36.842 --> 00:41:38.922 And I thought, "You know, that's really interesting," 'cause I would've-- 00:41:38.992 --> 00:41:42.312 I'm definitely not passionate about math, and I would use I would use 00:41:42.312 --> 00:41:46.412 an AI tool to solve a math problem ev-ev- any day, anytime I could. 00:41:47.622 --> 00:41:51.042 but you know, that's me, and I'm, you know, I, I have, I have that 00:41:51.272 --> 00:41:54.952 autonomy and freedom to use the tool how I sort of see fit that works 00:41:54.952 --> 00:41:56.592 for me and what makes me happy. 00:41:57.902 --> 00:42:02.042 you know, if I'm creating art or making music, probably won't use an 00:42:02.052 --> 00:42:06.632 AI tool because that's a space that I've, I've decided, I wanna have, 00:42:06.632 --> 00:42:09.232 like, m- my full expression in. 00:42:10.042 --> 00:42:13.362 but I'm not gonna judge somebody who, who uses an AI tool to 00:42:13.382 --> 00:42:15.552 sorta amplify their creativity. 00:42:16.362 --> 00:42:17.592 that's up to them, you know? 00:42:17.742 --> 00:42:22.072 But also it's up, it, it's up to the audiences and people to react 00:42:22.122 --> 00:42:23.552 to that artwork in whatever way they 00:42:23.618 --> 00:42:23.928 Jonny: Right 00:42:24.032 --> 00:42:24.272 Frank Bach: you know? 00:42:24.272 --> 00:42:29.412 And I think art

is a touchy place, for these tools, especially these, 00:42:30.512 --> 00:42:33.822 I think there was something around Spotify and Suno and all these things. 00:42:33.822 --> 00:42:34.882 You hear a lot about this. 00:42:35.612 --> 00:42:37.072 this is not, not new information. 00:42:37.072 --> 00:42:38.212 It's all in the news everywhere. 00:42:38.212 --> 00:42:43.512 But, yeah, I think mu- music and art is a, is a touchy place for, for that stuff, 00:42:43.512 --> 00:42:46.382 so it'll be interesting to see w- how that kind of shakes out in the future. 00:42:47.184 --> 00:42:50.384 Tom: So we have, a few questions we like to ask guests. 00:42:51.194 --> 00:42:53.714 short question, short-ish answer, short as 00:42:53.800 --> 00:42:54.020 Frank Bach: Cool 00:42:54.484 --> 00:42:58.244 Tom: and we'll ans- ask maybe one or maybe one or two each, depending on 00:42:58.244 --> 00:42:59.644 how quickly we are, and then we'll 00:42:59.722 --> 00:42:59.942 Frank Bach: Sure. 00:43:00.224 --> 00:43:00.684 Tom: after that. 00:43:01.154 --> 00:43:05.444 So question one, AGI takes over everything, what will you do 00:43:05.484 --> 00:43:06.554 with all of your free time? 00:43:08.942 --> 00:43:10.082 Frank Bach: I'll probably still make things. 00:43:10.082 --> 00:43:17.172 I, I've always loved making, little magazines and, and flyers for music stuff. 00:43:17.812 --> 00:43:18.422 I think I'll, I'll… 00:43:18.722 --> 00:43:20.992 It'll just maybe be a different, slightly different process. 00:43:21.762 --> 00:43:22.092 but I'll… 00:43:22.122 --> 00:43:24.042 I've always been someone who liked creating things, 00:43:24.042 --> 00:43:25.182 even if it was just for fun. 00:43:25.192 --> 00:43:29.952 So even, you know, if money, if money were no issue, I will still make things. 00:43:29.992 --> 00:43:33.772 I mean, even hopefully it all-- If it does happen, it all happens around the 00:43:33.772 --> 00:43:37.502 time I'm ready to retire, and then the, the outcome will be the same either way. 00:43:37.502 --> 00:43:38.322 It doesn't really matter. 00:43:40.398 --> 00:43:40.708 Tom: True 00:43:42.902 --> 00:43:45.892 Jonny: Still, still, yeah, 20 years away for us maybe. 00:43:46.402 --> 00:43:48.422 AGI can hold off just for a, for a hot second. 00:43:48.666 --> 00:43:48.926 Frank Bach: yeah. 00:43:49.006 --> 00:43:49.406 Yeah, yeah 00:43:51.512 --> 00:43:51.832 Jonny: okay. 00:43:51.872 --> 00:43:52.122 Yeah. 00:43:52.652 --> 00:43:59.282 question number two, is, what's a product that you think is 00:43:59.312 --> 00:44:00.672 exceptionally well designed? 00:44:00.952 --> 00:44:05.722 it can be physical, digital, anything in the world, something that you, you 00:44:05.732 --> 00:44:07.022 think you-- I wish I'd designed that 00:44:08.508 --> 00:44:09.488 Frank Bach: Ooh, that's a great question. 00:44:10.768 --> 00:44:14.308 man, I feel like this always happens when I see other designers' portfolio 00:44:14.308 --> 00:44:16.908 sites because I've been slacking on mine for many years, and then 00:44:16.908 --> 00:44:19.328 I see their sites and I go, "Damn, I w- like that's a great idea. 00:44:19.328 --> 00:44:20.108 I wish I did that." 00:44:21.108 --> 00:44:22.908 I'm just gonna look around the room real quickly. 00:44:24.068 --> 00:44:28.798 don't know if you, if you can see this, but I've got a, I've got a little… 00:44:29.258 --> 00:44:29.798 Can you see this? 00:44:29.808 --> 00:44:32.058 I've got a little television here. 00:44:32.112 --> 00:44:32.252 Tom: Yes. 00:44:32.688 --> 00:44:37.238 Frank Bach: It's an old school-- It's a, it's a little television. 00:44:37.238 --> 00:44:41.538 It's an old school Quasar, sort of like kitchen countertop television. 00:44:41.538 --> 00:44:42.798 It's about a five-inch screen. 00:44:43.278 --> 00:44:48.358 And I think probably because technology is moving so quickly, I found myself like 00:44:48.358 --> 00:44:52.418 yearning for old technologies and kind of wanting to tinker around with them. 00:44:52.798 --> 00:44:54.938 And yeah, this thing is a black and white TV. 00:44:55.108 --> 00:44:57.878 It has like one channel, and it has a radio on it. 00:44:58.208 --> 00:45:04.088 And I'm, I'm just really enjoying, sort of the, the balance of like working 00:45:04.088 --> 00:45:08.238 in such a high pace technology kind of environment and then having this like 00:45:08.248 --> 00:45:12.548 old retro television that like fi- I'm just listening to the radio, which is 00:45:12.558 --> 00:45:13.948 something I haven't done in many years. 00:45:14.958 --> 00:45:19.358 and it, you know, it's old, it sounds kind of bad, but it also sounds warm 00:45:19.408 --> 00:45:22.778 and fuzzy, and it's kind of nice to just play with the radio dial. 00:45:22.778 --> 00:45:26.758 And yeah, I'm just, I'm of maybe the s- same mindset that drew people 00:45:26.758 --> 00:45:31.718 to, drew people back to vinyl, after the, the era of s- of CDs. 00:45:32.098 --> 00:45:36.118 You know, there's just something about, about the tactile, that, 00:45:36.128 --> 00:45:37.598 that kinda makes us feel human. 00:45:37.598 --> 00:45:42.158 So yeah, I, I think that little television is poorly designed, 00:45:42.178 --> 00:45:43.768 but kind of perfectly designed. 00:45:45.554 --> 00:45:47.374 Jonny: There's something about the permanence of, of 00:45:47.394 --> 00:45:48.274 these things, isn't there? 00:45:48.274 --> 00:45:48.594 You know? 00:45:48.834 --> 00:45:49.074 Yeah 00:45:49.338 --> 00:45:49.668 Frank Bach: Yeah 00:45:50.626 --> 00:45:51.126 Tom: So true. 00:45:52.236 --> 00:45:56.766 I imagine working at Anthropic or any of the labs is a, a very intense 00:45:57.374 --> 00:45:57.704 Frank Bach: Mm-hmm. 00:45:57.736 --> 00:45:59.106 Tom: to be, to be right now. 00:45:59.856 --> 00:46:03.276 you have worked at Headspace, which I think of as, as we discussed 00:46:03.346 --> 00:46:08.076 earlier, probably the other extreme in terms of calm and thoughtfulness, 00:46:08.320 -->

00:46:08.690 Frank Bach: Yes 00:46:10.176 --> 00:46:15.676 Tom: Do you have any advice for, for anyone who is, in a very AI centric 00:46:15.676 --> 00:46:20.266 environment on how to not feel overwhelmed, to not feel like the world 00:46:20.266 --> 00:46:22.176 is running away, but to, to recenter 00:46:22.240 --> 00:46:22.510 Frank Bach: Yeah 00:46:23.296 --> 00:46:23.886 Tom: Any, any tips? 00:46:24.640 --> 00:46:28.880 Frank Bach: Yeah, I think, I think this is definitely not an Anthropic only thing. 00:46:29.770 --> 00:46:33.860 that I've, I've noticed in other companies and talking with other folks 00:46:33.870 --> 00:46:40.040 who are working in similar spaces that, we-- nobody has it perfect yet. 00:46:40.050 --> 00:46:41.140 You know, nobody knows… 00:46:41.180 --> 00:46:42.340 There's no perfect process. 00:46:42.340 --> 00:46:47.070 There's no, We're still in the stage where you can shape the 00:46:47.070 --> 00:46:50.780 process and the output and all the stuff kind of however you see fit. 00:46:51.250 --> 00:46:53.680 And there's, yeah, there's no right or wrong way to do it. 00:46:53.740 --> 00:46:56.190 I think just jumping in is the most important part. 00:46:56.590 --> 00:47:00.720 And we have lots of time in the future to, you know, to, to figure out 00:47:00.720 --> 00:47:05.960 what is, what's the perfect scalable way to work in all of these tools. 00:47:05.960 --> 00:47:09.450 So just don't be afraid to mess up. 00:47:09.460 --> 00:47:10.810 Messing up is part of it. 00:47:11.260 --> 00:47:14.770 Be willing to, you know, we talked about the impermanence thing of being willing 00:47:14.770 --> 00:47:16.840 to throw it all away if it didn't work. 00:47:18.300 --> 00:47:22.710 yeah, I w- I would lean into that and, you know, there's a, there's a, a saying 00:47:22.720 --> 00:47:26.900 something around like the, the opposite of anxiety is excitement or something, 00:47:26.900 --> 00:47:28.640 or it's like two sides of the same coin. 00:47:28.980 --> 00:47:32.280 And I definitely felt myself go through that around like AI anxiety 00:47:32.300 --> 00:47:35.370 and then AI excitement, and I- it just sort of flipped one day. 00:47:37.140 --> 00:47:40.320 but I think that's just a matter of, of getting comfortable with, with the tools 00:47:40.320 --> 00:47:45.500 and the, you know, just as a designer leaning into GitHub and understand, 00:47:45.500 --> 00:47:48.680 okay, this is how all the parts connect, you know, behind the scenes. 00:47:49.160 --> 00:47:52.620 And you went from like not-- it was a total black box where you didn't 00:47:52.650 --> 00:47:56.360 know any of it, and now you have some sense of it, and you can actually 00:47:56.360 --> 00:48:00.290 speak with engineers with confidence about what it is you're building. 00:48:00.290 --> 00:48:05.350 And I think like that's, yeah, it's a nat- very natural human 00:48:05.350 --> 00:48:06.350 feeling to go through that. 00:48:06.360 --> 00:48:06.800 And, 00:48:07.340 --> 00:48:07.360 Jonny: Yeah 00:48:07.480 --> 00:48:10.090 Frank Bach: but you know, it does, it kind of, it does, you know, 00:48:10.200 --> 00:48:13.150 sort of on- the only way out is through, so to speak, you know. 00:48:13.150 --> 00:48:18.150 Like you gotta get into it and there's no, You know, we used to 00:48:18.150 --> 00:48:23.150 have these, I've never really been a big believer in like tutorials. 00:48:23.320 --> 00:48:26.560 You know, I find them like really difficult to follow and like it's much 00:48:27.430 --> 00:48:30.680 like, you know, you're sort of-- a tutorial is like sort of practicing the 00:48:30.680 --> 00:48:32.690 thing before you need to do the thing. 00:48:32.690 --> 00:48:35.300 And so like why would I do that with… 00:48:35.340 --> 00:48:37.910 You know, I find that my brain just doesn't really work that way. 00:48:37.910 --> 00:48:39.500 I kinda just need to jump in and do it. 00:48:40.620 --> 00:48:41.030 so yeah. 00:48:41.850 --> 00:48:45.490 close-- I would say close your 100 tabs you have open that are giving you 00:48:45.490 --> 00:48:48.970 anxiety around the things you don't know, and just try to learn like one 00:48:48.970 --> 00:48:50.690 thing today, and you will get there. 00:48:52.854 --> 00:48:54.694 Jonny: That is a great bit of advice to finish on. 00:48:54.954 --> 00:48:58.054 I think we should, we should wrap up. 00:48:59.354 --> 00:49:00.444 Frank, where do we find you? 00:49:00.634 --> 00:49:06.224 just for, you know, for w- what, LinkedIn, Twitter, what's- 00:49:07.296 --> 00:49:10.376 Frank Bach: Yeah, LinkedIn, Twitter, I think those are both my jams. 00:49:11.236 --> 00:49:14.796 yeah, I'm easy to find or you can just Google, you'll find my 00:49:14.796 --> 00:49:16.256 website and all the links are there. 00:49:17.144 --> 00:49:17.344 Jonny: Okay. 00:49:17.456 --> 00:49:17.736 Frank Bach: but yeah 00:49:18.074 --> 00:49:18.454 Jonny: Awesome. 00:49:19.174 --> 00:49:19.984 Thank you very much. 00:49:20.014 --> 00:49:25.384 And, if you're in LA, quite unlikely that some of our audience are in LA, 00:49:25.404 --> 00:49:29.384 but, but if you happen to be in LA, go and check out Frank's, Frank's, meetup. 00:49:30.474 --> 00:49:31.494 thank you so much. 00:49:31.884 --> 00:49:33.374 it's been fantastic to have you on. 00:49:33.404 --> 00:49:34.504 Really fascinating. 00:49:35.194 --> 00:49:37.164 and, have a good rest of your day 00:49:38.184 --> 00:49:38.674 Frank Bach: Thank you guys. 00:49:38.674 --> 00:49:39.224 Great to meet you. 00:49:39.224 --> 00:49:39.964 Great to chat with you.