Go ahead and get started. Can you give us a little introduction to yourself and what you've done? And yeah, we'll kick it off from there.

Introduction

Sure. My name is Damian. I am both a product designer in the digital sense, and I also founded and curate Hatch Conference, which is a conference for senior product designers and UX designers. It's here in Berlin.

It's about 500 people each year that come here, plus some people online. Belinda was the speaker at the very first one, which was still tiny.

Basically, I do both of these things on one side, work with some clients on digital products, and on the other side, curate the conference. Amazing.

It's a really lovely conference, very intimate, a lot of really passionate people and great discussions. It was very fun to be there for the first one.

AI in Product and UX

With that experience, with both the conference and also the work that you do, How do you see AI shaping product and UX?

And maybe not just, I mean, AI is such a broad term. What does that mean exactly? But maybe some of the types of AI that you might see that are specifically impacting, yeah, the product and UX field.

I think like for most industries, it kind of took a little bit the world by storm. And, you know, I think the, you know, the first year kind of like not a very, you know, the conference is three years old now. The first year was not really like too much of a topic. And in the second year, it seemed like it was everything everyone was talking about.

But at the same time, no one really, you know, even us thinking of who do we feature speaking about AI, everyone was kind of figuring it out. The first LLMs were like publicly available and everyone was a little uncertain of what that meant. I think this year was much different because we kind of knew what to expect. And we already had cases of like how that was really affecting the product teams in general.

Generative AI and Automation

I think the separation that I always make is like, okay, what is like generative AI and especially in the creative field is very much related to everything that has to do with image generation, you know, content, et cetera. And then on the other side is everything that's like what we saw about like how automation can really generate like new views of and how it can help us bring in things in our process that we didn't figure out before.

I think the first one is definitely kind of the more attractive to discuss because obviously it's the one that's hitting us in the ego in terms of designers being like, hey, we're actually creating this. you know, illustrations or creating these websites. And suddenly you tell me there's like a computer that can do it.

But obviously the results right now are not something that matches kind of the value of what the market wants in general. So I think there we're still pretty much experimenting. The second part gets a little more advanced, but I think it's where people are most excited and most terrified about, particularly in the cases of designers that work in very specific parts of the profession, like people who are doing research, people who are doing design systems or things like that, which is very kind of procedural, but very specific knowledge that is required.

And there's where they're seeing the more... yeah, the most things that could actually affect what they're doing, especially when in terms of repetitive, consistent processes, that is something that AI is really good at. And there's definitely those teams who have downsized, they've mostly done it in those kind of specialist roles.

And even some, for example, PMs taking over the research or like, you know, other people kind of dwelling into like designing of some screens and stuff. So that's happening. But at the same time, you know, we are lucky to live, I guess, in Germany and not in San Francisco, which things here take a little more time to kind of like come through. So I think in what's like startup and local teams, they're still so used to working in a certain way. that I don't think what AI is capable of which changes every week can still be found in the core systems of companies that have been working on the way they were for so many years.

Impact on Product Managers

And so one of the roles that often comes up in this discussion is the product manager. Everybody always likes to talk about product managers and what do they do and how is it going to be important? But how do you see AI impacting that role or do you have any fear?

And I will get into this in one of the next questions on sort of the hype surrounding AI. But how do you see it kind of affecting like specifically also that role? I think it helps all the roles finally be able to venture into something that they were terrified and had no idea.

So as designers, it helps us with code, which was something that for the majority of us was always like, I don't know how to code. I don't want to touch it or whatever. And now I can just write something and change something in the app.

But back into the product managers, I think it's the same, right? Like everything that they were always not able to like visualize and represent, they're able to now with all these tools. And that can be good or bad.

I think what's important is like it's always interesting to as a designer to get kind of more consistent or more like concrete input than just like, hey, I was thinking, what if we do something, if you actually show me a screen or you show me a prototype, can be great. Then, as a professional, I can be like, that's great for a start, but obviously this wouldn't work, or we've tested this before, or this needs further testing.

But I think in terms of learning to speak other people's language, other team members' language, especially the PM being so in the middle of the team, that's really, really helpful. I would still kind of encourage not to overstep and just be like, oh, now the AI can do this.

Why don't we just do this exactly? Because someone who's been doing it for a while will have a few more ideas why that doesn't work. Yeah, everybody has their own background and experience and knowledge that, of course, they bring to the roles that they do.

The Hype Surrounding AI

And I think, yeah, that's something, you know, moving a little bit more in that direction of sort of the hype question. I mean, we've definitely seen a lot of hype, a lot of, you know, overpromising. I mean, like with many things, I mean, there's the Gartner hype cycle because we see this with more things than just AI.

But how do you feel that that's affected either, you know, maybe the adoption of the way that product uses that or the way of which we're talking about it or sort of the way companies are thinking about it now with product and design? What are your thoughts on that?

I think in general for designers, there's definitely this kind of anxiety that got generated about like which parts of the work are going to be taken over. And I feel it's like probably for most disciplines.

For me personally, I've always been more interested because I have to curate a conference and just put someone on a stage that's supposed to know about what they're talking about. And AI being so young and everything, it's always a bit nerve wracking. And there's definitely all those hype.

1I think I mentioned to you like one third of the submissions for the conference were AI talks. The majority was just like how to use AI to improve your workflow as a designer and stuff like that. And the problem with that is like no one knows yet what that's going to look like in the, you know, as we have like agile or we have like lean and all those type of frameworks that then people could grab and then modify.

No one knows what the actual role of AI is going to be in design. And whoever is telling you that they know, you probably should not listen to them.

And the other problem is there's this race as a new industry and something with hype where all this money is being poured in and all these people are trying to create content and courses and conferences and all this type of stuff saying, because you have to be the first. Consultancies that want to be the first AI consultants, the ones that have it right. This creates this very chaotic world where we don't understand where we should get our truth from and unfortunately it seems that while you can point to certain people who are doing great work, the best is to actually have your own judgment and to try to find out who is actually not full of BS.

AI Adoption and Implementation

um and it's it's the to be honest you always have to back down a little bit and try to understand like what all of these things that are supposed to be happening in ai how would they change my work from this week from this year you know and this is something that we really focus at hatch is like what does the market need and what is going to be useful in the next year we don't really want to dwell into like what's the future of design in like 20 years because we might not even be there, right? So for us, it's important that people who come understand, hey, this is actually something that next year, this seems to be going into this direction.

So here's some sources for you to continue investigating. And I think it is like, especially coming back to that education point, it's like things are changing so rapidly. And it is important that everybody does educate themselves as well.

Because yeah, there's a lot of talking heads that will say a lot of things about, yeah, AI being a hammer and everything being nails. But ultimately, we do have the tools and the access to education where we can make decisions ourselves and

Coming off of that, I'm kind of curious, you've mentioned it a few times about sort of the AI talks. What would be a really good AI talk that you wish you could have at your events or a topic that you think nobody's really touched on quite right yet?

I mean, I love when it's like actual implementations of, you know, specifically right now we're dealing with all this like uncertainty on the outputs of those LLMs. And we had one last year, which was about, you know, like booking.com's like travel planner and they were experimenting. And obviously the guy was saying, hey, we think we're going to get something that you never really know. And I think that's interesting.

It's been long enough that I hope that for next year we do have actual implementations where we're able to control that or at least try to understand how that output of AI can be really used as something. And obviously, definitely the content generation and the part where Right now we're at the point where AI can see and build in a different way and potentially the implications that that has over a company's design system or even directly prompting and modifying the design.

I guess on the control, right? Because everything was kind of out of control and now that it seems like we're getting closer, I would love to see someone actually having tamed the AI. Tame the beast.

Common Mistakes and Skills for Designers

And so looking at that, what would you say from both the conference, the talks, the topics, but also again from your own work, are some of the common mistakes you might see when people do start trying to use AI, whether it's in their workflow or whether it's to complete their work, are there some similarities, some patterns that you've noticed? I think it's the expectation that you can do everything with it and that you will just have a faster approach than what you've already been doing before.

I'm someone who always tries tools and stuff and trying to build interfaces with that and sometimes doing, I haven't seen the solo stuff before, but similar things. The problem comes when you have an idea of the control that you wanna have in terms of how something looks like and what's been tested before, it's still very hard to, to actually make that work with whichever product you have.

So 1I think the biggest mistake is just to believe that there will be like a direct translation of what you're thinking and what you're getting as output. And the second thing is to believe that the tools are what's gonna bring you as a designer or as a professional further ahead. And the truth is, you should be able to do with what you have right now, with the tools that you have, and not having the latest version of ChatGPT is going to make you a better writer, etc.

It's really about also constraining yourself and not falling into this idea of all these tools that are here are just, if I don't keep up, then I won't be good enough. And I think that's something even within our own organization is we try to help people walk the line between should this really be automated? Does this really need AI?

Because, yeah, it can get tempting with the culture and the way everyone's talking about it that, OK, everything must be AI based. But a lot of times it takes way more time. It's way less predictable. And you may not even understand exactly how it's working to be able to implement it correctly as well.

So I think that is a really important point as well of like, okay, yeah, really being critical about it and thinking about, okay, what's what's the real benefits?

Essential Skills for Designers

And then looking at that, like, what would you say are some of the skills that are most important coming up for designers that they should then be working on with sort of the fact that yes, okay, the these new tools are coming along, they need to pick and choose, they need to understand how this works in their workflow. And also, of course, in the products they build, what skills do you think they should be focusing on?

Definitely critical thinking, as always. You just need to really be sure that what you're getting as an output is something that you can use.

I think you need to be open-minded enough, but the good thing about all these tools is that most of them work with a natural language, but I think you always need to be able to do your best to understand And maybe just focus on one tool that is your go-to tool and try to work with that or see what the strengths are.

And especially for those that are trying to implement AI in their current software and trying to just not go and copy how someone else is doing it, but understanding what the real value... Because it's not just about having something generate some crazy images or just having some more context on some text that one can get. It's like...

Personal Experience with AI

to really exercise critical thinking and where the value of what you're building with ai within your product really makes you know the most sense and so then one last quick question for me and then we'll open up the floor for questions there is what was one of the things you were happiest to be able to outsource to ai in your own workflow I think in general, now that I'm doing a lot more marketing side as well with the conference, there's a lot of just checking text or trying to rewrite certain things, trying to get better copywriting in general, although sometimes it just gets way worse and much more generic and boring.

everything that has to do with like operating on images, like removing specific things and things that Photoshop could not do as well. It's super helpful.

And I mean, in general, just be able to be able to like have a better grasp at the overall kind of different sources of data and things that I have to be able to quickly understand where it's important to have impact. I think those three things.

Yeah, with ChatGPT, I always have to get it to tone down the boat analogies. Like for some reason, if you don't tell it not to be too whimsical, everything is out at sea or everything's up in space. And yeah, it can sometimes cause a little bit more work before it causes a little bit of help.

There's a very interesting joke that we had this year by Emily Campbell that you can watch on YouTube. It's called Shape of AI. And basically, she was just guiding us through the history of AI and how it got to this place and what it means for patterns, et cetera.

And one thing that she pointed out is that you can... now ask the AI like how did you get to this you know which are the the tokens and the and the things that you actually follow from my prompt into this and how do you get that and it's very interesting to see like the the back side of that and she specifically mentioned the boat stuff as well as a thing of like because they might just you know the AI might to get to what you wanted might have just prompted to itself the weirdest thing that made no sense to you but it got you you know to where you wanted

Indeed. That's super funny.

Audience Questions

Is there a question in the audience, something someone would like to ask product and UX and AI? Similar to your last question, because you said that you try out multiple tools. What were some that really sticked and that you're really applying in your day-to-day work and workflows, or like other designers

It's dangerous because I try to be like, OK, let's just have like 100 euros a month maximum of subscription of AI. So it's like, if I'm going to get a new one, I need to turn one down off.

And the flow of organizing the conference has different needs. So there's a point where I was like crazy on mid-journey and just going at it. And now I don't really have a use to it because I'm not really generating anything kind of artificial. So I stopped that.

And ChatGPT is something that I always go on and off, like stop paying. And then at the same time, one thing that I'm doing a lot now is just to modify code that already exists, for example, in our CMS. And I've just been doing between ChatGPT and Perplexity because ChatGPT gets me to a point, and then it just keeps in this loop where it gets me nowhere. And from there, I go to Perplexity. Perplexity solves it. I give it back to ChatGPT, and it's able to continue.

So I kind of wish those two could talk in a way. But yeah, those two are super, super valuable. And that's a really good point as well.

A lot of times when we think of a solution, we think just going to one tool for it and being like, okay, maybe ChatGPT can help me or Perplexity can help me or Cloud can help me. But indeed, a lot of times the power comes from knowing the strengths and weaknesses of each one and saying, okay, this can help me get this far, but then I know I need to go to something else, so.

I also had one of my clients who was like, hey, I've been, you know, started experiment with like v0 and just like creating some interfaces here and I do this and I do that. And like, how do you think we can implement this and whatever? And I went and gave it a try and I was like, listen, for you might be great because you don't know how to design. But for me, it's horrible because I can't tell it what to do because it erases this other thing and it just drives me crazy. So how can we actually do it?

And what I said was like, I think you can try to express your ideas better using this because this is obviously something that helps us communicate better, but then I have to walk in and just go into an actual prototype and try to do this the right way in a way. So that's also kind of like for each role and for each person and relationship, the different tools might actually be bridging some of that translation. Yeah, not having it go to 100% necessarily, but just helping, okay, now we understand each other better in that way.

Absolutely.

Conclusion

One last quick question from anyone before we round this out, because I'm sure you can all smell the pizza as well.

All right, well, thank you so much for coming, and I was really happy to have you here.

Thank you so much.