What’s Going On WithWebflow’s AI & Dev Features?

0:21Janne Parri: Okay, Samuel. So you have said that Webflow lacks a clear direction with their AI tools. Would you tell us what do you mean by that?

0:32Samuel Gregory: ⁓ Straight in there. Yeah, so it was kind of, mean, you know, the AI bubble conversation is very real right now. A lot of stuff ⁓ is, Everyone's trying to find their place in this kind of crazy world of AI at the moment. And Webflow is no ⁓ exception to that. And it was their last Webflow Conf where they literally came out with a whole heap of stuff. They started a few months ago with their little chat thing that could build basic elements, basic ⁓ websites. And we gave that a go and it was meh, as a lot of this stuff is. ⁓ And then the Webflow Conf just gone, they just came out with like, was nothing, one, there was nothing original in that.

1:22Samuel Gregory: You in that onslaught of things, you had the classic kind of, you know, vibe code tool. You kind of had a slightly refined version of their chat interface, which they released previously. And ⁓ I mean, really it was, it's in the response from the community and the designers where like they weren't impressed with it. There was nothing, it was like, what is this? Who is it for? And there was a lot of confusion amongst the community, which again, I sort of looked at that and assessed it and thought, well, you know, the AI stuff that Webflow are doing, it's not really, it's trying to take the job away from designers that they actually enjoy, that Webflow as a tool is actually good for, which is creating pixel perfect designs from the very hands of the designers that design it. So we really empower them all too much.

2:15Samuel Gregory: It felt as though they just, again, just copied a bunch of tools that already out there and slapped it in the face of their user base and didn't really ask, what is it that you want AI to achieve for you? You know what I mean? So it was kind of an amalgamation of those things where I just felt like, yeah, who wants this? It doesn't seem to, again, it doesn't seem to lean into the strengths of Webflow. So yeah, that's kind of where I was coming out with that. That statement.

2:44Janne Parri: Yeah, yeah. But I also know that you're a developer and you have been playing around with the Webflow MCP. So if you had all the strings in your hand and you were making decisions as a CEO of Webflow, would that be the thing you would bet all your money on? Or would there be something else?

3:03Samuel Gregory: Well MCP specifically or just you know...

3:09Janne Parri: Yeah, maybe if MCP is the golden thing that everyone should do or would there be any other directions you would go into?

3:18Samuel Gregory: I mean, definitely not MCP. Mean, MCP is just another tool and I definitely think that people should consider MCP if you want to be, I was gonna say, if you want to be lazy. I don't mean if you, as that as a bad thing, like if you just understand that you can literally write natural language and that will update the Webflow CMS at the very least, like add a new blog entry and write me a blog or whatever. That's just another tool in the tool belt. But ⁓ In terms of you know, Webflow's AI ambitions, ⁓ if I was Vlad, sorry, Vlad... ⁓ if I was, you know, I would, I mean, it comes down to the users. It comes down to the community. And I would do a real reset ⁓ on ⁓ what our users want to do ⁓ with Webflow. And they know this. I don't think they, ⁓ they knew this for a long time ago.

4:12Samuel Gregory: Understand what the users want and how they see how they use AI in their current workflow. How does that fit in? And then see if Webflow can facilitate that. You know... Maybe it's to do with content writing or something like that. Know, their designers are fed up of doing lorem ipsum and then they time and time again generate content. I don't know. It's ⁓ my philosophy is always. The technology comes last in everything. Even with tools, when a client comes to me, we'll probably get into all that, but when a client comes to me, I try and separate myself from what the tool that does it. It's like, what are your challenges? What is it that bothers you? What are you constantly doing on a day-to-day basis that really costs you money or costs you time? And then it's, okay, how can the tech solve this?

5:04Samuel Gregory: So it's really just coming back to basics. What do the designers do? Where are their pain points? And then having that creative mindset to think, how can Webflow solve this and how can Webflow AI solve this? And then really draw it back to that. And I think that they'll find a lot of open doors to new features, new technologies. That actually benefit the community and their users as opposed to let's just copy Lovable and hopefully they'll come. Webflow already has a very established user base. ⁓ They can't jump onto the scene in the way that Lovable has. They've already set expectations. They've already ⁓ gained a bit of momentum and they've already got a...

5:51Samuel Gregory: Loyal and dedicated community that they have to manage those expectations and over time convince them of a new way of doing things. It's otherwise people don't like change, right? People get scared of change. So ⁓ yeah that would be my approach Vlad take it or leave it ⁓

6:10Janne Parri: Yeah, what I have talked about, obviously about these things with some webflow users and my sense is that people are actually more excited about the little things like the alt text generation and meta titles and those sorts of things. Of course there is like potential in the whole AI builder thing if it really like does take a lot of your back like

6:38Samuel Gregory: Mmm.

6:39Janne Parri: Like the manual stuff from building, but people are not confident that it will do that. And yeah, now I'm with you. It feels like that they just had to do something and they just released it. And hopefully in future it will better, don't you think?

7:00Samuel Gregory: Like I totally get we're in a really exciting time when it comes to AI and I've described myself as a kid in a candy shop with all of these tools that you want to play with. And, you know, if anyone's followed my stuff, you know, that I'm kind of the guy that tries new tools and things like that. So I've, I completely sympathize with Webflow and any other person. Wanting to experiment and seeing where this all lands. Like we're not, you know, we need to give Webflow and anyone else a break and say, you know, they're allowed to experiment and not know what's going to happen with all this stuff. So just putting tools out there and learning from the data and then refining it is a totally cool approach. I just think that Webflow have such an established user base. They've got

7:52Samuel Gregory: Thousands of committed community members that would happily talk to them and influence the tooling and the stuff that they've built. I just think they missed on that part. And to your point as well about, you know, little things like alt text, whatever, Apple are great at doing this, right? They are great at abstracting the technology and having it where it matters. They don't need to tell you that it's AI that organizes your photo library or cleans up your images or whatever. It just, it falls into your workflow effortlessly. And I think that's where we'll get to with AI. Think that again, it's such a buzzword. Investors want to hear it. Consumers want to hear it. We're still in this exciting world right now, but it will fall into that subtle little alt text analyzer or whatever, which you.

8:46Samuel Gregory: Which you mentioned and it's just a time, you know, it's a timing thing. Think we're all excited little kiddies, you know.

8:52Janne Parri: Yeah, yeah, totally. But we're diving deep into the nitty-gritty already. We gotta do little intros before we continue.

9:01Samuel Gregory: Let's do it.

9:02Janne Parri: So my name is Janne Parri. I'm a Webflow partner all the way from Finland. Yeah, that one country in the Northern Europe. And with me today is none other than Samuel Gregory. Samuel, so you have your own studio, Jupyter and Giraffe. You have a podcast with Kabarza, command plus AI, and you manage to somehow publish videos on your personal channel, YouTube channel, I mean. So do you have any time for regular old client's work? And if you do, what

9:35Janne Parri: Sorts of projects do you like to focus on?

9:39Samuel Gregory: Really interesting you mentioned that actually because you know, it is October 22nd at the moment as we as we record this I'm feeling very very much like the content is it was becoming a problem where the content was just a cyclical thing where I just needed to get stuff out there that I wasn't taking the time to develop to work with clients and work with them on a deep level So I'm re-shifting things, maybe gonna be posting a little less frequently, but also to actually work with clients and learn the new tools, use the new tools and give my honest feedback rather than a panic rush to when as soon as something gets released, I'm the first person to get a video out and not actually really deep dive into the tools. Yeah, working with clients, working.

10:30Samuel Gregory: Mostly building applications and internal pieces of software and in the world of AI just leveraging this new and exciting area where what once used to take hundreds of thousands of dollars to build an internal piece of software to help you get your work done is now something that costs a fraction of that can be produced in half the time and start delivering value. So that's that's where I'm living at the moment and I'm dealing with a few clients in that world, which range from medicinal, what do even call that, like medical industries all the way through to HR, all the way through to real estate. Like it's very varied in who we deal with, but what we're doing right now is the building of the tools.

11:17Janne Parri: Yeah, yeah, beautiful. So do you often have a client that has heard about like the vibe coding thing and then they know that maybe it's not the right way to do and then they come to you and ask that, how could we do this in a lean way?

11:36Samuel Gregory: Not, no, in short, I am dealing with someone at the moment who has vibe coded an app, who saw the bright, shiny object and has got to a point where they've realized, crap, I can't build this. I can't take this further. So I'm working with them, but how I found those were that they're literally a friend, you know, who has their own agency and all the rest of it. So it was more of a, you know, a referral partnership rather than seeing an influx of people panicking and realizing that, holy crap, I can't actually build this application. Think the theory is that there's plenty of them out there. I've just not seen them just yet, apparently, unfortunately. Maybe there's something in the fact that people who are building the tools were never

12:26Samuel Gregory: Not even in this industry in the first place, so that they're just not able to, they're not even able to think, right, how do I take this or who do I speak to take this app any further, but beats me.

12:38Janne Parri: Yeah, yeah, totally. You have clearly more sophisticated clients and if they have vibe code, they are afraid to say it out loud to you probably.

12:49Samuel Gregory: Yeah, for sure. You know, I, you know, we might even talk about it. Like I'm fully in on the vibe coding thing. Like I think there's, it's recognizing the strengths in that and knowing where its weaknesses lie and just making the best of it. You know, I don't think it's a dirty word. So yeah, all for the, all the vibe and out you can do.

13:09Janne Parri: But relating to this, I'm very curious about your current workflow. You have mentioned it in the command plus AI stream at some point, but where are you at with your own workflow one tools and how much of the AI have you gotten into it?

13:30Samuel Gregory: So you are so lucky because you're the first person to hear this, but I do have a video going out tomorrow. So technically you'll be the maybe last person to hear about it, but who knows? So I've just recorded a video on this. I'll be releasing it tomorrow. And I feel like I've got a really, really good workflow down at the moment, which is really working for me. Think it, I mean, first of all, I think we're in an age where if someone wants, you know, if someone tells you about a piece of work they want to do, or a project they want, it's so much easier just to produce that work to show them a proof of concept. With that in mind, because that's the thing, the industry is in this weird world, where it's kind of a trust me bro system of what so you can build this thing that I'm going to build and you've kind of got

14:24Samuel Gregory: Examples which don't quite match what I want to do and this and that and the other and you're saying hand me over all this. I'm like let me just show you what we can do okay and so I will nine times out of ten thus far well actually ten times out of ten for now but I'll get into why I want to reduce that in a sec. I jump into Replit and Replit is a one of these fancy vibe coding things the history of Repl.it was a community or a collaborative coding environment. I'm not too sure, but now it's a vibe coding tool. And in my opinion, one of the more filled out, fleshed out vibe coding tools, because it has such an ecosystem around it. All of the others would throw the prompt to the AI and give you back the React thing.

15:15Samuel Gregory: Replit is a little bit more involved in that and has databases. It builds databases as the front and center to the app. I stick it in Replit and you'd be surprised at how often I've done this, but I've been on the train into the office and I'll whack out a prompt on Replit, set it to build over the next 40 minutes and then get to the office and it's mostly there and then I'll start working on it. It's crazy. So that's really good for me, because then I'll spend a bunch of time thinking about the prompt and whatever. And once it's in Replit, like I said, it's really good at starting the projects. Replit is quite expensive. It is. And I have a subscription to Claude Code. I will then push it to Git and then pull it down. And then I'll actually start getting the meat and potatoes of it done inside of Claude Code. I just prefer that interface and that way of working.

16:05Samuel Gregory: I want to, the reason why I said I want to kind of reduce that is, and I think the answer lies in a tool called Taskmaster. I want to replicate Replit's AI, whatever it's doing. Want to replicate that in a plugin and an MCP series and MCPs or agents or something. I want to replicate that so I can start projects in Claude Code. Because that will save me I have a lot of time. Don't need Replit. Although I wanna support Replit, I wanna just fully rely on Claw Code or some sort of local development. And then yeah, it's through, so that's completely AI up until this point. And then working with Claw Code is pretty much completely AI. And I think the secret sauce comes into play when you're looking at the files, you know the file structure and you're referencing files that you wanna manipulate or referencing files that you wanna...

16:55Samuel Gregory: Provide context to and working in narrow areas of your product. And that's it. It's just iterating and building from there. Sometimes I'm writing code, sometimes I'm whatever, but I think I'm good enough now where I can figure out the, or I sort of know the architecture. I know somewhat how something needs to be built and I'm there managing and monitoring the AI. And it's that rinse and repeat until it's kind of done. And oftentimes I'll present that alongside the proposal to the client and say, look, this is something we've whacked up, which has taken half a day, maybe even less. And if we get the job, then it's just a case of scaling that up and being a bit more diligent with security and scalability and just looking into the nitty gritty of what it takes to take an app from MVP into production.

17:47Samuel Gregory: And that's what's been working for me at the moment. I've been really enjoying that, you know, cause my interest lies in very heavily in the UX, right? Often people ask, what am I doing whilst AI is building my code? Well, what I'm doing is actually using the app, literally some, just thinking about what's going on and thinking, that could be work better. And I'll make notes on things I want to change, or I'll write a prompt, you know, preemptively of what I want to change and whatever. And yeah, I'm just able to focus on the UX and maybe write some tests or think about other areas where in traditional development, 100 % of your time is mostly just development. There's never breathing room to just think about other things. And that's what I love about AI development is that I'm able to, my brain is able to breathe, you know.

18:38Janne Parri: Yeah, yeah, that is beautiful. And that is the ideal. So the AI could

18:44Samuel Gregory: Mmm.

18:44Janne Parri: Do all the manual work and people could like focus on what matters. But are you like close to your goal to start every project in Claude Code or are there any technical like limitations at this point?

19:01Samuel Gregory: Yeah, so like I say, the Taskmaster thing is the closest way to do this. And this is an MCP where you can ask it to generate PRD based on a bunch of requirements. And then it takes that PRD and breaks it up into phases. AI isn't, this is why I always gravitated towards Replit. AI, I don't know why it's very hard for it to start up. Project. My theory is again, you know, first person hearing this, my theory is, it's just starting with a boilerplate, whether that's the Next.js project with the base technologies that you kind of want to use so that's the context that AI has to work with. So if it's a monorepo, react with express, or if it's the Next.js app, I'll install what I need and then ask it to build my app that I want in the context of that boilerplate.

19:53Samuel Gregory: So that's where I'm thinking. I just haven't had the time to really do this. I've got three or four streams trying to use Taskmaster. And I think I was trying to understand a few things at the same time. So I really struggled with it. But I think I'm slowly getting there right now. So yeah, TBC on that one. In theory, I think I'm there.

20:15Janne Parri: Yeah, yeah, I believe quite a few outside and inside of developments are trying to get some sort of unified or ⁓ one chat and chat bot to rule them all. ⁓ But of course, I'm a web floated. So I'm curious about where does Webflow lie in your toolbox? When do you use it? And when do you don't use it?

20:36Samuel Gregory: I mean, truthfully, I don't use Webflow an awful lot. Definitely, because my interest and my love for Webflow is that I saw the empowered designers and I saw the empowered designers who don't traditionally have a background in building websites. And for better or worse, Webflow lacked the translation. Or actually, no, actually they were very, very good at it. But traditionally, organizations and businesses, business accounts for things like the Webflow YouTube or the Framer X account or whatever, typically they never get a big following. It's always the raw, well, I say unbiased, like opinionated viewpoints of the community that are able to translate, you know.

21:24Samuel Gregory: How to do certain things and give different perspectives. We can be bit more loosey-goosey with our presentation style. All that being said, my history with Webflow came from seeing that I can provide a service to the community, teach them about website concepts coming as a developer. I've always been a developer. And so that meant that I would always use Webflow to teach things and whatever. So really my usage is never by my own choice. I never promote or sell Webflow to anyone. My Webflow work comes from people finding out and hearing me talk about Webflow and saying, hey, can you work on this website or do this? And I'm totally happy doing that. So it's weird. It's very much YouTube inspired or YouTube centric.

22:18Samuel Gregory: My involvement with Webflow. And I'm happy with that. I definitely see that it has pros. If someone was to genuinely ask me about what tool I think they should use, Webflow falls into that. They just don't want a headache, basically. They want something that's hosted. They don't have to worry about the CMS. They don't have to worry about all of the bits and bobs that go through to get a website online. They can just use Webflow and they are maybe design focused. Those are clear indicators to me when someone wants to use Webflow. Or I'd recommend, know.

22:59Janne Parri: But the landscape is changing quite rapidly. So I have talked about this a couple of times on this podcast, well, firstly,

23:07Samuel Gregory: Hmm.

23:08Janne Parri: The webflow has evolved and then there's all the competitors like framers and whatnot and all this web coding things and web studio was there at some point, maybe is still. But how do you see like, what are the competitors that Webflow should be afraid of?

23:28Samuel Gregory: Yeah, so it's funny because they have made a lot of developments and they've really, I remember maybe this time last year where Webflow really separated themselves from all of the other players. Really went heavy on the, let's call it advanced website builder tool. Frame will we're going in the direction of like, let's make it as simple and as designer friendly as possible. Frame will like, let's lower the bar. Webflow push the bar. And so I really don't think that there's much competition where frame it is. I mean, I'm sure you can draw some sort of like, graph or whatever. And I think Webflow do exist in their own category.

24:11Samuel Gregory: When it comes to like who Webflow should be afraid of, the way that I think about it is that there's no other businesses that really put them, you know, any sort of disadvantage. Framer probably the closest, but I think that they have, it's, how do I say this? It's from the inside. It's like, they're destroying them self-destruction. I they should potentially be concerned with themselves of losing sight of what they are good for and what service they provide. So, you know, the AI ambitions we spoke about before being one of those things really alienating some core, you know, user base because they've just thrown AI into the faces of a...

24:58Samuel Gregory: User base that inherently doesn't like AI. Designers don't tend to like AI. If there's one criticism the podcast gets is the creative aspect. They've stolen and this and that. So they have the potential to sort of self-destruct. I think that, yeah, I think that they've really got to bring the reins back in a little bit, I think. I don't think there's any companies out there that are really competing with the same sort of standard as Webflow to build the advanced websites that Webflow is capable of.

25:35Janne Parri: Yeah, yeah, I'm totally with you on that. Of course, there's the past, in the past, Webflow advertised themselves as a tool for designers to create their beautiful designs. But nowadays, of course, Framer, as you said, has a lower bar when you come out of Figma to replicate the Figma design or even design in Framer itself, because they have this new, was it design canvas or something like pages? Yeah.

25:59Samuel Gregory: Yeah, design pages.

26:05Janne Parri: But do you think that if there is a like developer using all the coolest news to AI tools, Claude Code, cursor, replete, whatever, they could like decide to just manually or manually with an AI

26:25Samuel Gregory: Hi.

26:25Janne Parri: To code a website instead of using Webflow?

26:31Samuel Gregory: Yeah, it's an interesting one. Think I produced a couple of videos touching on this aspect, like answering that question when it comes to it. I mean, I mean, first and foremost, like Webflow is primarily a marketing site builder, right? It's great for internal marketing teams to put together websites using components and this and that and the other. Whereas code is really good for building apps with lots of integrations and custom and bespoke, you know, it's software, it's for software. No one should be writing code to build

27:08Samuel Gregory: A website, really, most people actually. You're talking to the person who's got a coded marketing website, but that's beside the point. I know what I'm doing. I'm able to just dive in and I know what I'm trying to achieve. And for lack of a better word, design is not my main focus, although we can achieve that, certainly with Figma MCP. So I know the temptation that a lot of Webflow people have when it comes to, well, coding has AI and if I'm able just to whatever, spin up a website just using natural language, then yeah, that looks really, really appealing. Then they try and implement the design and that's where a lot of designers really struggle. Like a lot of designers really wanna like, they want the UI, they want to sort of drag slides around and pick colors and various things like that, which.

27:56Samuel Gregory: Typing into a prompt box is not enjoyable for them. So, and there's no good way, I mean, there's a couple of tools coming out. I think I reviewed one of them, I've forgotten the name of it, but they describe themselves as the cursor of web design or something like that. But there's no like start it in code, bring it up and then put it into this other tool or whatever, and then you can design with a UI. Like there's no real nice, way to do that at the moment. But yeah, can see, going back to your point, I can see what the temptation is, but I don't think there's any immediate threat in that beyond just designers curious to see if they can build websites with AI. Because then you introduce again, Webflow takes the security aspect away from you don't have to worry about that.

28:46Samuel Gregory: You have to build that in with code, you really took, you know, trying to think of analogy for it. Writing code is literally just the blankest canvas you can give someone. You write, sorry, you have to write everything that you need and a lot of those things go missed. Security being one of those things. So yeah, I don't think there's any immediate threat and there's absolutely a place still for no-code builders like Framer for Webflow.

29:18Samuel Gregory: It's just a shame that they're getting pretty bloody expensive now. They get really expensive, you know, but they provide a convenience, you know, so yeah, if you want all of that convenience, the security out of the box, then you've got to kind of be prepared to pay a little bit extra for it. But you know.

29:35Janne Parri: Yeah, yeah, totally. And I'm of course quite deep into the Webflow sphere. ⁓ And of course, as I have put all my eggs into kind of one basket, I'm, ⁓ I'm asking that should I be afraid

29:50Samuel Gregory: Hmm.

29:51Janne Parri: Of, but actually ⁓ if we take a step back, like let's say designers or developers should ⁓ all of us be ⁓ someone like, ⁓ afraid. I know that you have talked about that about what roles AI is really killing. So what's your stand on that at this point?

30:11Samuel Gregory: Yeah, so I mean, obviously, especially Dario from Anthropic is throwing out a lot of scary comments. And I think we've got to write, you know, we do need to be mindful of that and concerned about that. But my biggest message really is, it's not, you know, it's not killing off jobs. AI is not killing jobs. It's not eradicating it off the face of the planet. I describe it as it's just making the pool a lot smaller. So ⁓ now ⁓ you can build a re-advanced application or a really big Webflow website. You can use one person to do that. And this is actually derivative ⁓ from a long, time ago. You would...

30:58Samuel Gregory: You know, used to be you'd have a designer and you'd have a UX person. And now you just buy, you just get a product designer who can kind of do a good job at both of those tasks. You know, people are always looking to consolidate. And now they can get a full blown development team with one developer who knows how to use AI. So the pool is shrinking. And I saw a tweet from Primeagen the other day. Quoting Dario from Anthropic saying we're 36 months into AI taking our jobs. The thing is what I, you know, I'm going to debate with PrimaGin here. The thing is, there is already business owners that are choosing not to hire people or don't have to hire people because of AI. Yes, people aren't, maybe people aren't literally being fired and

31:51Samuel Gregory: Being told that they've lost their job. Maybe that's not, we're not there yet, ⁓ but I know for a fact that people have said to me, I was gonna just ⁓ hire X, Y and Z, but I just did it myself in ChatGPT. So there's a team there that just lost their job because the project manager, ⁓ whatever their role was able just to do an AI. So when we talk about people losing their jobs, they've kind of already lost them... And... If they are gonna get hired, you need to be a kind of standout person that's really leaning into the benefits of your skillset. When we talk about Webflow, it's a bit tricky because obviously Webflow, there's nothing directly in Webflow just yet that's really helping you get your work done. We can maybe talk about how they framed, sorry, the AI assistant that they released.

32:44Samuel Gregory: But I definitely think that Webflow designers should be looking into other parts of their workflow, whether it's, again, you'll see it in my video, I talk about requirements gathering and how AI can help you pick a part and understand a requirements gathering session in order to build features. Streamlining process, your processes around maybe being directly in Webflow. To help you stand out and be a better person in other areas. I really don't want to endorse AI doing absolutely everything that you can just sit around in your underwear and not really do any real work. I think the best people are those who are going to leverage AI as opposed to replace themselves with AI. So in all that being said,

33:30Samuel Gregory: The grand scheme of the AI job market is something we need. I mean, I was talking to my mum about it last night. My mum barely even knows what AI stands for, let alone, you know, the fact that she's telling me that she's afraid of AI taking over. It's a real issue. Some optimistic aspects of that I'd love to share are that I think that we will, or sorry, there will be mass layoffs. There has been mass layoffs in a century. In Duolingo, things like people have let people go in the age of AI. But we're in this excitable period where we're going to try out what works. I think 50 % of the people that were fired, you know, now will be invited back because we'll start to realize that AI is average at best. And we'll learn what AI is really good for and what it's terrible at and think, wow, we got rid of

34:25Samuel Gregory: That person doing that role, we need to get someone back because it could never do as good job as them. So I think we're just in this unknown time period where it seems scary, but life always finds a way and we will adjust to the AI world. Maybe that's more time and more energy spent with our families. Maybe that's a whole heap of new other pieces of work open up just because AI has taken that. So whilst the future may seem scary, it's only because it's unknown. But we've encountered this issue time and time again. New exciting technology comes in disrupts the industry. We work around it. We find workflows, we find mechanisms, we find jobs, we find fulfillment in spite of, I don't know if that's the right word to say, but like,

35:18Janne Parri: You

35:19Samuel Gregory: We work around it is what I'm trying to say. Don't think in spite of is the right word there, but you know. So change is coming. It will suck. It will hurt for a little bit, but I think we'll find our place in it all. And we will know no different in 10, 20 years. We'll just think, you know, COVID as an example, you know, completely shook up the industry. We know no different now. Everyone's like, yeah, we work from home. You know, doesn't, we figure out a way.

35:44Janne Parri: Yeah, totally. There was so much there. Just to...

35:48Samuel Gregory: I hope you're taking notes. This backtrack, right? Pool is getting smaller. That's what I started out saying. I think the pool is getting smaller.

35:59Janne Parri: Yeah, one thing I picked up from the internet, cannot remember who said it, but it was about like the future role or our actually current role as humans with AI. We have to kind of be a good initiators, like starting stuff and like the finishers, like doing the little details that AI just does not do so, take the project to finish. And yeah, I'm not sure if that will be the case in a couple of years, one year, 10 years, but I really resonated with that. And I think that was somewhat like on par with your statements. But yeah, change is inevitable and

36:49Janne Parri: We are definitely living some interesting times but i wonder if everybody has thought the same way in the history maybe

36:59Samuel Gregory: It's cyclical. It's happened time and time again. Know, when the car first came out and people were on horses, when the tractor came out and eradicated a field full of workers picking corn and things like that. It's cyclical. It's the same conversation over and over again.

37:16Janne Parri: Yeah. Yeah. And I wonder if we get like even more specialized in different fields, like I already like find it fascinating how, specialized people are. Like I of course think that everybody is doing like this same work that I am. But when I go outside, that does happen sometimes. I see that people are doing all sorts of weird, very specialized. Stuff and maybe AI and other tools will just like increase that. Would you think that?

37:49Samuel Gregory: That's what I mean, right? Yeah, that's what I mean. It's like we will find other areas around, it will open up doors, will open up opportunities elsewhere. That's the optimistic way of looking at things. The world as we currently know it is not relevant in the future, in an AI future. The way that we do things, the way that we serve people, the way that we earn money, it's just not relevant now or it won't be in the future. It was going to look like a very different place.

38:19Janne Parri: Yeah, yeah. And hopefully every episode of this podcast does not have to be about AI. I did not start this as a AI podcast, but it kind of feels like, yeah, yeah, but it's,

38:28Samuel Gregory: Yeah, you can't invite me anywhere without me talking about AI. It's just I find it so fascinating.

38:38Janne Parri: It's not you, but it's like everybody, like everybody has a stand. Stand and opinion and experience with AI and it's on everyone's mind right now. But okay, yeah, go on.

38:51Samuel Gregory: Yeah, and well, just to finish on that is like this, I'm very much, I love new technologies. You can see that in my entire history on YouTube. I've always loved new technologies, but AI is different. And the, saw a future very early on. I was talking about AI nearly two years ago on another podcast that I had with a friend. We were talking about AI and I have seen a I'm ⁓ scared, the easiest way to put it is that I am scared. And the only way I feel is that I stand a chance of success in the future if anyone wants to follow me or whatever, ⁓ is I need to immerse myself in every aspect of AI as much as I can to know where it sucks, ⁓ to know where it's good, ⁓ to know where it can help me, to know where it can hinder me.

39:43Samuel Gregory: And really just go full throttle into it because I want that head start that potentially others won't have. So that's why at the moment, my whole, every touch point of mine is somehow connected with AI. So for that, I apologize, but that's the kind of reason for being. I'm not on a hype train here. I'm on a, you know, I just think it's just my instinct to wanna do that, you know.

40:09Janne Parri: Yeah, yeah, I'm totally with you on that. But of course, there's a pressure because the pace is so fast that there's like a tool for every day that one should

40:15Samuel Gregory: Yeah, it's tiring. Yeah.

40:19Janne Parri: Like look into and maybe there was some good old days where you did the same work from start of your career to the end. I'm not sure if that's the case. But

40:31Samuel Gregory: Mmm.

40:31Janne Parri: I also wonder like I still have decades of my career ahead of me, like what is the job that I will be retiring from? Do you have a crystal ball to tell me what that could be?

40:46Samuel Gregory: I'm just looking around. Think I left it. I was down the pub last night. I think I left it there, unfortunately. Well, I don't know whether you want me to go down that AI route or whether we should pull that from it. But

40:56Janne Parri: Bring it on, bring it on.

41:00Samuel Gregory: I just wonder and it seems to be that we're all becoming managers more so than anything and that we are more strategic thinkers and we're managing the AI to do the grunt work, you know? So I wonder if it involves more of that kind of stuff than actually designing. Because here's the thing. OpenAI just released a new browser, you know, and I know that not the first to do this by any stretch, but it's changing the way that we interact with the web. And on my podcast, I said that maybe the, there's no need for design in the future because the web serves as an archive to the world's contribution information and up to date, you know, real time database of things and that

41:47Samuel Gregory: AI serves as the front end and we want to know, you know, certain designer or a certain person does this and this and this. And instead of me manually going out and reviewing websites, I just get the AI to pull that together for me and whatever. You know, we don't know what the future holds. We don't know whether design will serve the same purposes in the future. We just, we just don't know. But I see a version of the world where people don't want to be sat at their computer looking through hundreds and hundreds of CVs, looking at hundreds of websites to find their ideal partner. They just want the information brought to them and all of that stuff filtered out for them because yeah, we need to rethink how we go about our day to day. Maybe there won't be a need for UI, not too sure.

42:40Samuel Gregory: But I don't think that's the answer you wanted, it's a possibility

42:43Janne Parri: You

42:44Samuel Gregory: For sure.

42:45Janne Parri: So we can basically forget about CSS at some point. Everything will be just...

42:48Samuel Gregory: Hopefully, I'm sure many people will be happy about that. But, you know, or it might go completely opposite and people want maybe experiences because especially with SaaS in the last 20 years, you look at most SaaS companies and there is a very, let's say, efficient way of designing a website or a logo or whatever. Everything's gone serif, sans serif rather. Maybe there'll be this revolution of like, you know what? We want a Renaissance period. We want everything to be elaborate. We want everything to be animated.

43:23Samuel Gregory: We want Flash to come back in the year 2030 and every website is a completely different and exciting experience. You just don't know the effect of all of this stuff that's happening and that realism, you know, we talk about AI making films and things like that. People crave that real person. Telling them this, that and the other. Maybe the weather, we're happy with an AI person telling us the weather, totally fine. But the news, I want a real person to say that. I want that person, you know, actually in the war zone or the area of the world, which is happening at, no one knows, right?

44:02Janne Parri: I, yeah, I could totally get on board with that. And

44:06Samuel Gregory: Mmm.

44:06Janne Parri: I already see like in the, like artists communities in or channels in the internet. People are like adding these anti AI tags to the, to their products, just to label that this is hand written or had made by hand without AI. Yeah.

44:22Samuel Gregory: Yeah, yeah. Weird, yeah.

44:27Janne Parri: I also totally get that serves a value. I totally respect or value that product way more if I suspect that might be AI. Yeah, yeah, totally.

44:39Samuel Gregory: You know it was created by someone. Yeah. It's just a sad state of affairs that we have to say this is not AI. You know, it's just, we should be able to feel it, right? But yeah, it is what it is and people gravitate. Like I personally don't understand it. If it's enjoyable, then AI or not, I enjoyed it. But I can't argue with people like yourself who want, who crave that human. Element to things, that human touch.

45:07Janne Parri: Yeah, everybody has to go through this like dilemma in their brain that if you enjoy, let's say you talked about AI generated music in one of your streams and if you know that it is generated with AI, does it like take off the value or can you

45:28Samuel Gregory: Hmm.

45:28Janne Parri: Just enjoy it? Of course it's cooler if you know that it's like made by humans but

45:36Samuel Gregory: Hmm.

45:36Janne Parri: You can still enjoy it if you just know that it's made by AI. That's tough.

45:41Samuel Gregory: Yeah, yeah, I mean, like we can go into the music if you wanted to like music such a big part of my life and I sent my friend this morning, Oasis actually represent Champagne Supernova soul version. And he texts back saying that this is you starting to enjoy this stuff more than the original. Now, yes, the music is AI generated, but the touch points or the connection that we have with the song is just very different now. For instance, maybe you couldn't, you wouldn't enjoy that piece of AI music if it wasn't for the original one. For a moment, for two and a half minutes, you're living in an alternate universe where the singers and musicians of Oasis are soul musicians, right?

46:31Samuel Gregory: So you're connecting with that music on a slight different level. There was one piece of music by a band called Korn, heavy metal, shouty, screamy, all this sort of stuff. You're able to hear the lyrics in the soul version. So again, you're connecting with that song on a different level. But without the original, you could never appreciate that AI music in of itself. It becomes less about the actual production and you know, the actual piece itself and more about the gateways that it opens up for you. So this sort of loosely in some weird way ties back to the way that yes, the future is very different, but we're going to approach it and we're gonna engage with it in just a complete different way. We just can't, we just don't know it yet. We just can't, we just adjust as humans. But yeah, AI music's a weird one because I'm...

47:25Samuel Gregory: Time and time again listening to more AI generated music and without the original I couldn't enjoy the AI music.

47:33Janne Parri: Yeah, yeah. Well, I don't know if you know, but Finland actually has quite a few metal bands per capita. So, so for example, the corn, yeah,

47:39Samuel Gregory: Yeah. Yeah. Yeah. Meshuggah, from Finland. Were they German?

47:46Janne Parri: I'm not sure. Or are they from Sweden? I'm not sure. But yeah.

47:51Samuel Gregory: That part of the, you know, yeah.

47:53Janne Parri: But you mentioned corn, that's part of my teenage years. So

47:57Samuel Gregory: There you go. Freak on a leash, right? Soul version. Check it out.

47:58Janne Parri: Okay. Yeah. Let's take this back to the Webflow things and whatnot. There's this new thing called Webflow Cloud. And of course, I, as a Webflow user, I'm interested, could this be something that I could use or I could like maybe build a service upon? So I believe you have played around with it and could you explain what it is and who it is for?

48:32Samuel Gregory: So this is where the audience participation comes in, right? You being

48:35Janne Parri: You

48:35Samuel Gregory: The audience, okay? So I understand it, right? I understand what Webflow Cloud is. I have a course, is Webflow Cloud, React, that kind of world teaching you all about that. Tell me, like, what do you think Webflow Cloud is for? How would you describe Webflow Cloud?

48:53Janne Parri: I you can build web applications and host them on Webflow servers. Now you just have more open place to host your, well, not more open, but more connected to your original sites. That would be my understanding of it.

49:12Samuel Gregory: Yeah, yeah, I think that's pretty, ⁓ I think it's pretty good. I, the reason why I ask is I think that I'm struggling to ⁓ answer this very question really, because I don't know the angle of the person that I'm talking with or their kind of understanding, because ⁓ there are a lot of questions. But ⁓ yeah, I think you're totally right. I think it's that unified experience from Webflow website to application. To an experience that you can't achieve inside of Webflow. The most easiest example I can give is, you might have to do this on your own computer or whatever, but the WeWork website. There's marketing pages, there's about our business, there's obviously the landing page, there's this, but then you can go into searching, you know, WeWork.

50:00Samuel Gregory: What do you call it? Like locations and things like that, right? If you were to do that in Webflow, you'd have all the marketing pages that have all of the information and all the rest of it. And then you would have the Webflow cloud side of things being that search, something you can't really achieve natively inside of Webflow. But with code, you have access to APIs, you have real-time dynamic data and things like that. And so it just feels seamless to the user. You don't know that the website is Webflow Cloud. It just is. And then you can also drive the design of a experience like that with the Webflow website via DevLink. And it starts to get a bit messy there and we can talk about that.

50:43Samuel Gregory: You know, that unified seamless experience between the marketing site and the coding site. Yeah, think your explanation is probably a lot better than mine.

50:57Janne Parri: Yeah, I'm curious, is this a feature that is, how should I say, mature, like ready to use? Because I have experience with like DevLink. We did a

51:07Samuel Gregory: Mmm.

51:08Janne Parri: One project and I basically did some components and then they were exported to the actual project that was hosted somewhere else. And it kind of felt like it was kind of working, but it was not it was not like

51:24Samuel Gregory: Hmm.

51:25Janne Parri: There yet. So how do you feel about the majority of the these things?

51:31Samuel Gregory: Why, audience participation, why wasn't it working for you?

51:34Janne Parri: Yeah, there were quite a few external things too. There was a outsourced team and they kind of like used some of the components, but then like did their own components. So the project was not managed that well. And I believe that was the bigger issue. And yeah, I just got a sense that the the actual developers did not like to use those exported components. They just wanted to build their own.

52:08Samuel Gregory: Yeah, which is interesting, you know, I think there's definitely an assessment piece there where you find a development team who are more interested in putting the puzzle together rather than designing the puzzle as it were. My belief is a good DevLink project will be that you build essentially atoms, you build the buttons, you build the headers, you build... Maybe some slightly bigger components, then the dev team just wraps that in sending off or passing in the data or doing this kind of, again, putting together of the jigsaw piece. If they're a team that doesn't want to do that, then they're probably wrong for the project. But in answer to your question though,

52:53Samuel Gregory: I'll be honest, I'm having a lot of hiccups with the whole Webflow cloud ecosystem. Certain things not working. Me having to go in and change the code specifically of DevLink components to have it work within the next.js project. I'm working with Webflow to kind of amend that, but it's because they're using create React, I think, to build their components. And it's like, well, who uses create React anymore? And oftentimes I'll follow the tutorials and again, there'll be bits and bobs that are missing. So in terms of maturity, I would say that it's probably from a technology standpoint, it's probably there, it's probably really flexible, but to efficiently work with it, you're bumping up against a lot of things. I'm trying to get the, what am I trying to do at the moment? I'm trying to do the object storage.

53:46Samuel Gregory: Or the KV storage, is key value pairs, all powered by Wrangler, which is a Cloudflare service. It's not things aren't appearing things aren't appearing on the Webflow website and it's and it's constantly hitting these hitting these blockers really and it's I can't progress as far as I'd like with my course. I have all of these half finished tutorials that I can't release just because again the documentation is a little bit of a mess and it's not quite working smoothly enough for me to convey that or to

54:21Samuel Gregory: Give you a seamless experience on working with it. Again, from a technology standpoint, like it works. I don't think there's much they need to change there. They just probably need to work on the communication aspect to it. So yeah, wouldn't be, I wouldn't be concerned with, you know, it going down or breaking or whatever. I just think it will just take a while to kind of set up your project and your little ecosystem of DevLink and Webflow Cloud and stuff like that.

54:56Janne Parri: Yeah. That is that is fascinating. I ⁓ wouldn't consider myself as a developer, but I see, I see all of these kind of like, opportunities. And as

55:07Samuel Gregory: Hmm.

55:08Janne Parri: I learned, like the basics, like HTML and CSS through Webflow. ⁓ maybe this could be the triggers for me to dive a little deeper. And

55:19Samuel Gregory: Mmm.

55:21Janne Parri: With that segue, how does code components like fitting this puzzle?

55:27Samuel Gregory: ⁓ my God, how do you, okay, how do you think it fits into this? Because this is where I

55:30Janne Parri: You

55:31Samuel Gregory: Feel like it really starts to fall apart because ⁓ Code components are great... ⁓ right? They're great. This is where I was getting, hitting a roadblock actually, because I wanted to bring code components into ⁓ a Webflow cloud project, which I think is pretty standard thing, right? ⁓ But I actually needed to edit the code, sorry, the dev link. Component coming in, I did edit the code, which will get overwritten when I next sync the project. Anyway, there are two ways to create code components, right? You can create a React component and code and then just wrap it in the code component sort of function, right? Webflow just give you a function that you provide it a React component and then you provide it configuration. And then you can push that up to a library.

56:21Samuel Gregory: That you can use in your Webflow website. Or you can, which they haven't released yet, you can develop code components through AI inside of Webflow that you can use in your project. Ultimately, they both, the code looks exactly the same. They both kind of serve the same function. The differences, which I can't confirm just yet, because I don't have access, is that code components built inside of your code base that you sync up, those are libraries. Those you can use across multiple Webflow websites. Whereas the AI version is only a code component that you can use inside of that site. Saying all that, there's nothing wrong with you copying and pasting that code into your local code base and then syncing up and having that available across multiple websites.

57:15Samuel Gregory: But that's the kind of immediate difference I can see. It's co-components just in of themselves are just a way to provide advanced functionality, like areas or sections into your Webflow website that go beyond just simple date, CMS, reading the CMS or whatever. You can interact with APIs, you can just provide really interactive and dynamic areas. It's server-side rendered, is great as well, I believe. Yeah, co-components are server-side rendered. So that's obviously something you want from an SEO standpoint and just a performance standpoint. So yeah, I think it's an exciting technology. I just wonder whether it's too convoluted because yeah, it's all using DevLink technology. It's too convoluted for people to truly understand and it's too much of a...

58:11Samuel Gregory: You know, a thing in the face of information, it kind of paralyzes you all trying to think about how you can really leverage this, but not too sure. But yeah, if you want a little section that's pretty kick ass, then build it as a co-component. You can't build that in Webflow, build as a co-component.

58:27Janne Parri: Yeah, yeah, I can attest that I'm totally paralyzed with all these features.

58:33Samuel Gregory: Mmm.

58:34Janne Parri: Well, Webflow in itself is getting like huge and very complex. There's all sorts of features like, well,

58:43Samuel Gregory: Hmm.

58:43Janne Parri: To mention few like localization and analyze, optimize and now this

58:49Samuel Gregory: Hmm.

58:49Janne Parri: Like more geared towards developers. ⁓ features that we just talked about. ⁓ kind of maybe bringing it back to where we started about the ambitions and direction of the ⁓ Webflow. What do you think that Webflow is like trying to do with all these developer features?

59:16Samuel Gregory: Hmm. I think it just comes down to capturing a market share that they don't currently have. They've got two core user bases right now. Actually, I invite you to correct me. I think they've got the marketers or the marketing teams and they've got the designers and they've never really been able to capture the developers, right? Developers like myself can kind of come in. Webflow slows people like us down with what we want to do. So they just want to capture that side of the market or more specifically, probably the app development side of the market. That's where I think they're putting a lot of their intentions. Like they've kind of nailed the UI interface. Like, there anything that you really want from a UI standpoint in web... Or sorry, is there anything missing from a UI standpoint that you want from Webflow? Probably bug fixes and...

1:00:08Samuel Gregory: Maybe little organisational things. But like, you know, it doesn't need an overhaul. There's probably lots of things in the forum that people want wishlist items that they want. But yeah, I just think they're trying to capture the app development and the developers side of the market. You know, are they at risk at watering it down potentially, watering down the solution potentially. But yeah. That's what I think that they're going for. I don't know if you agree or if you agree that those are the two.

1:00:39Janne Parri: Yeah, I do. The designers, if you just want to do like smaller scale ⁓ websites you might go with the framer and ⁓

1:00:51Samuel Gregory: Hmm.

1:00:52Janne Parri: Yeah, I see more and more people just doing like the Webflow ⁓ development part not meaning like app development just like the Static CMS ⁓ website thing and letting some ⁓ other people do the design part

1:01:11Samuel Gregory: Mmm. Mmm.

1:01:12Janne Parri: So I think that they might be watering down the design aspect or trying to go towards the marketing team. That's the core. And then with the app thing, yeah, of course they have these features though. They are definitely going for that market, but my sense and correct me if I'm wrong, that my sense is that they are not there yet.

1:01:36Samuel Gregory: Yeah, yeah. So the thing that you know, Kabarza and I on the show mentioned is that what Webflow used to be great at is really articulating who it is that they're building for. I think every company could benefit from this, right? When they're going to introduce a new product or a new, you know, whatever, is that they start the presentation with the goal, right? So as an example, so then people can clearly understand or they can maybe even zone out of the whole thing. Like let's take Webflow's latest conference. They start that whole section with something along the lines of something very straightforward as we want to engage and we want to make Webflow the place for diverse and dynamic.

1:02:29Samuel Gregory: Website applications using React or something like that. Or it could be we want to capture the developer market. I don't know, just giving you that context in which that you can go in and start addressing, or sorry, go in and start understanding whether these tools are relevant to you. Because I think again, going that, talking about that paralysis thing, people just like, whoa, there's just a lot going on here. Is it relevant to me? I don't know. So you just want to segment. Those audiences make it super clear what they're trying to achieve with it and not have people confused. They're, yeah, analyze is for marketers. It's not for you as a designer, wouldn't have to worry about it. If you're interested in learning new tools, of course go and learn it. Like, it becomes overwhelming, too much choice.

1:03:16Janne Parri: Yeah. Yeah. And we can already see that people are like specializing within Webflow. So yeah,

1:03:22Samuel Gregory: Mmm. Yeah. Yeah.

1:03:24Janne Parri: Why not to say it out loud? Out loud. Okay, we're quite in the deep end right now. There's a couple of topics. Of course. I gotta ask you what is like your number one wishlist item for Webflow. You mentioned wishlist. I'm not sure if you have checked the Webflow wishlist lately, but there are ⁓

1:03:49Samuel Gregory: No, I haven't.

1:03:50Janne Parri: Some things there. But maybe from your standpoint, what would be a feature that would urge you to use more Webflow?

1:04:00Samuel Gregory: Well, I'm not obviously looking for things for myself to use Webflow or to encourage me to use Webflow. I actually thought more about the community and the user base more than that. Things I'd noted down was, and we came up with this on the show as well, that I think a really good place for them to add AI would be for marketers to generate pages with a prompt, but using components, using a framework that a designer has put together. So you're letting the designer be their best self in creating a system that Webflow, I think the page builder that they have components, things like that. You create an amazing framework for then a marketer to come in and say, create this page, fill it in with this information. Think just to

1:04:52Samuel Gregory: That would be the best use of AI inside of Webflow at the moment, I think. Who am I? The other one, and I know they've improved this, but like unlimited nesting. I would love to know why it's taken so long to, I think they're allowed three sub nesting of, three nesting CMS collections, I think. Do you know what the number is or do you know what it is? Because like I don't think you can nest... Go on.

1:05:23Janne Parri: They just updated that and I'm currently sitting on the same number. I'm not completely sure, but yeah, that's probably it.

1:05:32Samuel Gregory: I don't think you can nest another CMS collection inside of those. You know, this is basic WordPress stuff where you just, you loop within a loop and, you know, maybe they just don't want the site to break. But even then you can provide errors and things like that in the UI. So the nesting needs to better. That was a massive thing that put a lot of, where you have to manage some really wacky workarounds to try and get CMS collections nesting and people were drawn towards the Finsweet solution, which is JavaScript, you know, it's not really nesting anything. So yeah, the other point I was thinking is just again, more from a community standpoint, just being more clear as were just talking about, being more clear with who these products are for and

1:06:22Samuel Gregory: What they should be enabling them to do. Think just overall their messaging, which isn't really a feature. It's just, wish they would come back to something like that to sort of focus more on the community. They used to be so great in the community stuff, but for some reason, I don't know. Do you feel attached to web or do you feel more disengaged from Webflow as a company now than what you did maybe three years ago? I'd like to then figure that out a little bit.

1:06:52Janne Parri: Yeah, communication is always somehow hard and you can improve on that. But I'm obviously biased because I'm in the... Workflow partner group and chats and so on. And I actually do not follow them too much in the social media. Well, on YouTube I do, but not

1:07:11Samuel Gregory: Hmm.

1:07:11Janne Parri: On X and so on. Yeah, but yeah, was going back to your first point, which was about AI and using the components. That would be very cool because now that they are pushing the components based building people are already building a lot of like components and if AI could use those that would of course be ideal but isn't that kind of possible through the MCP? I haven't played around it enough to know but do you have a sense of that?

1:07:51Samuel Gregory: Yeah, no, I don't know. I mean, it could be. I'm not sure. I haven't seen that there's any MCP, you know, functionality to put together pages using specific components. But I also feel like it's a it's I mean, yeah, I mean, at end of the day, it's using AI, it's through a prompt. So just making that more of more of making it clearer, the communication may be a bit more, but I do think people like logging into Webflow and just being in that interface. It's maybe bringing it to light through that interface itself

1:08:27Janne Parri: See ya.

1:08:28Samuel Gregory: Might be the direction that I'm more hoping they do. Because yeah, MCP is a weird one. It's still something people haven't quite tapped into just yet. It's not, know, people don't make the association between Claw Desktop. And Webflow, because they're just completely attached. Whereas when they're logged into Webflow, it's like, I know I'm in Webflow right now. So yeah, maybe they have figured that out in the MCP, but it might be something to look into.

1:08:52Janne Parri: Yeah, a little, noted detailed question. How many MCP servers have you like integrated on Claude Code or Replit or whatever?

1:08:58Samuel Gregory: ⁓ That's a very personal question, Janne. That's very... No, so... ⁓ God, this is too nerdy. It depends. It depends. I mean, I've got a bunch of MCPs, but there are certain ones that I want access to at all times when I'm coding, and that is Ref and Exa. Ref is a documentation MCP, so whenever it is that I'm... You know, using, let's say, React, yeah, React, go on React. If they add something new to that, then the LLM isn't gonna have that information. So ref always goes to fetch the latest documentation and builds code around that documentation, or at least verifies it. Exa is a search, a dev-specific search engine, which I'm...

1:09:45Samuel Gregory: Attempted to update to perplexity, it's yeah, it goes off and fetches things, but it's always in the context of code as opposed to, know, searching something and getting some result from slightly relevant, but just not helpful. So those are pretty cool. And I want to, what do you use for project management? Or do you use project management?

1:10:07Janne Parri: Well, at some point in time I used Notion, but now

1:10:10Samuel Gregory: Hmm.

1:10:12Janne Parri: I'm actually going quite lean. So I usually just end up getting myself into clients project management tools. So in Slack or teams or whatever, but I do

1:10:26Samuel Gregory: Hmm.

1:10:26Janne Parri: Have a to-do list app, actually to-do list that I use for my own stuff. So nothing major.

1:10:34Samuel Gregory: Do they have an MCP server? Does Todoist have an MCP server?

1:10:38Janne Parri: I'm not sure. Actually, I do not know.

1:10:40Samuel Gregory: See, you should look. Get it? Well again, if you want to create a bunch of to-do lists and this and that and whatever, an MCP might be the way to go. Especially, a lot of people are using their voice now to say things. So you were gonna, in that prompt box, put on your voice thing and say, create a to-do list, or like, you could rattle a bunch of stuff off. But that will be instead, I'm trying to get into linear. I'm hearing an awful lot about it. It has an MCP server and I wanna try and work that into my dev workflow. I think those are pretty much it because the thing is about MCP, especially when you're coding, it eats up context. So you don't want MCPs running all of the time. Obviously the Figma MCP, gosh, like I've got a video on that as well. This, MCP, Figma MCP is...

1:11:33Samuel Gregory: Incredible for implemented design. But again, I don't want it on all the time because I'm not in design mode all of the time. Yeah, you've got to be quite selective.

1:11:43Janne Parri: Okay, since you mentioned the Figma MCP, I of course wanna know how close are we that we can just hook the whatever AI with Figma and Webflow MCPs and then just ask it to fetch the Figma design and do it in my preferred framework in Webflow.

1:12:04Samuel Gregory: That'll be the dream, because then you've, I mean, that's what Framer is great for is that you have that design interface and then it's also a website. You're just, Figma is always gonna be the dog's bollocks when it comes to design, because that's just all they do. Yeah, I probably wouldn't think it would be that long. Like I don't know enough about the Webflow MCP to know that it's actually capable of like... Taking design or like, you know, whatever, but there must be this... I'll have to get back to you on it, because there must be a way to do it. Whether there's an intermediary to say Figma to Cursor and then Cursor up to Webflow. Yeah, probably not too long, but...

1:12:49Samuel Gregory: Are you saying then, because I asked this question years ago, are you saying that you would just completely omit Webflow and you want to just work in Figma? That you'd be completely happy just working in Figma for the rest of your life?

1:13:01Janne Parri: Not really, I prefer working in a Webflow. That's

1:13:05Samuel Gregory: Hmm.

1:13:06Janne Parri: Like my knee jerk reaction. Of course I'd like it to do the heavy lifting.

1:13:15Samuel Gregory: Hmm.

1:13:16Janne Parri: And maybe even if someone else does the design. Yeah, I'm not sure.

1:13:21Samuel Gregory: They give you a starting point.

1:13:24Janne Parri: Yeah, yeah. And a lot of the...

1:13:27Samuel Gregory: Like break point the difficult

1:13:31Janne Parri: Yeah, yeah, and of course there, I believe there are nothing towards the animations, not integrations.

1:13:39Samuel Gregory: Yeah, yeah, animation.

1:13:42Janne Parri: So hopefully that those will be updated. But yeah, again, I'm, just looking for some automation to decrease my workload on every part.

1:13:55Samuel Gregory: Have you tried the Figma to Webflow plugin?

1:13:59Janne Parri: Yeah, actually I used it today. But I just synced a bunch of variables, but it didn't work. So

1:14:12Samuel Gregory: Hmm.

1:14:13Janne Parri: Yeah, I used it back in the day and

1:14:15Samuel Gregory: Hmm.

1:14:17Janne Parri: Found that it didn't work as well as I'd like it to. So today I tried it again, just a minor like. Case added again did not work. So I

1:14:31Samuel Gregory: Hmm.

1:14:32Janne Parri: Have heard that people are actually using it, but I also wish

1:14:35Samuel Gregory: Hmm.

1:14:36Janne Parri: It would better.

1:14:38Samuel Gregory: Things about setting expectations and knowing that it won't just build your website for you but it might give you a starting point. It might do your desktop design and then you can do your tablet and mobile from there.

1:14:53Janne Parri: Yeah, I sometimes teach like total newbies to use Webflow and the look in their eyes when they find out that there is a plugin called Figma to Webflow.

1:15:05Samuel Gregory: Done

1:15:06Janne Parri: Then I'll have to tell the bad news that it might not be the thing they are looking for.

1:15:12Samuel Gregory: Yeah, that's the that's the top search probably on YouTube. If you were to type Webflow, it'd be like Figma to Webflow. People want this magical solution that just does it, you know, in their minds is logical. But from a technology standpoint, it's actually really friggin hard. Auto layout made it easier because it's basically using flex. But yeah, it's yeah.

1:15:32Janne Parri: Yeah. And I believe there are like agencies or studios named like something along the lines to Figma to Webflow.

1:15:38Samuel Gregory: Yeah. Yeah. Yeah. For sure. Yeah. PDF to Webflow or whatever, you know.

1:15:47Janne Parri: Okay. This has been a blast. We got to land this plane, but there is still one thing that we have to go through. This is a weird off topic thing. I'm not sure if you have listened or watched any of my episodes till the end. There's this little section that usually comes as a surprise for the guests. At this point at least. So what follows is that I'm gonna tell you a Finnish saying or an idiom or metaphor. I'm gonna translate it to English word for word and you have to guess or know what it stands for. Is it clear?

1:16:29Samuel Gregory: I think so, I think so.

1:16:30Janne Parri: So yeah, the saying for you is firstly in Finnish, repiä pelihousut and in English it is to rip game shorts. So what does it mean when I say that guy ripped his game shorts?

1:16:46Samuel Gregory: Okay, I think I think I get it. So game shorts. Is it that you've just worked really friggin hard?

1:16:54Janne Parri: I can see the logic in that, for some reason these things are they are not...

1:16:59Samuel Gregory: They never make sense.

1:17:01Janne Parri: Well, it kind of makes sense. It just means that the person is getting angry or is angry. So he rips his game shorts. Like you can imagine that when people play football and they get angry and then they just... Rip their game shorts.

1:17:21Samuel Gregory: Got you, okay.

1:17:22Janne Parri: Yeah, just a little.

1:17:23Samuel Gregory: Yeah, kind of, when was the last time you said that?

1:17:25Janne Parri: That is not actually too like unique, so to speak. So I actually would use that, but I cannot remember. I guess I have been around people that are not getting angry. But if you

1:17:39Samuel Gregory: There you go.

1:17:40Janne Parri: Ask a Finn, they will definitely know that one. Okay, so Samuel, if people want to know more about you and maybe contact you, where should they go?

1:17:51Samuel Gregory: First and foremost, my website, samuelgregory.co.uk, not.com,.co.uk. And anywhere on the socials with 0x5am5 should do the job. SamuelGregory.co.uk should just point you in all the right directions.

1:18:08Janne Parri: Yeah, and I obviously will put all the relevant links in the descriptions. Also the command AI and Jupiter and giraffe and are there any else? Of course, there's a bunch of AI tools that we mentioned, but people already know about those and yeah.

1:18:27Samuel Gregory: Check in the link description.

1:18:28Janne Parri: Okay. It was an honor to have you. Thank you so much. And for the listeners, watchers, see you in the next one. Bye.

1:18:36Samuel Gregory: Keep on vibing.