Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Mind the Model: Episode 011: Pip Bingemann
29 Plays12 days ago
Hosts: Nathan Guerra & Emmalee Crellin
Guest: Pip Bingemann, Co-Founder and CEO of Springboards.ai
Hosts Nate and Em welcome Pip Bingemann, CEO of Springboards, a company he co-founded with his wife, Amy Tucker, after a 2022 layoff. Initially built to address agency "pitching problems," Springboards evolved with the addition of co-founder Kieran Brown. The company uses AI as an inspiration tool, baking in variation to "inspire the most interesting idea" rather than giving the "most likely answer."
The Meat & Potatoes:
- Springboards: AI as an Inspiration Engine: The core mission is to inject variation into AI, shifting its role from an "answer generator" to a tool that inspires original ideas, countering the homogeneity of models like ChatGPT.
- The Future of Agency Billing: Pip advocates moving away from head-hour billing to value- or outcome-based models. The current system misaligns incentives in an AI-accelerated world, with pressure for change possibly coming from traditional sectors like law and finance.
- Creativity as Pattern-Breaking: Pip defines creativity as "breaking patterns." He questions the capacity of pattern-following AI models to genuinely achieve radical, unpredictable creative input, necessitating architectural changes.
Quote of the Week
"The idea of billing for time in a world of AI doesn't make any sense.” - Pip Bingemann
Mentioned Links & Resources:
- Anthropic’s Super Bowl ads that takes aim at OpenAI - https://www.thedrum.com/news/claude-s-first-super-bowl-campaign-asks-if-ads-belong-in-ai
- Springboard’s recent ad that mimics/mocks OpenAI - https://www.bandt.com.au/stop-outsourcing-creativity-to-machines-springboards-highlights-the-peril-of-ai-produced-ads-with-parody-spot/
- Check out the work of Springboards here - https://springboards.ai/ and their monthly newsletter Sparks - https://www.linkedin.com/newsletters/7393517069868228608/
- Mind the Model: Got questions? Email the show at mindthemodelpod@gmail.com
- Reach out to Pip via his LinkedIn if you’re interested in beta testing their latest version: https://www.linkedin.com/in/pipbingemann/
Recommended
Transcript
Introduction and Super Bowl Ads
00:00:11
Speaker
Welcome to Mind the Model. I'm Nathan Guerra. And as always, I'm joined by my co-host, Emily Crelland. Em, how's it going? Hello, hello. it's It's all good here, Nate. How are you? You know, it's ah it's a bit hectic in my life right now, but generally positive. Things are moving forward.
00:00:26
Speaker
I like that. And um Nate, I don't know about you, but it is coming up on the greatest time of year for my American heritage and my passion for advertising. Yeah. And we're not talking about Thanksgiving. We're not talking about Thanksgiving, which is my number one holiday. Let's be
AI Advertising: Claude vs. OpenAI
00:00:44
Speaker
clear.
00:00:44
Speaker
Super Bowl ads. Tis the season for joyous advertising chaos. Yeah. Advertising chaos. And I think last I looked it up, it 30 seconds cost $10 million dollars USD.
00:00:56
Speaker
Bargain. That is wild. um I've seen a few. have Have any popped out at you that you... I think I know what you're alluding to, Emily. You, of course, are talking about the interesting ah throwdown from Claude about no ads in AI.
00:01:15
Speaker
Yes. Yes, of course. I have seen those because they have been all over my socials. um Yeah, I mean, it's ah it's a bold move. Subtle, not subtle, calling out your top competitor. of You'll never put advertising in RAI. That's why you should go with Claude. and There's a couple different variations. We'll link to all of them in the show notes. My personal favorite was ah a man who is speaking to a woman who, it seems to be a therapist, asking how I can communicate better with my mom. And the person...
00:01:45
Speaker
replying as a therapist is very like stop-starty, like typical, great question. Thank you for asking. And then goes on to, you know, here's how you can interact with your mom better and blah, blah, blah.
00:01:56
Speaker
And if you want to speak to older women, here's where cougars, where you can date cougars, where you can date cougars. And goes on. And then the guy's just like, what is going on? And, you know, you yeah kind of feeds in the little advertising and it's just,
00:02:11
Speaker
It's such a dig at open a i OpenAI. Hilarious. I loved all of them. Look, it's a great way of attacking OpenAI right now. I do question whether Claude will be able to maintain no ads going forward forever because as we all know, ads seem to really drive most things on the internet. True. And I mean, you look at Netflix that started as subscription revenue and now they have been putting advertising and partnerships for that matter um into their content. yeah um There's actually really a really good ad as well that I saw that took a ah subtle dig at ah OpenAI i and it is actually the brand of um ah the founder that we're going to be meeting with
Interview with Bingaman: Creative Process and AI Tools
00:02:52
Speaker
today.
00:02:52
Speaker
um Well, tell me more about who we who we have today. Yeah, so um i I'll explain the ad. We'll again link it in the show notes, but effectively the ad was ah very similar to how the ChatGPT OpenAI i ad that was released previously started of we're going on a road trip, you know, where should we go? and you know, looking very like a vintage nostalgic feel. And then the like the camera pans out and you just see,
00:03:16
Speaker
so much traffic because everyone had the same idea and was endeding up going to the same place, which was very clever and on the nose from Springboards ai Today we have Bingaman, an ex-agency strategist turned co-founder and CEO of Springboards.
00:03:31
Speaker
Springboards is built to keep creative people in the creative equation by using AI to spark ideas but not replace them. With over 15 years in advertising, media, and marketing, Pip has focused on building tools that help keep people at the center of the creative process.
00:03:46
Speaker
Welcome to My The Model,
00:04:07
Speaker
where it's almost frame by frame perfect but then we change the end so instead of them going off into the sunset having this magical secret uh road trip they all end up in the same place um and and as well actually as we're going through that process we got to the point where we're like holy shit this is so damn close to what the ad originally looked like that we had to go back into production and change the artist's likeness on purpose because we didn't want to cross that boundary We wanted to cross the boundary of stepping on open AI's toes, but not on the actual the actual people. So it's very it's amazing what these tools can do. It's also scary what they can do. And I think it's horrifying that most people don't really understand how these things work, um which is the the bigger point. Yeah, we're we're thrilled to have you here. And we typically start off with a few quickfire questions. and What is your favorite AI model and what's the last thing that you've used it for? like a lot of people, I default to ChatGPT.
00:05:03
Speaker
It's not my favorite, but it's just default there. And it's like, it is the power of being a first-over advantage and awareness. It hooks you early. ah And you go for on that's a lot on the language model. but I first fell in love with Mid Journey. So the image the image side of it back in the day, that was my first love. And that's where I really opened my eyes up into shit, what these things can do.
00:05:21
Speaker
But I do default to ChatGPT. And, but, but soon that will have to be replaced because we are now just starting to build our own model. So ah that's a whole separate thing down the line.
00:05:33
Speaker
Yes. Well, I can't wait to hear about that. um Speaking of models and and how you work with them, do you have a ah prompting methodology or an approach that you you you apply?
Prompting AI: Context and Creativity
00:05:44
Speaker
Quickly, it's very simple.
00:05:45
Speaker
ah Give it a persona, give it a task, give it some context, give it a framework of thinking and giving it a a response format to come back to me. And so like, if you keep that in mind, it's pretty good. My other big hack that I tell everyone is just ask for lists.
00:06:00
Speaker
Don't ask for one answer. Ask for more, ask for more, ask for more, ask for more, because that's what these models are really good at is... working really hard. And so it's up to you to work them. Yeah, exactly.
00:06:12
Speaker
Yeah. Yeah. I've used that actually even on holiday planning. Yeah. So you get your holiday planning and it goes, okay, here's everything you can do in seven days. I'm like, great. Now give me another seven days worth of stuff because yeah so even to your ads point of view, the first seven days is exactly what everybody else is doing.
00:06:29
Speaker
We really want to know more about the origin story of Springboards. And as we mentioned in the intro, you know, it started after a career transition, but what was the specific why problem that you're looking to solve and how did that drive you to create Springboards?
Founding Story of Springboards
00:06:45
Speaker
Springboard started um with my wife and myself. So Amy Tucker and myself, Pip. um And we have spent 15 odd years in agency. So we started in media agencies in Australia, ended up in creative shops in San Francisco and then in in-house marketing teams. And so Amy was at Twitter and Shopify. ended up a brand to marketing team at a...
00:07:05
Speaker
at a Crypto Bro Web 3 startup, which I'm not going to go into because it was whatever. i'm just going to pause just there and shut up. um Anyway, ah anyway we we yeah we were doing those jobs over in SF. We moved home and we ended up settling in Noosa with a child. We were working these big global remote jobs from from the beach. Everything was rosy. But then all the tech layoffs started coming around in 2022 and we both got whacked.
00:07:31
Speaker
I like to say. So within three weeks of each other, we both got laid off from our separate jobs. We had heaps of time. um And yeah, it was it was actually really good. Like like just before we on the call, like ah I mentioned to you guys, like i was actually dancing when I got laid off. I was so happy. It was one of the happiest moments of my life. back into my marriage and like at my birth birth of my children. ah But um we, it was it was such a nice time because we actually had time to work out what we wanted to do. And so we we were freelancing and we're doing anything, like mainly just working with old clients and like freelance brand strategy, creative strategy, whatever it was. But we also had the time to experiment and AI was starting to bubble up at the end of 2022, 2023. when, ah
00:08:16
Speaker
Our first experience with like ChatGPT specifically was that it was amazing, but it was very average and it was very shitty for what we were trying to do in our own work. It never got us to good places unless we had to really work it and really spend time with it.
00:08:29
Speaker
And so when the API ah was released, we decided to teach ourselves how to code um before Vibe coding was a thing. Like we actually had to learn Python yeah because it like an API, like, oh, what's an API? Oh, you can call things. Yeah.
00:08:42
Speaker
Yeah, exactly. We were lucky that we're surrounded by great people working this little tech hub up in Noosa. A guy called Chris Bowden runs out. It's amazing. um And so we had people help us along the way. and And so we built this thing to just help us in our own freelance daily life. That was the inspiration. There was no like master plan. There was no vision. It was like experimentation and see what you could get these things to do by going to the the source versus relying on the app that you were forced to have.
00:09:11
Speaker
I love that. So you identified the need yourself as a problem, both yourself and Amy freelancing, and you just thought there must be a better way to do it. Might as well learn how to Python code and just do it myself. I love that.
00:09:24
Speaker
Yeah. And so like originally it was about us, like being able to explore more ideas. And and so where like the the problem, like where it hit us that, oh, there's a problem this can fix was the the pitching problem within agencies particularly. um Because I just think that is what's killing the industry. um You have agencies working for free, giving great thinking away.
00:09:45
Speaker
ah The clients ultimately then throw that work away. So it's like a wasted effort. It's such a, like, I think in the US, agencies spend $12 billion dollars a year pitching against each other, giving away free thinking.
00:09:57
Speaker
And that's like was the original thing was like, oh, shit, what if we could just help people explore more ideas faster and just focus on that one little thing? And then it's kind of expanded. What about the pitch process then is the thing that you're trying to solve? like what what's the What's the fix that you guys believe Springboard drops in there? its Yeah, it's evolved. It's evolved quite a lot.
00:10:16
Speaker
um And so I should mention our third co-founder because he he's the actual secret sauce behind Sprintboards and what we're doing. And so our third co-founder is guy called Kieran Brown. He actually doesn't have a LinkedIn profile. You can't find him on the internet. no I think I saw that reference actually when you guys were on the cover of like Forbes Australia some magazine cover. He started as an artist.
00:10:35
Speaker
After doing art school, he then decided to get a PhD in machine learning and then then he actually ended up at Google. Yeah. He ended up at Google, um, in the creative lab team, um, playing with a lot of these AI tools before they were AI tools. because obviously like Transformers and stuff were all kind of built out ah out of Google. um And so he also got laid off because they wiped out the whole creative tech lab team in Sydney. And so he yeah and so he just got whacked as well. um
AI as Inspiration: Moving Beyond Standard Answers
00:11:03
Speaker
But he's this brilliant, lovely person. And then we got introduced to him by a friend of a friend and started working with us. um And so his website, it's got his artworks, it's got his research papers. And so there's not many websites where it's half art, half research papers, but that's like, that's Kieran and that's what it is. That's a brilliant kind of like combination, right? It's it's such a unique and interesting kind of point of view when you can bring those two worlds together. Oh, 100%. And so the reason why his vision has changed is it's a lot to do with him and what our research team has been doing at Springboard. And so the reason for that is that we –
00:11:38
Speaker
Doing a lot more like, I'm just going say nerdy research work, um we we quickly came to the realisation that these models are all giving people the exact same answers.
00:11:50
Speaker
And it doesn't matter who you are, where you live, yeah, Yeah, it's not just it's not even that. it's seven followed by a three followed by a nine, like 99% of the time. yeah And so these models, and it doesn't matter what model you use, it doesn't matter if it is Claude or DeepSeq or OpenAI's models or or whoever's models, yeah? They all give you the same answer. So...
00:12:12
Speaker
no matter who you are, your history, where you live, all the rest of it. And so these models are all funneling people into the same place. And so our vision now has has largely evolved because of that, where it is more about how do you use these tools not as an answering tool, but as an inspiration tool. yeah And so how do you break these models not to give the most likely answer, but to inspire the most interesting idea in someone else's head? And so I guess, is that the big difference between using a chat GPT or a Gemini and springboards? That's that's ah that's the the biggest thing. So we have at the moment, last time we checked like 10 to 30 X, the variation of these models combined because They don't have interest in variation. They have interest in being the next Google, being the next front page of the internet, being the best coding agent. Yeah. And in those worlds, you don't want variation. do You want answers and you want to send people to the right spoil place. so So they're not focusing on that. They're just very different ideas. Yeah. um And so one of the major differences is is that we we bake variation differently. into the model, the architectural type layer. But then in the UI UX layer, which is just as important, it is designed around um a traditional like creative agency process or a creative thinking process. And so the tools to play with don't look like just like chat. If you want to look like chat, can. but we have like an infinite canvas where we can throw stuff on and it's got more about like pinning and highlighting and finding the little nuggets. And so the UI UX experience is very different. And then the the model layer is also different. And the model layer up being you're currently using foundational models and then...
00:13:55
Speaker
adding additional, I guess, layers of prompting and and changing the temperature and doing things to it to make it fit for your purpose. Is that essentially? Correct. Yeah, so correct. So at the moment, we use like eight different models. um And we play around with some of the controls that you do have. But, you know, your temperature scores, your top p yeah type metrics, they don't work that well for driving variation because they they fall off a cliff very, very quickly where they go into gobbledygook. And so so so there is a limited amount of variation that you can drive at both the the models and the controls that they give you. You can drive some more variation by having different models as well. um
Human Input vs. Agent-Based AI in Creativity
00:14:32
Speaker
But historically, the way we've drive variation is it is at the prompting layer. And we do... we we We basically go through prompting techniques in the background that the user doesn't have to know about to to push that variation further and further and further. um But that is, a if I'm really honest, it's a short, that is a band-aid solution to the problem.
00:14:52
Speaker
Mm-hmm. And so this need for ah models where which have variation baked into them is is is like mega important. So that's why we've been doing that part of it. That's why you're building your own model. And look, ah it's not a foundational model. Like if we were trying to build a foundational model, I wouldn't be on this fucking call because it it takes too much time. Because you need billion dollars. or And you need a billion dollars. Exactly right. And so, yeah, yeah it's definitely not foundational model. It is playing with some of the open source. One thing that you mentioned that I find to be quite interesting and is
00:15:23
Speaker
saying that like you almost bake in the kind of the different roads and different paths a user can take behind the scenes. So it's not like you necessarily need need to do a lot of like prompt engineering and context engineering training to these teams. Yes, and.
00:15:41
Speaker
Yeah. And so like the challenge that we have is that um the the the job that ChatGPT and Anthropic and Gemini have done is, and the beauty of chat interface is how simple it is. Mm-hmm.
00:15:53
Speaker
ah And originally we hated that, but now we've had to come around to it because like there's a challenge of user behavior versus understanding the need and and when you need both of those things. And there's less things you can do when you're in a chat interface. And so so when we can, we can more heavily control the prompts in these, what we just call them, we call them springboards in the tool. um but So that is not chat-like.
00:16:17
Speaker
Um, but it gives us a lot more affordances to drive variation within those two things. And so it's kind of about balancing act between the two simplicity versus variation. the need for answers, the need for inspiration. When do you want those things? One thing that's been very hot on the news lately is agents. And I feel like Agentic AI ah was spoken lot a lot last year, but I feel like this year, especially with, you know, Claude Bot, Moltbook, whatever's going on in the world. and What role do agents play in the Springboard stack in what you guys are building? a Zero at the moment.
00:16:50
Speaker
Really? we've We've been anti-agents for a long time. um i think they're getting better now and we probably do need to start thinking more about that. But but the reason why I've been anti-agents is that... um you particularly like ah if it's ah if it's a very basic request, there is a type answer thing. These things are probably pretty good, yeah? But whenever there is a need for subjectivity, tiny decisions and branch off
00:17:22
Speaker
into very different directions, yeah? And so this idea having an agent which has a little bit more autn autonomy, can't even speak, and the ability to like
Springboards' Role in Creative Agencies
00:17:34
Speaker
go back and forth and prompt between each other, like one slight decision in a complicated decision-making process or something that requires objectivity ends you up in the wrong space, Right, so you're just really highlighting the need for human in the loop almost at every single stage. Yeah. Yeah, and so we design around people. That's what we do. We use AI to help explore ah ideas.
00:18:01
Speaker
Get out of the way whenever the decision needs to be made because from all the research that we've done, it' these these language models are not good at running. Like an LLM as a judge is is, we've tried that multiple times, and it just gives you, ends you up further down that path of averageness. um And so that's in our use case. I see in a lot of other use cases, agents make 100%. Especially with creative thinking, like that's what is, creative is innately human.
00:18:32
Speaker
And I'd like to make that statement, but you know perhaps in years to come, maybe creativity is not innately human. But what you guys are solving for right now is maintaining that human in the loop and that autonomy over the process.
00:18:45
Speaker
Yeah, I don't know. i always hope that creativity will always be innately human. Like, that's my hope. Whether or not that doesn't end up that way is another thing. I just think, like, my definition of creativity, and everyone's going to have a different definition, but my definition is anything that breaks you out of a pattern.
00:19:07
Speaker
And I think if based on my definition there, i think why would we expect if creativity is about breaking your patterns, and you can argue that it is not that, yeah, but if it is, why would a model which is very good at recognising and following patterns be any good at breaking patterns, yeah? They just don't go together. um Yeah, and then the other half of that is the human part, which is like the the the purpose of creativity is a very human thing which involves human emotions of having your heart broken or surprise or joy. And again, these things don't really understand any of that unless they are secretly sentient and we've got no idea and they will kill people if you just threaten and to shut them down, which is was what was in the news. It's fascinating the work that you guys are doing. um
00:19:58
Speaker
Just thinking about who your ideal clients are, who who do you find to be the super users of Springboards? Are they more heavily geared towards creative agencies? Do you see a lot of appetite coming from brands themselves or media agencies? Yes. So if historically creative agencies and specifically strategists within creative agencies. um I think ah my my background is being an ex-strategist and I think we kind of built it for ourselves. But I think the other reason for that is I always found myself as a strategist being a frustrated creative. Yeah.
00:20:33
Speaker
where I always had ideas in my head, but I could never bring them to life in the way that I wanted to, or i couldn't communicate them clearly. um And so these tools allowed me to do that. And so I think a lot of strategies feel that way. And so now you've got tools where you can you can stress test your strategy to see what it might look like and use that as a way of communicating. So i think that's...
00:20:52
Speaker
where is that kind of the core audience? Um, but, but it, it does evolve beyond that. So there are media agencies that use it. Um, there are, yeah, we but just started like chatting to in-house teams as well, uh, from around the world. So in-house like creative teams, studios, um,
00:21:08
Speaker
media agencies, but that core user that I think like the super user is like the strategist who wants to explore lots of territories, explore it where they could go. in As an Australian-funded company, I would assume a majority of your users are Australian-based. About 40% is Australian, about 60% is probably overseas. so it's probably like, oh, maybe it might be 50-50 now. Yeah, yeah. So we've got teams in New York, teams in London, and so they're like sales and customer success. and so So they're the main markets, U.S.,
00:21:35
Speaker
ah Europe, but it is primarily London, Australia, but then we have people in Singapore and Amsterdam and New Zealand and so there's there's countries all over the all over the place. Shifting gears a little bit just in terms of How do you see this impacting the evolution of kind of creative agencies? I don't think we're going to impact the evolution of creativity, is like us personally. Okay, war well, all right. Well, let me ask a different question then.
00:22:00
Speaker
yeah Looking at the next five years, what do you see the evolution of creative agencies being? The obvious answer is smaller teams. Yeah. And I think it is smaller teams that can do more. Now, um because I think that's what these tools do. I think the tools the people that these tools benefit are really great generalists.
00:22:19
Speaker
But understand good advertising, yeah. And you can be a great strategist, but you also understand what good creativity looks like. Sure. And you can be a great creative who just, like, focuses on creativity, but you also understand what strategy, like, good strategy is going to start idea. And I think what these tools really allow people to do is to, yeah, have their perfect swim lanes, but then stretch.
00:22:38
Speaker
more into other areas within specialty of advertising and marketing. And so what that naturally does is means that you do need smaller teams to get equivalent outputs or better outputs. Yeah. Now, smaller teams doesn't necessarily mean smaller businesses, I should say.
00:22:56
Speaker
And so i think there are some really, really large companies which are going to struggle because if they have to now change their their their entire business around smaller teams, that's just a very hard thing to do. And so I think you'll have a ah a world where there is smaller teams but not necessarily small
Value-Based Pricing in the AI Era
00:23:18
Speaker
businesses. There will be some that lose. There will always be winners and winners and losers. um I think ah at the same time,
00:23:25
Speaker
ah Brands that couldn't traditionally get access to great creative thinking will now be able to get access to great creative thinking. And so there's downward pressure on prices then, but price will like the ability that you're probably able to charge a client will go down, but then you'll be able to service more clients. And so I think these things will just balance out naturally. And so is that what Springboard is? Is it a...
00:23:52
Speaker
A creative accelerant is a a creative efficiency driver. It's up to the user on how they want to use it, if I'm really honest, because you can, if you want to do it quicker and faster, you can. It's not the way that I love to talk about it.
00:24:05
Speaker
What I love to talk about and these two these two things are very connected. I like to talk about exploration. yeah If I can do something faster and quicker now, it means I can explore more ideas and more territories. And in that exploration of the work and how it could manifest itself, you will get to better work.
00:24:23
Speaker
yeah And so i could still take a week of thinking, but instead of coming up with one or two different directions, I can come up with 10. um It's still a week of thinking, but i've've I've expanded my thinking and I've now gone down more rabbit holes to get to a better outcome. yeah um And so so I think this idea of efficiency and speed um is one part of it.
00:24:49
Speaker
But in saying all of that, there is something to be said about the ingredient of time to get to good work. And so just because you can get something faster, there's something about stewing on an idea and a problem not trying to get it done fast, that has an unknown value. And I think this is true in a lot of different fields and disciplines, everything from making bread to coming up with advertising ideas.
00:25:21
Speaker
Agencies are so traditionally time-based in those billing and procurement systems that are just looking at time as the output, and it's not necessarily an outcome-driven system. How do you crystal ball thinking, how do you think agencies will begin to change over the next few years? Okay, I think they have to move away from head hour billing.
00:25:44
Speaker
um But I also think this is true of not just the advertising industry, but every industry in the world. The idea of billing for time in a in and a world of AI doesn't make any sense, okay? Because it is not representative of the value that you provide.
00:26:01
Speaker
And so the other thing that I hate about billing for time is that the incentives are misaligned. Because if if if you're charging me by the hour, your incentive is to go slow. And my incentive as a business operator or an owner or client is I want the work fucking tomorrow. But then if you can deliver it tomorrow, that means you're going to get paid less And so this idea of billing for hours, it has got misaligned incentives. It doesn't make sense. Almost no agency has has successfully been able to completely cut themselves off from the hourly way of working. And that's partially because of procurement teams and everything else. So like practically what is that? Is it is anybody doing it right or? Well, I think it's going to be a slow burn. I think when it the time that when people start accepting value-based pricing is when they start doing it themselves. Okay. okay And so the world is really run by ah lawyers and finance people. and no by the hour Yeah. You know what AI can do? It can do lawyer and finance work really quick and they're all being billed by the hour. And when they start finally going, oh shit, we can't keep billing by the hour because people go to Chachamity and get fucking legal advice really quick, then they're going to go, it's value-based pricing now. And so I think that pressure will actually start from the lawyers, the strategists, the procurement people that don't need hours and hours and hours to go through sifting through documents now, go, oh, click button and they have to
00:27:18
Speaker
like talk about their own value internally. I think that is the way it's going to happen. It's going to be slow, but, um, that is, that is the gamble that I would
Tailoring Springboards for Creative Output
00:27:26
Speaker
have. you would say that chip has to originate from the very, very traditional time-based industries. it's It's very hard to be the first in anything. It is always hard to be the first. Um,
00:27:37
Speaker
But, you know, if I if i was a client, actually, funnily enough, before we started Springboards, we had an idea. And actually, our still our website domain that people people on our Google domain drive isn't Springboards. It's 1010.au. And the reason why it's 1010 is that when we were freelancers, we said, we'll do anything in 10 days for 10 grand. I don't care what it is. I'll get it to you in 10 days for 10 grand. If you want a brand strategy, 10 grand in 10 days. you want a creative strategy, 10 grand 10 If you want a media strategy, done. 10 grand, 10 days.
00:28:05
Speaker
10K, yeah. And and so i think theres you can do that. I know I can i could do that without AI. i can definitely do it with AI, yeah. And so this idea of going, well, cool, in a world where the world is speeding up and there is quarterly reporting and people want faster, you lean into that pressure.
00:28:21
Speaker
You lean into like, shit, you want this tomorrow, we can get it to you tomorrow. This is what it's going to cost. Well, how do you think that, I mean, it's the triangle that everyone knows, the good, fast, cheap. Like how, how does AI play into that? Cause could you also do cheap and it will be fast. Like, do you think that triangle is starting to collapse? Maybe it breaks it to your point, Nate, as well. Um, I, I, I don't know. I don't know. the honest answer that was built in a time without AI. And so I know that is true or not.
00:28:53
Speaker
Yeah, that's true. You can argue the toss. It's an interesting one to consider though. Right. I mean, I'm going to have to ponder that one myself. i So pivoting back to the platform, ah we've done some research and we've looked at past interviews you you've done and you've discussed having two different models of springboards, you know the off the shelf versus the custom. What percentage of your clients are using more of that custom engagement? It's a catch 22. Yeah. Before someone wants to pay for custom, they want to trial it to make sure they like it. And then when they trial it, they want it to fit their process. They want custom before they like it. Okay. And so it does, it's a total catch 22. So not many us like literally single digit ads. Yeah. But in saying all of that, the big thing that we're actually doing that work this is like we're about to launch beta users onto now is is making springboards customizable. And so you let up to the person to customize it to their own process because this is one of the things that we've realized is we originally we were selling springboards. it It was our process. It's the way that we like to do strategy and we like creativity. And they're all, if we're really honest, they're all pretty much the same. what do you mean? The the circle and the and the infinite loop are are aren't different?
00:30:01
Speaker
Yeah, problem insight idea, and you could call it whatever you like. Four Cs or three Cs or four Ps or whatever. You can call it whatever you like. they're all pretty much the same, but still people want their own.
00:30:12
Speaker
yeah And so um really what we're doing now is that if you really want to drive adoption of a technology, that I always say, i've I've come to the realization that tech is the easy part.
00:30:24
Speaker
The hard part is people and processes. Okay. And so if you're not wrapping around the process, it's very hard to get the people on board. um And so customization is at the moment one or two.
00:30:37
Speaker
Hopefully in the future, it's everyone. so So talk me through some customization then. So let's say I'm an agency and I am all about behavioral economics. Like that's my whole thing. So does does customization like mean that I'm adding additional information into your system or like what, what, it talk me through it. The easiest way to think about it is like, instead of going creating a custom GPT, which there's a bunch of flaws with that based on behavioral economics, you can do a very similar process within our tool, but then we drive a lot of variation of the model layer as well. And so the ability to customize that, it could be either a very simple prompt or it could be like textbooks of PDFs and whatever else. we we we can customize that perfectly around that process. And so I'll give you an example, a real example for behavioral economics. You could go input, you have a field which says the headline is grab a can of Coke today.
00:31:28
Speaker
The output could then take grab a can of Coke today, rewrite that within 10 different heuristics, behavioral economic heuristics, and so that then I can go boom really quickly. But that that would then be integrated as part of the flow. And so there's two parts of customizations. One is like the, we'll just call them,
00:31:45
Speaker
I don't want to call them custom GPTs, but it's easiest way because people understand that. Custom GPTs, but better. Let's just say that. And then the second is is ah frameworks. And because we have the infinite canvas, so think like a mirror board built into the platform, you can also have templates in the canvases. And so, you know, like old, like trend canvases that people use or, you know, audience canvas or whatever else, you can templatize. A brief template? A brief template. Exactly right. And so there's two parts of that. It's you can build your templates, your artifacts, as well as your boards that match your process. So you can add additional layers of input information. And then again, how you want that data then formatted and and kind of pulled together for you. Exactly. Right. and that makes perfect sense because ge yeah to your point earlier, every agency wants their own process, right? You've hit the nail in the head in terms of inputs and outputs. Where Springboard started was a bit of paper where we just wrote, well, what are the inputs and the outputs in a journey of going from brief to ideas? Yeah. And so input is client brief.
00:32:47
Speaker
Output is four C's. Output of four C's becomes the input into the next thing, which is a creative jump off point or the middle, center of the circle, whatever you want to call it. Center of circle becomes an input for a creative, yeah? And so you can daisy chain these things, inputs and outputs on a whole
Competing with ChatGPT: Springboards' Unique Approach
00:33:02
Speaker
different process. was just wondering, and I'm sure if you can share the secret sauce, but do you have any clients that are using springboards and have done some incredible work? Like the Springboards has been used to launch new brand strategy um with like a a massive, um you know, out of home TBCs. Can you share any clients that are doing really, really great work with Springboards?
00:33:24
Speaker
I would love to. we all like One of our values is it's not about us, okay? um There is like literally some of the best agencies in the world have been using springboards for a long time. And they've they sit they often send us like spots like, hey, see this spot? like There was one that someone sent us a while back, which was featuring Gordon Ramsay, like, see this spot?
00:33:42
Speaker
Like this idea was sparked out of springboards or is ah there was a one for a car oil ad in the US like, hey, check this spot. Do you like it? Yeah, it was inspired out of springboards. that the The best things we get though is that when we get people where it's helped them win pitches,
00:33:58
Speaker
um and And so we we, the other major thing with us is that we're not like a go use the tool, go sort yourself. We have like customer success teams and we run workshops. So we will go like, hey, we will run this workshop for you in springboards. We'll match it to your process. We'll help you on this pitch to get you accelerated, to go quicker. And then like getting the feedback that people go, oh shit, we this helped us win business.
00:34:19
Speaker
um And so like the best use case of that is one of our very first users, I used to freelance with them. And i when I built the first prototype, I said, hey, instead of paying me my day rate, like which was stupidly high for a day, get this tool for the whole month.
00:34:35
Speaker
and see how you go by yourself because this is basically me in a box at the time is what it was. I'm like, yeah, okay, sweet. And then they went from one user ah to rolling out to like 200 users in their agency. they've been with us for three years um and because it's been paying for itself in terms of helping win pitches. Who's your competition? ChatGPT. Really? Like it doesn't sound like a competition, but it but it is. um there's no There's no like direct competition that I've really seen that when I look at them and see them as like, oh, this is like trying to do exactly what we're trying to do. that the
00:35:07
Speaker
the the competition is is ingrained behaviour of thinking ChatGPT is good enough. Like literally that is like, oh, that's good enough for what I need. Let's use that. Yeah.
00:35:21
Speaker
And I think that will always be our competition. Even if a direct competitor comes up, guess who, like, even if a direct competitor of us comes up, they're not going evaluate us and the direct competitor. They might, but but really it's going to be like, shit, it's the ingrained behavior and using Jachibut.
00:35:36
Speaker
Exactly right. And so I think that will always be, and it could be, and you could replace Jachibut to Anthropic or Google or whatever else. um I think Google's made some massive strides in the last year. like definitely like most improved award for 2025. But, and they're ingrained into a lot of agency ecosystems.
00:35:57
Speaker
um But, you know, so is Microsoft with ChatChapT
Reflections and Future of AI-Driven Creativity
00:36:01
Speaker
as well. So like that that's the that's the hard is, is these megacorps that are going to, um who have got, who have got,
00:36:10
Speaker
much bigger piece the pie across the entire business infrastructure systems, not agency infrastructure systems, which are different things. And so do you see yourself like plugging in or working more closely with like a Figma or Adobe in the future then? Because they have...
00:36:24
Speaker
complimentary audiences. sir We need to get better at using plugins because um I think we need, we just need to do that better. It's just, we've been so busy trying to build our own thing first. yeah um And we do have an infinite canvas built into our tool now, which is built on the, like the same underlying technology as like a Figma. Right.
00:36:44
Speaker
but but they've got like thousands of engineers and we definitely don't. well What's your ultimate ambition for Springboards? The honest answer that I tell people is that we will do this for as long as we're enjoying it. And as soon as we're not enjoying it, we'll go find the next thing. And that's been the truth of my entire career there's been two reasons that i've left like two things that i evaluate my job on one is am i enjoying it two am i learning and when like those conditions aren't met i naturally will go and find something else i'm learning a shit ton at the moment because it's around product development i've gone from doing like creative strategy to looking at financial models and helping run products teams and everything else which has been been a really interesting journey and i'm still enjoying it and when those things are when those conditions are met then it's trying to find a way out um and so i don't know usually it's about four or five years for me though on on our picture it's your Yeah. Well, Pip, I think that's probably a ah pretty good place for us to close. I mean, it's a a perfect encapsulation of everything we've talked about. Um, is there anything that we should have asked you that we haven't asked you about yet? No, not really. Um, the, the, the only thing is that we'll have a little beta program going on in, in a couple of weeks. yeah Um, the new version of springboards, which has a whole bunch of customization of some the things we've talked about, um, today. And so if people want to give it a run, drop us a note, if the beta program still going, get in on it. Um, Do we have like a secret code we can use or anything like that? ah Just like hit me up on LinkedIn or something and we'll get you. There will be like a, there'll be a form filled. there'll be a form film on the website, I believe. um And yeah, just do that. Have a play because we want people to break it before we start charging people money for it. So um that's the way go. Well, look, we have a final question or two to ask you. So ah do you consider yourself to be an AI optimist, realist, or pessimist? Depends on the day. ah
00:38:44
Speaker
I'm going to have to say optimist. Some days I'm very pessimistic about it. My biggest fear is brave new world. and And so like this idea that we think we've got everything that we love and want, but we don't really because we've got AI creating content that feeds into a social algorithm, which decides then what we get to see and it becomes this horrible loop of like passivity when... And ignorance. Yeah. Yeah, just hitting the dopamine, just enough dopamine signals just to keep us quiet. Like, that's my fear. um
00:39:20
Speaker
But I think I'm optimistic because I think most people don't really want to live in that world. It sounds horrible. If you can finish this sentence, AI will be blank for creativity in three years' time.
00:39:33
Speaker
Unlocking. There we go. Pip, thank you so much for your time today. It's been a really very interesting conversation. I've learned a lot about Springboards and yourself. um So yeah, really appreciate you spending some time with us. Yeah, you too, guys. Thank you very much. Very much appreciated. And i hopefully get to speak soon.
00:39:50
Speaker
Yeah. Sounds good. Thanks, Pip. Thanks, guys.
00:39:56
Speaker
Nate, epic chat with Pip. What did you think? Like such a smart guy and such a down to earth guy and building product. It, I think has real legs for the longterm.
Conclusion and Listener Engagement
00:40:08
Speaker
For sure. I'm, ah I'm kind of jealous. I haven't invested in the company because I can see some Yeah. Like amazing opportunity ahead for them. Cause it's, I don't know in the market the last week, uh, software companies have been beaten up in the marketplace. Um, and that's, say you know, Salesforce and ServiceNow and all these companies that create software, but the thinking that they're going to be replaced by generic Claude, uh, Gemini, whatever, creating SaaS products really quickly.
00:40:40
Speaker
But I feel like that they have created a platform that is going to be very difficult to replicate. And it sounds like the way that it's built, it's not just relying purely on the open AI API. Like it's an all-game made in all of these different foundational models. And as we see in the market, like I think I just saw this morning, Claude Opus 4.5 got released. So they will need- 4.6, thank you. They will leapfrog Gemini and then OpenAI will release, you know, 5.5 and they just keep on leapfrogging. But if you start to combine all of them, you start to see like getting the best out of all of them, but not relying purely on one subscription either, which is fascinating.
00:41:22
Speaker
One thing I really took away from the chat with Pip, and I should full disclosure, huge fangirl over here. I've been trying to get him on the pod for so long. So I'm very glad that our mutual friend Colts had put us in touch with him. But one thing that Pip mentioned that I really, really appreciated was when i asked him about the definition of creativity.
00:41:41
Speaker
And he just said purely it's breaking out of patterns and it's capturing that human emotion. So succinctly, you know, goes back to the foundations of creativity is innately human.
00:41:53
Speaker
And we'd like to hold on to that. Can I show you my t-shirt though? Oh dear. Let's see. Everything is a remix. Which, I mean, like, because I almost believe that creativity is a and you know, there's a real, there's really very different schools of thoughts about this. And in fact, if you go to the Spring Words website, a couple of people talk about the fact that creativity is i ah combination of things that have previously existed.
00:42:22
Speaker
And that's every artist, you know, you see all these problems with artists, you know, accidentally copying a music track from somebody else or a little riff or, and so it's an interesting challenge with creativity is to defining it. But yeah, I mean, I think that is certainly one perspective on creativity and and there isn't any right answer to it.
00:42:40
Speaker
But yeah, i'm I'm a big believer in the the remix culture. And that's, cause that's why I'm a little bit of a DJ, you know, Em. It explains so much. um But yeah, it was a wonderful chat with Pip today. I'm, you know, really looking forward to next time we catch up with him and hopefully Amy as well to hear her perspective on building springboards alongside Pip. But, Thank you to all of our listeners and watchers of the pod. If you have any questions you want us to chat about or anyone you want us to get in touch with, please email mindthemodel at gmail.com. And remember, the intelligence might be artificial, but the wins are all real.
00:43:20
Speaker
Cheers. Thanks.
00:43:28
Speaker
This episode was brought to you by Testchimp. Testchimp is a web developer's best friend. Testchimp uses deep runtime context to automate your testing workflows, generate scenarios, write scripts, and fix bugs all in one place.
00:43:44
Speaker
Stop wasting time and start shipping quality code faster. Bugs suck, but finding bugs doesn't have to with Testchimp.