- Salt is drawing on his many years of games experience in new independently lead endeavours
- Catnip has been developed with the help of various AI tools such as Midjourney and Stable Diffusion
Get Industry News In Your Inbox…
Sign Up Today
The games industry has many big-name studios with eager eyes on them, all waiting to see what their next big hit could be. But giant studios aren’t necessarily the home to what could be your next favourite game.
There are thousands of independent game studios and solo developers, all hard at work crafting new game experiences with often unique ways of looking at games. In this series we aim to highlight those very games and the developers behind them.
In this interview, we spoke with Bryan Salt, who, after many years of experience in the games industry, decided to develop his first indie AI mobile game, Catnip. We discussed Salt’s history in games, the development of Catnip, and how AI tools were used in its creation.
PocketGamer.biz: First, can you tell us a little about yourself and your history with the games industry?
Bryan Salt: I sold my first game back in 1982 when I was very young, but my first full-time job in the industry was with Incentive/Dimension, the creators of the Freescape series – a very early 3D game engine for 8-bit machines.
I later became a pioneer of real-time 3D and Virtual Reality in the early ’90s and played a key role in developing some of the first 3D games and engines for mobile phones.
In 2001, I led the creative team that helped establish the first global Java standard for 3D on mobile phones. So, throughout my career, I’ve been deeply involved in pushing the boundaries of game technology.
Most recently, I worked at Kongregate as creative director of the South American studio, where I led the K Squad, a lean three-person team focused on rapidly prototyping innovative ideas.
We spoke last year about your work on Kongregate’s AI adventure creator. How was your experience working on that?
I had been exploring AI for about ten years, and my first commercial use of generative AI was in a hidden object game for the Discovery Channel called Red Crimes in 2017.
However, it wasn’t until early last year that AI became mature enough to consider using it again in a commercial title, this time for the AI Adventure Creator.
“It wasn’t until early last year that AI became mature enough to consider using it again in a commercial title.”
Bryan Salt
I’ve always been involved in creating editors and empowering others to build games, so I wanted to demonstrate how AI tools could assist more people in being creative. I developed the AI Adventure Creator in under ten weeks as an editor to enable anyone to create adventure games.
While it was simple, it showcased the potential of this technology by blurring the line between players and creators. Now, as an independent developer, I’m eager to explore how far we can push the boundaries of AI-driven game applications for both players and creators.
Fast-forward to the present day, and you’re working on your first indie AI mobile game! What can you tell us about the game itself?
When you’re looking to innovate, it’s wise to start with something simple at the core and build the magic around that. I chose one of the simplest word games – word search – and wrapped it in a “wishing” concept that wouldn’t have been feasible before AI.
In the game, players are given quest tasks, like repairing a garden. To complete the task, they search for “wishing words” that help fix it. Instead of the three fixed options typical in these games, AI generates thousands of unique outcomes depending on the player’s chosen words.
We took a simple concept and elevated it by offering a vast array of visually stunning content. Each player will only experience a small fraction of that content, which would have seemed like an enormous waste of resources just a couple of years ago. But thanks to AI, this level of customisation and variety is now achievable and creates a truly unique experience for every user.
And why choose to do a game like this with AI? Where did the ideas come from, and did any market conditions help push you in this direction?
We chose a simple word search mechanic for our first AI-driven game because the target demographic for these games doesn’t really know or care about AI. The goal isn’t to use AI as a gimmick but to create a richer, more personalised experience for players.
In modern mobile game development, the word “content” has almost become taboo. I’ve lost count of how many games were rejected by publishers who saw content as a trap and feared the need for constant updates.
“Publishers often want games that can be replayed endlessly with minimal fresh content.”
Bryan Salt
Publishers often want games that can be replayed endlessly with minimal fresh content.
However, in my experience, the highest-earning games I’ve worked on made most of their revenue from live-op events that featured unique, engaging content.
Players crave new and exciting experiences, and AI enables us to deliver that level of variety and freshness without the prohibitive costs traditionally associated with content generation.
For me, AI is not about doing the same old games for less cost but about being able to create much more unique content-rich experiences for the user. Despite its simplicity, I think we’ve achieved that with Catnip.
What can you tell us about the game development itself? What AI tools did you use?
We started our development process with Midjourney and Stable Diffusion for generating visual content. Recently, I’ve been working with Scenario, which has made significant strides this last year and has become one of the most valuable tools in our toolkit.
Moving forward, we’ll focus more on Scenario due to its advancements and versatility.
For 3D elements, I experimented with 3D AI Studio, but while it showed promise, I ultimately opted for a traditional modeller and animator for the 3D work in Catnip.
Additionally, I’ve developed custom tools to interface with the AI tools we use. Catnip uses a proprietary editor that streamlines the more tedious aspects of content creation while still allowing for human oversight and curation of the AI-generated output.
How many people are working on the title? And did you gain important feedback in the soft launch?
Catnip is primarily my creation, but I strongly believe that working in isolation isn’t ideal.
That’s why I teamed up with Roger Gilabert on the business side. We’ve been collaborating since the 90s, developing countless games together, and we share a similar creative vision. Roger’s sharp mind has played a significant role in shaping the game.
“The first MVP of the game was built in under three weeks, allowing us to quickly soft launch and gather user feedback with minimal initial risk.”
Bryan Salt
The main 3D elements and animations were created by talented local artist Diego Meza. We also used off-the-shelf assets from the Unity store to speed up development.
The first MVP of the game was built in under three weeks, allowing us to quickly soft launch and gather user feedback with minimal initial risk. We were pleased to find that the concept resonated strongly with players.
Interestingly, 10-15% of our retained users shared their “creations” on Facebook, indicating that the app fosters a sense of ownership over the content they produce.
Over the past months, we’ve refined the design and added new features based on live user data. Now, we’re confident in its readiness for a wider release.
How have you found the experience of going independent to develop a game? Market conditions are still pretty tough; how do you combat this?
One advantage of being a veteran in the industry is the perspective that comes with experience. You become realistic about the odds – only a few of the games you make are likely to become truly successful. That’s why it’s crucial to develop your idea quickly into an MVP and test it live with real users as early as possible.
This approach reduces risk and makes abandoning projects that aren’t working easier before you become too attached. The key is to fail fast, learn from it, and move on until you find that elusive winning formula.
Too often, developers cling to ideas that aren’t working because they’ve invested months of effort. It’s far less painful to discard a poorly performing game you’ve only spent weeks on than one you’ve laboured over for months.
Because of this, I firmly believe AI will bring about a new golden age for indie developers. Small teams can now develop MVPs quickly and innovate with less risk.
“This seems to be the right time to be an indie developer.”
Bryan Salt
While many large companies focus on using AI for cost-cutting, indies have a real opportunity to take the lead. It feels reminiscent of the 1980s, when solo developers and tiny teams could create incredibly diverse games with minimal resources.
This seems to be the right time to be an indie developer.
Do you have any ideas of what comes next for you? More support for Catnip Word Wishes or some other titles? Will AI play a prominent role in most of what you plan to work on?
Absolutely. We plan to start live testing our second AI game in Canada at the beginning of October, and we have a third game in the early concept stage that we aim to release in November.
“In the coming months, we’ll be seeking investment to grow our team and enhance our user acquisition efforts.”
Bryan Salt
We will maintain this fast-paced development cycle and adapt based on live results. If Catnip proves successful, as early numbers suggest, we will allocate resources to expand its content.
Similarly, if any of our soft launch titles perform well, we’ll develop those ideas further. In the coming months, we’ll be seeking investment to grow our team and enhance our user acquisition efforts. AI is an essential part of our process; it reduces our exposure to risk and provides opportunities to innovate.
And finally, is there anything else we should know about future plans for Catnip or anything we should be on the lookout for from you? Where would you like to see yourself within the games industry in a few year’s time?
User-generated content has always been a personal passion of mine. For over 30 years, I’ve been involved in creating productivity tools and editors for both in-house and end users.
It’s incredibly rewarding to enable others to express their creativity. The AI Adventure Creator was an interesting experiment, but my larger goal is to empower a diverse range of content creators to use AI for game development, helping them bring their unique visions to life.
This journey begins with Catnip, where users engage with their creations in a simple but meaningful way. I plan to build on this by offering greater levels of AI control and flexibility in end-user game creation applications, enabling a diverse range of creators to innovate and create. Expect our future soft launch titles to build upon this vision
