PROFESSIONAL PROFESSIONALS feat. Dave King

*Welcome to PROFESSIONAL PROFESSIONALS, a new curated content stream where I interview people who are extremely good at what they do, so that people building businesses who need some niche specialists to come in and help them grow, can find them easier. Enjoy! - Gaz

I met Dave King quite a few years ago: introduced to me by a mutual acquaintance Kate Dinon → spoke on a couple of events I was running, with one of those event appearances being quite prescient for this here chat: an event called ‘Machines with Brains’ which Dave nailed.

We’ve caught up every now and again, and I’m always super impressed at his curiosity, humour, and a certain rascal-ness that rascals like myself seem to enjoy hanging out with.

I had one occassion in early 2024 where Dave was once again speaking on an event of mine, so I promoted the event out to the email list I would semi-regularly promote to. Here was the response from one subscriber:

Here’s Dave, probably the most formative influence on me for turning attention to everything AI *back in the day*, that I would regularly interact with IRL. North Melbourne Kangaroos supporter too, so even with that continual angst, he is a lovely, funny guy, that Dave.

Gaz:

Dave King, welcome to the next edition of PROFESSIONAL PROFESSIONALS, thanks for taking the time, let’s kick off with a brief background of your work with Move 37, The Royals, and wherever else this career story kicks off?

Dave:

Yeah cool. So I fell in love with the internet in the early 90s and I did psychology and English stuff at uni. Then in 1994, I was working at Monash Uni, I was the only person who really knew how to use a computer.

So I worked out how to make a website, the first graphical browsers landed while I was working there. So I made the first Monash Arts website, and then worked in internet jobs, strategy, production, all sorts of stuff during the 90s.

And ended up working in corporate strategy at Sensis on the Telstra stuff.

Then someone said “you should work in advertising!” and I thought, “oh, what's that mean?”

So I kind of applied for a couple of jobs and got a job at Clemenger (actually an agency bought by Clemenger), then lasted there a couple of years then thought I'll do my own thing, so I started The Royals in 2008.

So, that was a bit of a life story but I got there quickly!

The original idea for The Royals was kind of a weird multimedia collective. No one would use the word multimedia anymore, but 2008 it was: “I wanted to make things.”

Not ads. I wanted to make stuff. I imagined that I could take a marketing budget and make substantial things: that entertain, things of utility, of substance, rather than just banners and gifs.

To a degree that worked and then sort of went into partnership with a couple of other ad guys who knew a lot more about actual advertising than I did, which wasn't hard. And that became the next iteration of The Royals in 2011 and I worked in that business for 10 years.

I mean, the T in ChatGPT stands for transformer. That was basically invented by a bunch of researchers in 2017, which is the same time as when we started. So we had that on our radar the whole time as we were kind of training our own models, building products.

And then, ChatGPT dropped and the phone started ringing off the hook with all sorts of organisations saying “how do we take advantage of AI?” And so we rolled our product ambitions into helping other organisations and we've been doing that for a few years.

I mean, the penny has just dropped. ‘This is fucking insane. I can make a no-code thing because I can just describe it.’

So over the last bunch of years I've seen a lot of people have those light-bulb moments and it's when they're in it, doing something, saying ‘holy shit, what is this magic new type of software?’ It's not like other software. It's unpredictable, it's non-deterministic, it's interesting. And then you use a no-code tool or an agentic sort of no-code tool and you think, holy shit, if I can communicate my ideas, I can have it.

So yeah, it's been an interesting journey. I think if I had my time again, it would have been good to start Move 37 a bit later, when the market and the opportunities were more mature and more emergent. But fuck, we learned a lot in that first bunch of years.

Because product development in AI is a weird beast. Because there's a tension between disciplined product development and then rapid prototyping and experimentation that you need to do when you're trying to wrangle LLMs.

Gaz:

So, on The Royals, independent agency… what’s one very hard thing from when you started it, but also super glad in regards to embracing the indie operating way?

Dave:

I haven't worked in big agencies anywhere near as much as the other people, but I think it’s very different when you basically have your destiny in your own hands. And I think that spirit is shared amongst all the people, not just the partners. For example, recently they had this big kind of round table meeting with all The Royals people about some changes happening. It's just like an incredibly flat structure.

You know, the partner’s sitting there, the CEO is there, the most junior graphic designers sort of interrogating with questions. You can only do that when the radical transparency doesn't come back to bite you and the others.

And I think that's part of the culture. But also the structural thing that you kind of alluded to for being an indie is: there's no mothership in New York telling you to spend less money or cut costs or save, it's all up to you.

Which is quite frightening for some people. Kids need boundaries, so it can be quite daunting to kind of go: destiny is completely in your own hands. On the flip-side, there is literally no one who's about to call you and say you need to cut staff.

And so I think that indie spirit and the partners, including myself when I was more active in the business, we did preemptive marriage counselling for 10 years.

As an indie, you live or die by your own perspectives and your own opinions and your own creativity and strategy. I think you're more willing to - this is the theory - it depends on your appetite for ambition or destiny at any given time, but: for indie to thrive, you need to be more willing to back yourself and go with your gut and be respectful with clients.

The Royals’ clients would say that they have great relationships because it's not like a master-slave thing. We're all grown-ups, we talk respectfully, we want to grow your business, we want to achieve revenue growth or brand sale into whatever it is. And we’ll work our hardest to do that. But also we're not going to work in damaging ways for our people. So I think you're more in control of the nature of your relationships with people, in an indie.

I think the other thing is that The Royals is probably the most awarded agency in the country for culture. Stacks of awards for that, to the point where they kind of go, ‘oh would be nice if we won more creativity awards as well.’

They've won a heap of those as well, but they just seem to nail the best places to work, the culture, all that kind of stuff, the AFR stuff, because they've worked very hard at cultivating it.

Gaz:

Yeah, very cool. I want to go to the product building aspect? I'm guessing having the indie aspect then allows you to take note of some shift in culture easier, and move with speed client-side. But on the product-building side, I'd like to just ask did you have ambitions from the outset that were realised immediately, or did that lag a little bit and you eventually got there?

Dave:

Yep that’s a good question - throughout The Royals’ history we’ve always tried to create things of value, of utility, communicative pieces like that could be in an application form, or it could be an interface form, or it could be a widget or a device, whatever. So not just an ephemeral campaign.

The ambition in Australia for clients to fund that, is not as great as other places and I also feel like over the last bunch of years - maybe kind of related to COVID - that just seemed to ingrain a sense of conservatism into how people think about the world and spend their money, even if it's not their money.

And so this is just my perspective - I'm not sure if this is true - but it just feels like there's not as much appetite for risk and adventure, as there was. But to answer your question, The Royals, we did do a lot of stuff. We created voice assistants, we've created a Hamburger phone such as from the 80s (and Juno film) that used AI to dial up a special number. We ran a hackathon for Mercedes Benz called Hack My Van and had all sorts of people creating incredible stuff with the van.

Gaz:

Wasn't Henry Rollins involved?

Dave:

He was involved in another campaign we did for Mercedes, yeah, which was Tough Conversations. And that was really interesting too, because that was, I thought, that selection of him was right on the money, because we had to pitch the ute, which was a juxtaposition of luxury and ute, and people were very skeptical of it.

And there's a guy who's traversed, you know, all sorts of aspects of culture. And he travelled around Australia, interviewing all sorts of people about what it meant to be tough in Australia, redefining what tough is. I thought that was very good.

But going back to the product stuff - I think it's a different kind of client and a different kind of marketer. And I think where The Royals have had the most success is where we can kind of go slightly outside the marketing function in a business and kind of go, “Well, hang on, we've noticed something. You could be building something rather than just communicating something.”

Gaz:

What's your general perspective of the overlap between the type of work in the technology space you’re doing, then that startup side of things which is a bit over there

So that's a world where you're not really as involved, but they’re both technology and there could be overlap. What’s your perspective on that?

Do you feel that there's a massive missed opportunity in terms of the disconnection between the two, or it doesn't really faze you?

Dave:

Like, theoretically, agencies should be very good at spinning up startups because they spend a lot of time thinking about society, thinking about trends, thinking about consumer behavior or whatever, and they’re very close to insights that normally a startup would kind of go “Oh, I've just realised I've just made this killer observation and I’m now gonna explore an MVP or a gap in the market opportunity.”

What they're traditionally not great at is working out how to fund that. And The Royals have had a couple of startups and we've learned a lot from that. And it's really about the kind of people who love coming up with stuff versus managing stuff. They're not the same people.

And I think you've got to go through that, to understand that probably most of the people who are populated in agencies really get a buzz from coming up with stuff, really get a buzz from pitching it and getting that kind of chemical rush from getting affirmation and awards.

There's no other industry that obsesses about awards as much as advertising.

So I think there should be a better connection between how startups think about tech and how agencies think about tech. But technology needs to be built, managed, run, evolved, and agencies don't often see that opportunity.

Gaz:

Yep that’s a fair view - this was raised in one of my earlier interviews with Jess Wheeler of SICKDOGWOLFMAN, where I remarked on a tale about attending a talk at SXSW Sydney with a panel of some pretty well known agency people talking about the future of agencies - including technology - but the answers all centred on how to better serve clients, as opposed to creating products in-house, which I was more interested in hearing about.

Dave:

It's a different kind of person and I think it's quite tricky. I've done it here at The Royals and other agencies have done it but it is hard because you've got to, at the end of the day, you've probably got to raise money rather than just think you can bottom up, sort of fund it.

The other thing about technology in the advertising space, and I've noticed this with AI, is marketers tend to be more operators than builders and so they are aware all of the vendors, all the software vendors, all the dashboards they use, all the marketing tech is getting AI in it.

So when you talk to them about what they could make, “imagine what you could build with AI” they kind of go, “yeah, but there's something off the shelf.”

Gaz:

I'll go some of your external involvement - I know that you mentioned about being involved with Monash University way in the past, but then you had returned to Monash with SensiLab? (a team of story tellers, artists, makers, hackers, designers, developers, musicians, coders, scientists, theorists, luthiers and builders)

Dave:

Yep through Move 37, we've done work with SensiLab at Monash - it's a creative technology lab in the sense that it's not kind of theoretical AI, it's quite practical and artistic. So for example, there was a great PhD student a few years ago, when GPT-3 or 2.5 or something was in market, where you could look in a mirror and the mirror would sort of create poetry about what it was seeing.

It's a really neat, new, avant-garde idea that now you kind of go, ‘that's what Meta's glasses can do’ or whatever. But those kind of new frontiers, that lab is really good at. Creating an AI that can paint and more. I love those kind of ideas. They're quite natural and seed curiosity in more mainstream kind of building.

So yeah we've had involvement with, well, The Royals have Melbourne University as a client now, but Deakin Uni continues to be a great relationship for The Royals also, and I think getting exposure to universities is a super interesting way to sort of maintain your interest in R&D and curiosity.

Gaz:

What about from a trends or keeping updated on the pulse with the rapid changes in development with numerous types of platforms, tooling, etc. What’s your process for keeping up to date and absorb the context?

Dave:

I'm still a massive fan of RSS feeds, and as much as I don't love the values of the platform, it's hard to break away from Twitter for R&D insights in AI specifically. Following individuals who are publishing and making things…following along on Bluesky or Twitter are very good for staying very close to the source.

We share a lot in Slack and Pan (Demosthenous) our CTO at Move 37 has made a little agent called Slinky, ok, and Slinky you can ask: “What was that thing we posted six months ago about? You know, about reinforcement learning?” and Slinky talks like a character out of Red Dwarf, a famous episode - well, maybe it’s not famous - an episode about a talking toaster.

And it's a really fucking annoying little toaster called Talkie, but Slinky is a little similar, with it’s little quips, and will then link us to the knowledge we need recalled for our Move 37 Engineers.

Our engineers love to share the latest research papers, cutting edge bits and pieces, practical examples of AI models being applied, which is very different from benchmarks and research.

Gaz:

Awesome. Well, maybe on the translation of knowledge front, regarding emerging talent: What's something that you enjoy doing or a particularly useful way of communicating to the emerging talent that you might work with directly (or in a public forum), first and foremost amongst your team?

And then some advice for those looking to tinker, experiment and get creative with getting their head around the tooling but sort of like having this permissionless base to just trying things and moving with the technology.

Dave:

Yep, there’s a things. We run AI Advanced Party Parties, which is a four hour thing on Fridays, at The Royals. So it's a sort of collab between Move 37 and The Royals. And that is getting small teams together practically, juniors involved, who might not be cross-functional, making things with AI. Starting with an idea, an insight, and translating that into, “okay, what's the idea and why would that matter?” And then start prototyping experimentation.

There's no substitute for doing rather than just thinking about it and so we've all got these ideas 

Because as soon as they're in it, like making something with Replit, as soon as they're in it, that’s when they kind of say “I'm feeling the interaction with the language model of the technology.” It's not hypothetical anymore.

I really like Build Club too, I think that crew have done a great job of cultivating a bunch of people who communicate within Slack, and who are trying things, experimenting, doing little challenges, doing networking stuff.

What is a real thing for entry level people is: it’s going to be harder to start with a speciality. You should probably think of yourself as a generalist, and then the great opportunity with that is if you get a job, and I think these people can be very attractive, is that you can do a lot of things.

You don't have to just learn how to edit film. The tooling is progressing so quickly that you’re going to have this kind of toolkit that is all your own.

I say to my son who's 22, who is into music and he's studying film & TV at Swinburne, you're gonna be able to make your own tools. You're not going to have to rely on Adobe Premiere or anything or similar tools for music.

You're going to be able to go to an AI and say “I really need a tool that takes an edit and adjusts it for sentiment” or whatever it is. Just describe it and you can have it. So we're not teaching people how to do things, we're teaching people to communicate their ideas, how to lean into their sense of agency, how to realise that no one's stopping them and they don't need permission.

They're not even soft skills, they're like personality traits we need to bring out of people. And I don't know whether that's coming out of high school or university, I doubt it is, because those, especially universities, tend to kind of tip people into being a specialist. I think there's going to be a new line of education at universities.

Gaz:

Yep very interesting times. Yeah, a high irony discussing the use of AI agents, etc. whilst referencing the need of people to initiate their own agency…

Dave:

Yep the term AI agent is a little bit, or quite a lot frustrating in a way, because the way we're creating agents at Move 37 - AI agents - is putting them on rails.

There are going to be general purpose agents, like Deep Research (OpenAI) -  those kinds of things that can do a lot of things. They're going to be crowded by a few companies, but for the rest of us, we are going to make these super niche specialist agents who are on rails, are fine tuned to a task, and can only do a couple of things, but they're really well and really reliable.

And then there's an orchestration layer that sits across the top of that, that interprets a human query or question or idea and then invokes the different agents to play a part in satisfying that query, or creating something, or making something, whatever it is.

And so yeah, the idea that any AI agents have agency is an oxymoron. Because they shouldn't have agency. They've got too much agency. They're going to run amok.

Gaz:

A little point blank, a little out of the blue: even with high ambitions, are you satisfied and happy with how Move 37 has progressed since starting to where it's at now? I know you mentioned before that you wish you'd potentially waited just a little bit for maturity in the world, but no doubt you were sort of quick to market also and with the established standing in that world communicating about it for years?

Dave:

I'm hungry to get more about our ideas out there. What I really like, and I talked to Pan who's my co-founder and business partner at Move 37, and our CTO - I talked about what I really like about Move 37: that we are remaining in our lane.

We are very strong at creating AI products for ourselves and for other people.

At the moment, we're predominantly building for other people, but of course, we're building things for ourselves.

And we're very strong at wrangling large language models and the engineering around them, including agentic AI now. And that's the thing that we did a lot of R&D around. We started off with computational creativity and linguistics - we had a computational linguist on staff at Move 37 for a while. Super bright guy. And that was because it was at that era of, I guess, where the models were at.

Now, yeah, we call ourselves an AI product studio for the ambitious. And the kind of people who are drawn to that, the kind of client partners and staff who are drawn to that, feel like ambitious people who want to create incredible things. And so building a reputation for that is different from being an AI agency that's making ads or making marketing. That's not our space.

We're very interested in sort of capitalising on this incredible opportunity that for medium to larger organisations, is a transformative once-in-a-generation kind of technology. And when I say one, a technology, it's not one technology, it's like a bunch of different technologies.

But it's also not even really once in a generation, is it? I know people sort of compare it to the arrival of the calculator and how that didn't stop maths or whatever, but I think it is close to the kind of feeling, I remember showing people the internet in 1993, I remember showing my mum's cousin who was an actor, and she's just like, what the hell is this? You're dialling into a university across the world and you're browsing through their computer and looking at images and bulletin boards and stuff.

I haven't seen that as much in the last 20,30 years or whatever it is, up until now. You show people things and they just go “holy shit.” It's a cultural and transformational kind of change management challenge more so than a technology or equation as a technology challenge, but the shape of industries the shape of sectors will look very different in the next couple of years. Anyone who is making something is looking over their shoulder.

Gaz:

Last one Dave, you’ve been at it for a while: the a-ha moment for companies you work with?

You could show them the world and you could sort of explain and go through the scenarios until the light bulb moment hits. So, what are better questions that they could be asking from the outset as opposed to you justifying or explaining it to them, so that they get it?

Dave:

Yep that's a great question, the thing we bang on about - because we've been working, both Pan and I and other people in Move 37, we've been working in sort of consulting for big business for ages  -  it's basically when we start to talk about the AI opportunities, our clients tend to fall into two buckets:

One is how do we take advantage of this? It's a very open-ended question.

And the other is: “I've got an idea. Can you help us play this out, potentially build a prototype?” Whatever it is. The most fundamental thing is the same as any other kind of form of working with a client is: what's keeping you awake?

Most of our clients, the more sort progressive, sophisticated organisations, kind of get that the AI technology is profoundly innovative and can make dramatic shifts in the way of their business, but it needs to come with all those other things, like people and culture and change management and all those things. In answering your question in brief, we try to say, “let us in, let us in to what's in your head as far as what's motivating this idea, or what's motivating this interest in exploring AI.”

And just by virtue of working on a stack of different AI projects over number of years, we're able to go “that one: that challenge is something that's really interesting” and we can list a bunch of examples or case studies. “That one is going to be very hard. You might get 80 % there, but the last 20% is going to really test you.”

You are literally providing a new tool that lets them do something that they couldn't do before. So that's, I find that really exciting. So any kind of approach that's like, we have the opportunity to provide tooling for people to enhance their life, give them superpowers, is my favorite stuff.

Move 37: https://www.move37.ai/

Follow Dave on LinkedIn: https://www.linkedin.com/in/iamdaveking/

The Royals: https://www.theroyals.com.au/

MORE GROUPED: