Subscribe
Share
Share
Embed
Conversations with designers, founders, and builders behind some of the best work
Ben Blumenrose Full Video Podcast.txt
English (US)
00:00:00.000 — 00:00:57.960 · Speaker 1
Ben Bloom and Rose sits at a pretty useful seat for where design is at right now. He's not just watching one team figure out AI through designer fund. He gets a front row view to how a lot of different companies are absorbing it. The messy version, not the keynote stuff. Who's changing how they work, and who's just layering tools on top of old habits, and who's building new leverage.
Not just a new kind of chaos. That's why I wanted to talk to him in this conversation is a lot more about what happens when that floor rises, when more people can make decent things faster, what actually becomes valuable? What does AI fluency look like beyond a buzzword? And what happens to junior talent, to hiring, to judgment, to process, to the shape of a design team when the old lanes start breaking down?
Ben has a clear point of view on all of it. This is state of play. Let's get into it.
00:01:05.019 — 00:01:40.100 · Speaker 1
It feels like for me, a lot of the things we're experimenting with on our very small team is, is project Car, and I've got lap project car running on some pretty good cylinders in some areas, others not so much. But I already start to fear I'm I'm putting so much investment into like calcifying a few things around this project car and all of a sudden in perplexities personal computer is going to come out and it's it's like all this work I put into it is not done optimally.
And I can just like throw it away. And it's like, did I really make headway here? I'm not sure. Yeah, yeah.
00:01:40.100 — 00:01:45.020 · Speaker 2
I think you almost have to be as flexible with your design process now as we are with our designs.
00:01:45.580 — 00:02:27.590 · Speaker 1
Yeah. And you know, like what's interesting and I know that you're collecting, um, data points about this with your state of design survey, and we're trying to do similar around prototyping. And I'm, you know, we've seen a couple of these surveys and these data points come out and there's a lot of just there's a lot of conflicting information.
I think it's hard to you know, it's certainly hard to plan a flag anywhere. But I look at Figma and the thing that was interesting was to see how down the middle or almost like in three columns, it appeared the sentiment amongst at least designers that things were getting better, or that they were moving faster, and it was almost this like three way split.
What are you hearing from your portfolio about how their teams are embracing it, and do they feel more productive?
00:02:27.630 — 00:05:33.920 · Speaker 2
Look, we have a range of companies, right? Some are small and just starting out. Some are like 2 or 3 years in. And then you have like a stripe or Augusta, which are very like many, many years in. Certainly I would say most companies are now using AI in some shape or form. Right. It's very rare that you talk to someone and they're just like, yeah, we're just not doing that.
We're just like sticking with like Figma mocks, giving the mocks to the engineers. You know what I mean? Like, uh, doing research by old schools, like Talking to folks, doing data by working with the data. You know, it's like there aren't any examples of that, but and so then so then it's like we're in the shades of gray.
Then you have, you know, companies that are full on, um, you know, designers, shipping code, uh, designers collaborating with engineers on internal tooling, designers building tools that people use for design work. Or you have teams that are like leaning more into AI in the research phase, you know.
So using it really as a way to like, gather insights and understanding of the space of what's happening. Um, and then and then as they go down, maybe like they use AI a little bit less in other parts. So it's, it's a real like range of both like depth and breadth is a lot of there are big differences. But I would say like almost across the board.
Um, folks are starting to like bringing AI in. They're having it be like the de facto way they do things. Um, there is a lot more formal sharing of, uh, how to do things with AI inside companies. Now, compared to a year ago, a year ago, it was like it was a real just like, uh, companies telling everyone like, you know, use AI for everything.
And then employees are like, okay, how do we do that in the companies? Like, I don't know, spend nights and weekends and figure it out. It's like, okay, and then what? And then just like start doing it that way and hopefully it like spreads, you know. Whereas now we're hearing a lot more like, oh yeah, every Friday we get together, we're sharing best practices.
People are creating tools. We have internal, um, like an internal tool, uh, development environment to make it easier for designers to build things. You know, it's like it's starting to, like, codify a lot of, like, best practices for different companies and how they work. So we are seeing more of that.
But it's still, um, I would say, whereas like three years ago it was probably quite homogenous. Right. It's like, oh, you know, you know, it's like we design in Figma, we do Figma prototypes. Maybe a few people do like code prototypes, but that's like more the exception than the rule. And then we work with, you know, it's like it was very limited in terms of breadth of, of process and tooling.
And now like there's a much broader set of tools and a much broader set of processes.
00:05:33.960 — 00:06:39.300 · Speaker 1
What's what seems clear is that smaller teams like mine and startups that are still very young, have the opportunity to kind of just like, throw away a lot of ideas and immediately reconstruct how they create software. But the companies, the companies that are really interesting to me too, Or how are enterprises?
Because it seems like they're trying to adopt this a lot faster than we've typically seen enterprises adopt. And when I talk to companies that like Carvana, for example, they're just saying, go wild, do whatever you want. They don't have like an AI embedded engineer kind of managing this, just like here's tokens, here's the software we're allowed to use.
Do what you want and let's like talk to each other. Meanwhile, you have companies like eBay who've just been around forever. Lots, lots of processes that are kind of calcified. And they might have like a lunch and learn with somebody who's been experimenting. And so, you know, the question is for enterprise teams or these larger teams that are trying to adopt this, what are you seeing seems to be working right now?
Is it having one designated person or is it just letting everyone run wild? Is there a middle ground?
00:06:39.860 — 00:07:02.200 · Speaker 2
Yeah, I don't think there is a one. Even when you say enterprise, like, I don't know that we could put like stripe in the same bucket as Ford, Right. Um, so I think it really depends on like, what's your internal culture or what's the internal, uh, skill set. And then you have to adapt to that. Right.
00:07:02.440 — 00:07:04.040 · Speaker 1
So if you.
00:07:04.040 — 00:08:31.770 · Speaker 2
Have a company filled with technical people, with tinkers, with a culture that is like flexible and malleable, then yeah, you probably want to be doing something of more experimentation, more bottoms up. Um, maybe like embracing a little bit of the chaos and starting to see where people naturally gravitate.
But in a very, very large org that might just be like, I mean, we're even seeing it in not large orgs. Like a simple example is basically like these prototypes. And this is something I'm hearing time and again, um, everyone's excited about having like higher fidelity prototypes, but they're all in URLs and there's no standard way that those URLs are cataloged.
And, uh. How is feedback given? Okay. Now it's like looms or slack messages, right? It used to be just like comments in Figma. Again, it was all in one place. Now it's like it's not in one place. It's in different tools and all over the place. So then it's like, you know, 2 or 3 years, 2 or 3 weeks down the line.
You want to know where's the latest version of X? And instead of just going to Figma. Com now I have to go scour slack messages and see where the latest prototype URL is. And it's in this dock linking to this URL and this piece of feedback. Do you know what I mean? And you multiply that in a team of 100 and oh my god, we have a total shit show.
And why don't we do that?
00:08:31.810 — 00:09:00.310 · Speaker 1
Well, because you have so many people now different roles. You have non design roles contributing to that ideation phase with these artifacts. And so it seems what I'm hearing similar that, you know there's a lot of throwaway work that was relevant to a conversation at a point in time. And it's not making it to other conversations.
and it's hard to know. Version control and sharing with it. But are there are there things you're at least seeing or hearing right now that appear to be emerging? Cautionary tales don't do this. Maybe avoid doing it that way.
00:09:00.710 — 00:09:58.650 · Speaker 2
I would say if you're if you're like a big company, do not just like let anyone use all the tools and do whatever. Like I actually think, um, what I would do is basically, uh, create like some sort of. So I had this probably talk to one of our companies and we were talking about like AI ops, and we were talking about the role of basically, uh, someone who would come in and just look at, like the design process and codify, codify the process, uh, research the tools, help with education, all that stuff.
And so there was a thing called like design ops, or there is a thing called design ops. And usually a company is like a team of 100 designers or more before you think about a design ops role. But I've talked to a few people now, a few design leaders who say they are thinking about bringing that role in very early, like a fourth hire,
00:09:59.730 — 00:10:42.330 · Speaker 2
which you're just like, wait, what? Like a person that's just looking at that. And the reason is because someone like that could just basically, if you can get one and a half to two x out of every designer, it sort of makes sense, right? Because instead of that person could basically double the design team with operational excellence, you know what I mean?
So if you have a design team of 5 or 6, that person can make them be as good as ten. So it's actually like kind of makes sense. The math makes sense. But we were we were talking about like, well that the design like AI ops, AI design ops or design AI or just like that, that that role sounds crappy or whatever. And so I was I'll tell you, you should call it an AI Imagineer.
00:10:43.330 — 00:10:44.010 · Speaker 1
Oh, boy.
00:10:44.140 — 00:10:44.940 · Speaker 2
Yeah.
00:10:45.260 — 00:10:47.100 · Speaker 1
That's right. How did that fly?
00:10:47.180 — 00:12:07.760 · Speaker 2
I think she liked it. We'll see. We'll see if I can. But like, it's basically like, you know, how would you get a designer to get excited about a role like that? And it's just like, hey, come in and envision just a new way of working, a new way of, you know, how do we do these tools? That's kind of like, you know, the Disney Imagineers basically had a big hand in that.
Like they looked at the processes, new technologies and how to how to also apply those things. And so I think it's kind of like an interesting conceptual model versus if you think about it, most designers, I would say for a long time, we spent like 95% of our time on the design output. Right. It's like something gives me a problem.
I think about how to do it well. And but I'm not thinking about like, how do I improve the way I work so that I can do design better? Maybe you do that a little bit. You know, that's we call that professional development. And I was like, you spent like you went to, like a conference once a year, right? So I was like three hours out of the year, one day out of the year.
So one 356 right. You you would go and better yourself or better the process. Well, this is like, uh, saying almost like the way you work, uh, should be a huge chunk of how you spend your time,
00:12:08.800 — 00:12:11.320 · Speaker 2
which is kind of like an interesting thing.
00:12:11.920 — 00:13:03.660 · Speaker 1
Well, and to your point about professional development and, you know, people leaning into some of these, you know, a design ops role, there's a lot of systems thinking that appears to be emerging is really beneficial right now to kind of like, think on that high level and allow that to distribute through these new ways of working.
Um, and I know a lot of, you know, senior designers are finding ways to fill those roles in some capacity. I think the number was something like 56% of design job postings. And it's so hard to cite this data because this stuff is changing in real time. But, um, you know, especially so 56% of design job postings seem to be skewing senior.
Think there's like a little bit of the middle and then something around 25% or junior openings and that's on the decline. How how are your teams thinking about the junior roles and like professionally developing this, this entry level market.
00:13:03.700 — 00:13:14.100 · Speaker 2
The whole like idea of a junior designer. First of all, no junior designer goes out and thinks they're junior designer. I remember like ten, 12 years ago. Um,
00:13:15.220 — 00:13:21.940 · Speaker 2
it's like we saw people applying for a senior level roles, and they're like two years out of school
00:13:23.060 — 00:14:23.350 · Speaker 2
and you're like, how are you? Well, how do you think you're senior? And they're like, well, I've been designing since I was like 15, 16. You know, I've had like internships and blah blah, blah, blah. So yeah, I'm like ten years of design experience, but I graduated in 24, you know. Um, and so designers have always been very comfortable with, like, overstating their seniority.
Uh, right. So I'm like, okay. Right. Yeah, dude. Same thing. Same thing. Right. I'm like, uh, you know, uh, how many like, design directors, you know, that are like 25, 26 or whatever, and they're like, yeah. Um, and so I think it's it's more just like a junior. The junior thing is more of like a skill set thing.
Um, and I actually think you can get a lot of the junior skill sets out of the way pretty quickly. And so if I'm a true junior designer, like, if I'm in school, uh, and there aren't a lot of junior design roles, what I would be doing is basically like thinking about how can I just start,
00:14:24.470 — 00:16:08.630 · Speaker 2
you know, doing work that looks and feels real. Um, and that could be through internships or freelance or whatever. And, uh, and then you apply for not the junior roles, you know, um, but like, the flip side is actually that I for companies, I actually think they should be looking for people who are earlier and I won't call them junior.
I would just say like earlier career people, because these folks do not come with the baggage of working the old way. Right. So, like, it's kind of this thing, it's like I want junior people, but not the ones who call themselves junior. You know, it's like you should be looking for early career people who don't think they're junior, who think they're there.
They are working the way that all of us should be working, and those people are actually super valuable, and they probably can do the work of a lot of like mid to senior level level people in terms of pure output, the mid to senior thing like the true mid to senior people. The value they bring is basically like taste and experience and just just like a knowledge of how?
How to work, you know how to be a professional, which I think it takes a little bit of time. It's amazing how many people just, like, don't know how to, like, respond to an email well, or show up, you know, just show up to meetings on time, you know, like you're already like in the top ten, 15% or literally just do what you say you'll do by the time you say you'll do it, like, hey, I'll have that for you on Thursday.
And then on Thursday you send them the thing and you do that. Ten times out of ten, you're in the top 10% of designers.
00:16:08.950 — 00:16:39.560 · Speaker 1
And nothing you just said, by the way, had anything to do with being, you know, AI native. But I do think you alluded to this idea that the junior or the entry level designers, um, they're not being protectionist about like a way of work. They're just, hey, this is how I'm doing it. These are the tools I'm learning.
Is there is there some like, assumption inside of that that they are adopting some of these new tools or ways of work with AI? What do you think of the term AI native? What does that mean to you? Are your teams in your portfolio using that?
00:16:39.840 — 00:18:19.020 · Speaker 2
Yeah, I think so. Either like AI native or AI. AI fluency is the other like term I've seen, but it basically means okay, like what does it mean? So I'll give you like an example from my life. So, uh, this is forget design. Um, you have like a thought. Oh, uh, that's like an interesting flower. Um, I wonder if that's, like a flower native to California.
Um, if you had that thought 40 years ago, you would say, uh, I'll look that in my Encyclopedia Britannica, and hopefully it's in there, you know, and then this thing called Google came out, right? And then it's like, oh, actually, when I have a question now there's this thing called Google, and you slowly train your mind that actually instead of asking your friend or whatever, it's like, just Google the thing.
And it kind of led to the meme of like, you know, when people ask you a question, it's like, let me Google that for you. It's like, just Google it, right? The next iteration now is like, let me AI that for you. So basically the questions that we have in our heads, there are certain things like Google, you know, there are questions that we have where like you wouldn't respond with let me Google that for you because it's a much more complex question.
So for example, um, we were talking I was with a group of friends this weekend, and we were kind of, um, thinking about how, like the 80s had a lot of one hit wonders. And we were we were like, who would win? Like, what's the goat one hit wonder? You know, like, who had the most like the song that was like most influential, um, that
00:18:20.100 — 00:19:06.160 · Speaker 2
the band was it was just really a one hit wonder, you know, and that's a question that if you Google do, you probably get nothing, because it's just such a weird it's like a combination of data, um, that you have to and it's like, what does it mean to be the best one? One hit wonder. You know, you have to actually, like, start thinking about the different.
And then there's probably a discussion that you would have with AI to figure that out. How would you like? How would you actually like consider what the Goat one hit wonder song is. And then it led us to thinking like, well, what? Why do we have more one hit wonders in the 80s and 90s? That's kind of interesting.
And we just went down the rabbit hole. And with AI, you can go down that rabbit hole. AI native means that when you kind of have this sort of like, I wonder if instead of saying like, well,
00:19:07.200 — 00:20:50.430 · Speaker 2
I wonder if, but I don't have the time or energy or whatever to to go down the rabbit hole. AI native people are like, I wonder if and I know exactly how I would go about doing that. They know the kind of questions to ask the AI tools, which AI tools to use, how how to like structure their process in those tools, to what level of detail they want the output to be to answer the question they need.
They know when to say no to like. I wonder if so that you know it's like time box the thing so you don't like lose a whole day. And I wonder if one hit wonders or like, you know, like, where is this going to take. Because you could really, especially me as a designer, if I find that I'm my whole life is about connecting the dots and things like this, you can just spend your whole day going down these areas, because it really is like you have an expert in all the things in the room next to you in that world.
You have to make sure that you're not just, you know, that you're focusing on the right things and going down the right pathways and coming out of those pathways and going to the right level of fidelity. I think that's like what AI fluency means is just like that, it becomes a natural thing. And for young people who grow up on this stuff, like my daughters, they're just going to it's just going to be a muscle that's going to be automatic for them.
They don't have to unlearn anything for the rest of us. We have to unlearn the constraints that we had before. We have to unlearn the processes that we had before. Right? We had to unlearn the roles. You know, how many times have you heard, like, stay in your lane, right. Everyone's being told to stay in their lane.
Well, actually, like the beauty of AI is that allows us to not stay in our lane,
00:20:51.550 — 00:21:04.190 · Speaker 2
you know? And in fact, we should probably be, as a society, being like, uh, there needs to be some new thing which is just like, don't stay in your lane, you know, go down that lane and explore that thing and then come back to me.
00:21:04.550 — 00:22:11.930 · Speaker 1
I love the way you framed this AI fluency as like, what do you do with I wonder if and then when we think when we kind of like lean into that and we say, okay, and now that person plugs into, um, a team that's, you know, been moving fast and doing what they're supposed to be doing. That's where I feel like it gets a little messy because we have a lot of conflicting information.
Some teams are saying, yeah, we're moving fast. Things are great. You have reports like Peter coming out and saying they gave a bunch of developers, uh, with, you know, they said, hey, here's a bunch of tooling, AI tooling. Go use it to do work with these open source code bases that you were very familiar with.
And it came back saying most of them felt like 19% slower, uh, working with those tools. And so how do you if that's true, how do you. Is it does it become kind of the factory problem? Like, we're we're trying to we're just trying to, like, can add some piece on top of like old factory floor or is it like we need to we need to maybe advocate for teams to redesign the entirety of everything.
Like get rid of lanes completely. Where do you fall on that spectrum?
00:22:12.330 — 00:24:08.959 · Speaker 2
Yeah. So here's where I'm pretty sure it's going. We've had T shape. Take the T shaped you know T shaped designer. You usually have depth in an area. Right. Um and and breadth in a number of areas. but really good designers. Uh, they're kind of like they have a thing that they're just exceptional at. And they have a certain competency in the others.
Right. So it's like, let's say you're an incredible UI designer. Well, that doesn't mean, you know, it's like you're probably, um, not as good as someone who's excellent at UI design, but also has a little bit of technical knowhow and a little bit of like understanding of user research and a little bit of understanding of like, product design strategy.
Right. The T now is going to be a lot broader, right? Because now it's like, uh, if you're a UI specialist, it actually is. You're going to be even better if you actually know. Yeah. Like true design, if you can actually do like your own design research. So not just, oh, I sort of know a little bit about it, but I actually can run a little bit of design research on my own.
I know where to get the data, um, why people are clicking on certain things and why not? I can understand load times. I can understand how to pull all the all the, let's say, all the icons we have. And I can build a tool that allows other people to build new icons. That's like an, you know, it's like so I know the research, I have technical know how to build some internal tools, have enough PM ING understanding to run a little bit of a process with some other folks who might want to be helping and supporting this, this effort.
So the T becomes much broader. And then I think the level of depth is also going to get deeper, because if you have an expert with you at the thing that you're doing
00:24:09.960 — 00:24:44.850 · Speaker 2
all the time, right. So imagine like how often we tell designers like, it's great to work at a place where you have a mentor. Why? Because that's kind of where growth happens. Or a lot of growth can happen. Right. So you're not just like, you know, someone who can help you see around corners and help you avoid mistakes.
Well, not like everyone has something like someone like that available to them, you know? So if you're about to do design research on your own, well, you can just ask Claude. Hey, what are like, the things I need to be watching out for? If I wanted to pull this data, how would you pull this? Right?
00:24:44.890 — 00:25:20.890 · Speaker 1
When I use certain tools, I experienced this level of phantom competency. And if I live in that space long enough, I find myself actually learning real things in that phantom competency starts to become kind of my flaw. You know, and it's like there's a weird sort of apprenticeship model inside of using these tools for a long enough period of time with everything you just said.
Though my question is, how do you evaluate somebody, you know? Okay, we're looking for people who are AI fluent. And we have this this idea on a on a larger scale, like what that looks like. But how do you say you are? How does somebody demonstrate that they're AI fluent?
00:25:21.050 — 00:27:16.610 · Speaker 2
I mean, you basically have them. We do this all the time. When we say, give us a case study. Show me how you build this thing. And for the last 30 decades, usually when someone gives you a case study, it's like, here's the five person team I worked with. Here's the researcher I worked with. Here's like the technical team, here's the PM.
And, you know, here's my role. And it's like it was a team of 8 or 7 or whatever it is. And I did this like one thing. Um, and I think now it just is going to look, it's, you know, it's like, hey, I, I came up with the like way to do the research. Um, I asked the researcher if that that seemed right, if we had that access to that data, um, if there was other things that they would pull.
But ultimately, like, I was able to do that. Then I built this tool internally, using Claude to visualize kind of the research and make sure that people were on board with, like, this approach. Then I, you know, started to build a prototype. You know what I'm saying? It's like at every phase. Did you do it the old way or did you explore the new way?
You know how that. And I think you know what's interesting that I'm hearing from folks and this is going to change. But for now they are setting a clearance bar that people have to or like a hurdle that someone has to clear to get hired. Uh, that is an AI hurdle. Like they want to see AI fluency of new hires. Um, they want to see AI fluency, uh, internally.
But so far for most companies, they haven't gone so far as to codify it into like, hey, if you are going to go from a IC five to an icy six, we need to see that you use AI in this phase of this. You know what I mean? Like it hasn't gotten to there. So it's almost like the new hires actually have a higher burden of AI fluency than internal people.
00:27:16.610 — 00:27:46.300 · Speaker 1
I've heard from some teams who will say, hey, we'll we'll give candidates, you know, they'll do the take home thing. Probably still pretty controversial, but they'll say, hey, we'll give them a quick take on thing. But by design, it's like just a ridiculous amount of information that you really only could reasonably synthesize with AI.
And through just the act of doing that, it's a it's a self disqualifying for some people. And how do you how do you feel about exercises like that? Do you think that's potentially one way to get that done?
00:27:46.340 — 00:28:24.280 · Speaker 2
I mean, I play games, you know what I mean? I would just be like, you know, like I think that's a fine exercise. But then just like, that's like a test, like, will they use AI or not or something? It's just like give, give someone that this is just say, hey, look, this is something that a couple years ago we would have given someone a week to do.
We want to see what can you do in a day and use any tool, any whatever you want to do, do. But we want to see high signal, high clarity, um, like high intent. Um. Right. And spend three hours. Time. Box it to three hours.
00:28:24.760 — 00:28:31.440 · Speaker 1
I want to I want to hear where you're at with these these tools. Um, because nine months ago, we we last spoke, you and I, nine months ago.
00:28:31.480 — 00:28:33.160 · Speaker 2
And nine months ago. Oh, my God, it.
00:28:33.160 — 00:28:35.240 · Speaker 1
Was about nine months ago. And if I recall.
00:28:35.240 — 00:28:41.720 · Speaker 2
We were children. We were. We were little babies walking through a primordial ooze.
00:28:41.920 — 00:28:59.720 · Speaker 1
Uh, opus 4.5 I don't think was out yet. Um, and who knows all the other iterations. But since then, and you had a, your take at the time was that the tools were really good at giving you a lot of options, but as soon as you wanted to, like, refine something, that's when they started to fall apart. How do you feel about that right now?
00:29:00.320 — 00:30:05.139 · Speaker 2
You know, it's still kind of like that, right? I don't think that that's been solved because I actually think like they're so good at engineering because engineering is just like, I just need to get the thing to work. And they have so much data in terms of like, um, what works and what doesn't. But like with design, it's just not.
It's not clear. Um, you know, I could show you, like, five marketing site designs for a company, and you could make a good argument for all five. If they're, like, pretty good formally, you know, in terms of type, color, contrast, all that stuff. And then it's like, you know, one thing might convert better, but one thing might give you better brand lift, which we don't really know until we get data ten months down the line.
And like, you know, even humans don't know and AI doesn't know like the, the, the the connecting the dots is just not there. It's just such a it's a much more complicated system. It's in people's heads. Um, so I just think like that's just a harder thing for AI to do is to like do that. Well. So I think it's very good at getting things like
00:30:06.180 — 00:30:29.070 · Speaker 2
pretty good or even good. Like I would say, you know, it's like the floor is, is quite high and probably It's definitely higher than when it was, uh, nine months ago. But it does feel like there's probably some ceiling where AI will never get there on certain things.
00:30:29.110 — 00:31:29.010 · Speaker 1
I think it does a really it does a really good job of solving the blank canvas problem. Right? Like it really gives a good jump off point these days. I've found and a lot of the people I've talked to in a lot of different roles, there's this there's there's a little bit of this like, you know, I guess I could call it moral panic about, should we be doing this?
Is this the right thing to to adopt these tools? We're pushing, uh, technology driven design? I've kind of heard of this where it's like we're mandating usage around the tooling, and this is the first time this industry has ever seen this, which I don't know if that's entirely true, but we are a younger industry.
But I think historically we've seen a lot of like big disruptive innovation, followed by a big kind of outcry of I don't know if you know, oh, I don't like the way things are changing. CAD. Um, I think we had, like, the electronic, like, health records was something mandated by the government. We've seen a lot of examples through time.
Do you feel like the this sort of moral panic is, is warranted right now? Do you think that's what's happening?
00:31:29.050 — 00:31:52.330 · Speaker 2
Like we're basically going to create if you think about how many designers around the world that number or how many people can produce good forget like how many people would self-identify as designers, the number of people that can produce good design is going to go vertical and is going vertical basically.
And then the question is like, um, in that world, then like, what is your value as a designer?
00:31:53.450 — 00:33:26.960 · Speaker 2
Um, and if you are a designer that can produce good design, it's going to go down because there are more, way more people that could produce good design. Now, if you're a designer that is exceptional, that means that you're not just, um, someone that can produce beautiful pixels because that's going to that value is going to go down, but it's going to have to be like, like I said, your tea is going to go broader, you're going to be the designer is going to like be like, okay, what is needed here?
What type of product is it? Software? Is it a user experience thing? Should it be? You know, I was talking to a founder today, um, and he was talking about like this dashboard that he wanted to design and redesign to make sure try to get more people to use this dashboard. And my I was like very quick. Within a second I was like, the dashboard is the wrong thing to build.
It's I'm like your customers. Yeah. They don't they don't like the dashboard, not because of the esthetic or the information in the dashboard. They don't have time to log in to some website. You need to push the data to them, whether it be over text or email or voice, whatever. And that needs to be how you communicate to them.
Progress. Not by iterating over 100 dashboards. I know this because I have 20 years of building things and knowing that how people behave with these tools, right? Um, that's the value of right, like experience, right? That a designer who's actually going to continue to add value in this world brings.
00:33:27.000 — 00:33:50.440 · Speaker 1
Yeah. And I wonder the same. You know, is design engineer going to become sort of a default, at least for a moment in time? I'm not sure. But for you, with the designer fund and the thesis built on this idea that designer roles, people who are, you know, fill that that seat, make really great founders as well.
Does that thesis need new language or do you think it continues to for now?
00:33:50.440 — 00:34:34.020 · Speaker 2
No, I think like we still have we still know the power. The power of design right now is like, you know, it's still it's still clear who is a designer and who self-identifies as a designer. And the power that they bring to the founder role. And I think now it's even more powerful because it is true that, like building things is getting easier and design is getting easier, but not at the same rate.
So, um, so it's like much easier to get great engineering done, but it's not that much easier to get great design done. So it becomes just like a bigger differentiator, right? Um, and I think designers are just going to be in in a world where anyone can build anything,
00:34:35.700 — 00:35:28.040 · Speaker 2
the value goes to the person who understands kind of the direction to go, the things to say no to. Um, how do I, you know, designers are great at like, inspiring folks and telling great stories and should be great at recruiting great people. Um, and so recruiting the right people also becomes even a bigger deal in this world.
So I still, I think, like it's even more so now that a great designer as a founder brings an even bigger, um, uh, like level of excellence to the role potentially than before. So I think for now, no, but yeah, I don't know, in 4 or 5 years, is there a different term for a designer founder like, I don't know, it's a good question.
We'll see.
00:35:28.240 — 00:35:56.440 · Speaker 1
Well, and you know, when we last spoke to you, I like the way you painted it that VCs are really providing, you know, Kilimanjaro gear where, you know, we're going up that big monster of a mountain if if it becomes true that, you know a designer, one person and AI, it can equate to a team of 4 or 5 people. Do you think the VC path becomes more accessible?
Do you think it becomes less necessary? Where do you think that evolves?
00:35:58.960 — 00:36:03.600 · Speaker 2
Um, I still think for ambitious teams, uh,
00:36:04.920 — 00:36:06.200 · Speaker 2
venture capital
00:36:07.610 — 00:36:53.130 · Speaker 2
is still a is a great path still, because it is true. You'll need. You need less money to do an X unit of work. But the expectation now is like that unit of work needs to be done way sooner, that we want more units of work done right by the teams, and the goals are way more ambitious, right? So in that world where.
And so we're already seeing like, um, you know, take like we have a company called Physical Intelligence, which is kind of like open AI for the physical world. It's like an arguably hundred x more difficult problem. And so you probably need 100 x the capital,
00:36:54.610 — 00:37:12.470 · Speaker 2
you know, and so there are a lot of those kinds of problems where because of these tools and this technology, we now are just more ambitious and that will require more output per person. But you still need a lot of people. Because of the scale of the ambition of the thing.
00:37:12.550 — 00:37:33.710 · Speaker 1
With the frontier models getting better and better and the products built around them getting better and better, how do you differentiate somebody who's building these companies and say, that's an extraordinary person, or yeah, they're all right, but they're using these extraordinary tools. Where do you find.
How do you evaluate the the gap that that phantom competency that I spoke about?
00:37:35.110 — 00:37:43.270 · Speaker 2
Yeah. Okay. So you're saying basically like because someone has access to tools, how do you how do you evaluate, like let's say you see a good product.
00:37:43.310 — 00:37:53.110 · Speaker 1
I can ship overnight. Maybe I have a great marketing campaign. I've got some users right away because I've got some of the things done. But am I still, you know, how do you how do you evaluate that situation?
00:37:53.110 — 00:40:16.840 · Speaker 2
Well, it's still so let's say you have a team of two and they're highly technical. And if I look at the website and it's beautifully done, and if I look at the product and it's beautifully done, and if there are thought processes effective and all that stuff. It's like, okay, well, how did you and I just ask them, okay, like, who did the site?
How did you think about, like, how are you thinking about brand? What role does brand play in your company? Um, what is the product? What do you see around where's the edges of your product experience? So I think you, you know, through a conversation I get a good I'm not going to get it from a 2 or 3 minute presentation, but in a half hour with a founder, I can get a good sense of, um, how they leverage the tools, how they leverage the tools to work with experts, um, what they value, what they don't value.
Are they doing a lot with a little? Um, there are certain industries where you can't cheat, you know, it's like there are certain industries where you're just going to have to put ass in seats and grind and do the hard work, you know, highly regulated, like, we really like, like highly regulated spaces.
We really like the spaces where it's like, you have to embed yourself into the communities and like the networks or else like no one's going to give you the time of day. Um, so those are the kinds of things like we have a company called Handbook that that is started as like a, like a basically like the financial tool for farmers.
And so you can imagine like if you're like a tech company from Silicon Valley going to sell to farmers, there's going to be like a healthy dose of skepticism when you start doing that. Um, but Amber, over time, it's just like they embedded themselves in those communities. They really, like, went farm to farm to understand, like, well, what is the pain point you're having?
Um, not just like with financials, but just like in general with your farm. You know, they just like they had all these insights into how farmers work, the unique, um, frustrations they have, how they share those frustrations with one another, and in what communities and in what platforms, um, how to market to those folks like they're using, like a very different way.
These aren't like, you don't do Facebook ads and Google ads. To these folks, it's very different, right? And I think they earn the respect of people when they show up. And then you show up again and again in person. Right. And so go. Okay, good. You know, if you're the next person who is like, oh, Hamburg seems good.
Like maybe I'll do that. It's like, okay,
00:40:18.080 — 00:40:24.000 · Speaker 2
go get in your car and start going. Let's see if you let's see if you have the grit and stamina to do that.
00:40:24.040 — 00:40:54.840 · Speaker 1
I want to rap by asking you something that's, I think, close to you. It's close to me. I have a son who's going to, uh, college this fall, and. And, like, I hope that's the right answer for you, you know? Um. And I still think. I still think there's a lot of value in that. But my question to you is, you know, you were talking about with your daughters having them use AI, and it's just kind of a muscle they're used to.
And I'm trying to think about the same thing. I have a brand new daughter now, and I'm like, how am I going to build curriculum for her where it's the first time we're thinking about homeschool.
00:40:54.880 — 00:40:57.320 · Speaker 2
Or not, she's going to build it for you, right?
00:40:57.360 — 00:41:07.500 · Speaker 1
Yeah, actually, That's the take that'll hold you to, because it's probably going to come true. How are you thinking about it, man? How are you teaching your children about this today? Right now?
00:41:07.660 — 00:41:28.100 · Speaker 2
So I'll tell you for now, I'm keeping them away from it mostly. So the girls are six and a half and four. Um, I expose them to it. I'm exposing it to, like, a little bit. But I think just just as important is like. And it's going to be even more so is like
00:41:29.660 — 00:41:40.220 · Speaker 2
you first need to really know, help them flex these like, creative muscles, these like person to person, um,
00:41:41.260 — 00:41:43.860 · Speaker 2
like the ability to communicate and connect.
00:41:44.460 — 00:41:47.140 · Speaker 1
In independently on their own. Kind of.
00:41:47.740 — 00:42:06.430 · Speaker 2
Yeah. Because basically, I mean, we're already seeing the research come out that you lean too heavily on these tools and it's just like You're you're you're we're creating, you know, just like masses of dum dums where it's just like, I don't know, you know, it's like, um, what should.
00:42:06.430 — 00:42:08.710 · Speaker 1
I have for lunch today, Claude? Yeah, exactly.
00:42:08.750 — 00:43:16.169 · Speaker 2
It's like, hey, I have these six things in the fridge. Like, what do I do? Yeah, exactly. I don't know how to put a sandwich together, you know, with these six ingredients. Like, that's that's that's exactly right. And so, um, we need to first, like, be, I think, be teaching the kids. And it's even more important just all these, like, I, you know, I'm focusing with her on, just like, how do you build things with your hand?
Um, here's a, uh, the, you know, it's like a paper towels done that you have the inside of the paper towel, like a little roll. What could this be? What could we turn it into a sword? Uh, an eyeglass, uh, you know, like a baseball bat. You know, those are probably like, the the the, like, really getting those muscles strong right now, early on, I think it's becoming even more important because it's going to be so easy for them to just, like, lean on these other tools and then slowly expose them to how to utilize these tools in an effective way and not over rely on them.
Uh, and I think that that's basically going to be
00:43:17.290 — 00:43:22.730 · Speaker 2
the challenge. Um, for like kids coming up is basically like,
00:43:24.290 — 00:43:52.370 · Speaker 2
you know, where, where do I use this tool? Where where do I not? What are the things that, um, a tool like this allows me to do? Um, but also, where do I take over and where do I bring in, like, my own ingenuity? You know, but I'm hearing also just like kids, like they're terrible at, like, communicating with one another.
Just like 1 to 1 face to face, you know? Um, have you heard that, like tin can.
00:43:52.650 — 00:43:53.930 · Speaker 1
I haven't. What is that?
00:43:53.930 — 00:45:09.240 · Speaker 2
So I just bought one. We'll see. We'll see how it goes. But a few of my friends have and they say it's fantastic. So it's basically it mirrors a landline phone. So it's a it's a phone. You pick it up. It only has. So it's audio and you can talk to it and it has you whitelist numbers. And so like a seven year old can call their friend on this like faux landline.
It's like a faux landline. And so you no longer have this thing where it's like, I don't like, I don't want my kid to have a cell phone at age seven, but do I want her to be able to call her cousin or grandparent or like 4 or 5 of her closest friends? Yeah, and I have a friend who has a kid who's a little older, and he says now instead of just like, hey, dad, can you call so-and-so's mom to see if we can have a playdate on Thursday?
They have a teen. And he goes, I just tell him, you call her and you have that conversation and you tell me if you can do it. You know, it pushes the it. So this is an interest. It's like, hey, we're actually like using technology to remove features, which then actually creates better outcomes and
00:45:10.280 — 00:45:14.120 · Speaker 2
more opportunity for them to like, use it in novel ways.
00:45:14.160 — 00:45:52.120 · Speaker 1
What it took from Ben is that the bar is not just moving, it's moving sideways. It's not enough to just be the person who can make good screens. A lot of people can do that. Now. The real value is shifting toward judgment and systems and being able to cover more surface area, not just designing the artifact, but designing the way work happens.
And I think that's actually the more human read on this moment. A lot of the fear around AI is real. A lot of the disruption is real. But Ben's point, at least the way it landed for me, is that that doesn't remove the need for great designers. It raises the standard for what great actually means. That's it for this episode.
I'll see you in the next one.