With decades of experience in the gaming industry, Jani Penttinen played a role in developing extremely popular games while witnessing the industry go through multiple transformations. His latest project, Bitmagic, is on course to be popular and embody a massive shift in the game industry. This is because Bitmagic is an AI gaming platform that enables players to create new games using natural language. Wanting to learn more about Penttinen’s background and Bitmagic, as well as his thoughts on the gaming industry, I was able to interview him for ScifiPulse.
Nicholas Yanes: Growing up, what were some video games you loved playing? Are there any from that time you still enjoy?
Jani Penttinen: The very first video game I ever played was the classic Pong, on a black and white TV, but obviously it didn’t have colors anyway. I don’t remember when and how, but this was at my grandma’s place. Later on, we’d sometimes rent an Atari game console for the weekend, and I remember enjoying games like Donkey Kong and Kangaroo.
The first game I really fell in love with must have been Civilization on Atari ST. This is something I might still play today, as it’s a game franchise that still lives on. I also remember enjoying the Ultima games, as well as space shooters like Oids and Thrust, which were later inspiration for me to become a game developer.
Yanes: When did you know you wanted to pursue a career in gaming? Was there a moment that this goal crystalized for you?
Penttinen: I didn’t really plan on becoming a game developer. This is something that just kind of happened. I was in this thing called “demo scene” for some time. The idea was that coders were challenging themselves and each other in trying to create things that should not be possible with the computers we were using.
I had an Atari ST, and there was a rivalry between the owners of the ST and Amiga back then. One day a friend of mine was showing this really smooth space shooter duel game on his Amiga, claiming that it’d be impossible to do a game as smooth as that on the ST… challenge was accepted! I had played space shooters on the ST before so I knew they were not playing nearly as smoothly as the game I just saw on the Amiga. A few months of hardcore hacking, and I had a game called Utopos running on Atari ST just as smooth as the Amiga game had done.
I copied a demo version of the game on a disk and sent it to my friends. There was a notice in the demo that if you wanted to have the full game, please send $10 worth in your own currency to my home address in Finland. Obviously, there was no www back then. It took like a month, and I got a letter from someone I didn’t know, with money inside. Then another, and another. First, they came from Finland, then the Nordics, then Europe, eventually the US and all the way from Australia and New Zealand.
I learned that when you send cash in an envelope to some faraway country, you’re not just going to send the cash, you’re going to write a nice letter to tell how much you love the game. If you’re asking what is the single biggest thing that made me a game developer, it would be this. I received hundreds upon hundreds of super nice and encouraging letters from all over the world over the next couple of years. How could I even consider any other career after that!? I ended up being one of the first employees at Housemarque and Bloodhouse, two of the first game companies in Finland.
Yanes: You have been in the industry since the late 1990s and it is not easy to build a career in the gaming industry. How do you believe you’ve managed to survive all the ups and downs in this sector?
Penttinen: I think it comes down to being one of the first game developers in Finland. While the early years were not so much struggling for my career than struggling to keep these early game companies in business at all, it gave me a chance to wear many hats. I was a programmer, but I was also a project manager. When it was time to market the products, I would join Housemarque’s CEO on business trips to trade conferences. These days people are much more pigeonholed in their roles and they get less exposure to different roles. In my case, I met a lot of people back in the day and many of these people are industry executives of successful companies these days.
Yanes: Prior to the current wave of GenAI software, what did you think of when you thought about artificial intelligence and machine learning in game development? For instance, what was AI/ML to you 10 years ago?
Penttinen: I started getting seriously interested in AI/ML about 6 years ago when I started Utopos Games. Prior to that, I knew AI was getting better at translating text, but the methods at the time seemed to have peaked and it wasn’t clear what is the path forward.
With Utopos, we started creating a game called Raivo, where you train a robot into a fighter and have those robots fight against each other on an online arena. We’d use techniques such as reinforcement learning, which was slow and tedious and not exactly easy to run on the consumer hardware of that time. But it was a great learning experience and it in many ways made it easier for me to understand how GenAI works, and how to get the most out of it.
Yanes: Bitmagic, the company you are the CEO/CTO of, recently announced a new platform that allows people to use natural language to create video games. What was the inspiration behind this development?
Penttinen: It’s funny, actually, we started building the tech before GenAI was a thing. I was aware of OpenAI’s LLMs and I knew GPT-2 was quite capable compared to things that had been done before. And then GPT-3 came and was out-of-this world. It hasn’t taken over the world yet, but it certainly struck a chord with developers who were following the progress.
We started out with a goal of building a role-playing game where one of the players is the Game Master. The GM would need the ability to quickly create new rooms, even new worlds. We didn’t want the GM to spend time in an editor, as this would easily ruin the game for the other players, so we started building something that would allow the GM to describe what they want, and the world would be built.
The role-playing game didn’t find investor interest the way we wanted, but everyone was really keen on the technology we had built. All of a sudden, GenAI was a big thing, and while others were trying to build a better NPC generator, we were way ahead of everyone else with something that could essentially build a whole game, including those NPCs, from a prompt.
We ended up pivoting the company direction, dropping the plans to create an RPG and instead focused our efforts into building a gaming platform where anyone can create a game without an editor. There is some serendipity there, we sort of stumbled into a great idea without thinking of it, but we found ourselves in a position where we had built an ideal team to build just what the world will soon be looking for.
Yanes: As a seasoned veteran of this industry, when did you realize that the current wave of natural language processing and artificial intelligence was going to be a big deal? Was there a moment in which you felt the ground shift beneath your feet?
Penttinen: When GPT-3 was available, I saw this was going to make a difference, but at the time I had no idea how big it will be. It was not until GPT 3.5 and ChatGPT when it became clear the advances in LLMs were so big that this has a potential to change everything in the world. So, as much as I would like to say that I saw it before everyone else, I feel like nobody saw this coming. We were working on GenAI before others, not because we knew it would be super-hot, but simply because I felt it was the best solution to our problem.
Yanes: As you bring Bitmagic’s latest project to market, what are some unexpected obstacles you have experienced?
Penttinen: It’s been an interesting year for sure. One of the biggest challenges is that the rapid advancement of AI technology means that we can’t plan far ahead. Almost all the goals we had for the full year 2023, we’ve already achieved. We constantly get things done faster than we anticipated, and often the reason is that a new tool becomes available that speeds up development.
It’s also been somewhat of a surprise how much coding there is to do when you’re building a platform that allows making games without coding. At least in the near-term future, the amount of coding is not decreasing, quite the opposite. AI tools make it easier to write larger amounts of code, but this results in building larger and more complicated projects, which in turn need more code.
Yanes: On the flip side, what are some impressive elements of Bitmagic that you are particularly proud of?
Penttinen: I’m really excited about the fact that the system works so well. There’s nothing like this out there, so this was something we didn’t know until we built it. Due to the way generative AI systems work, it really seems to be working as a catalyst to human imagination, not the other way around. Every time I play with Bitmagic, I come up with new ways to do things, and I feel we’re just scratching the surface on what will be possible.
I’m really most proud of the Bitmagic team – we’re a team of brave doers, and you can see it in the rate of progress. For example, our Tech Lead never says, “I don’t know how to do this.” Instead, he says “I’ll need to learn how to do this”, and then he goes to learn it. When you have a team of talented people who believe nothing is impossible, there’s really no limit on what you can achieve.
Yanes: When people finally get around to using and enjoying Bitmagic’s offerings, what do you hope they take away from the experience?
Penttinen: My hope is that it enables people to do things they’ve always wanted to do, but never had either the skills or the time. I know how hard it is to make games. Even if you know how to code, and you know how to use all the tools, it will still take months or years to build even a simple game. For this reason, most cool game ideas in the world have never been implemented. That’s what we hope to change. I am hoping to see something completely unexpected, some true innovation in game design.
Yanes: Finally, what else are you working on that people can look forward to?
Penttinen: Bitmagic is my full focus right now, and far into the future. We’ll open alpha testing in November and then frequent updates all through the next year. I think we’ll see our Midjourney moments along the way. Remember they used to have a problem where people had too many fingers? It was just a few months ago, but it feels like a long past. So, over the next year you should start seeing some really cool games coming out that were made by users of Bitmagic, and those games will keep gradually getting better and better. Our goal is that in the not-too-distant future, games that the “regular” people can create in Bitmagic will look as good and play as well as any AAA-game out there.
However, there is a group of talented developers in Brazil called Mathilda Studios, who are working on taking some of my game designs to the market. Raivo, the robot fighting game I mentioned earlier, should be coming out in 2024. And Guntech 2 will see reincarnation in the web3 world, including new mobile builds, in the near future. I’m really excited that while I need to dedicate my full focus on Bitmagic, I’ve been able to keep these projects alive by finding new homes for them, and then seeing the results of the hard work by these talented developers come to fruition.