WHO WE ARE
Thunderbloods is just two friends, Jacob Moretti and Aaron Marquez.
Our goal is to make a game that is uniquely us and to bring some of our friends and family along for the ride.
OUR FIRST GAME
CENCON: AWAKENING
A story-first 2D pixel-art RPG. In active development.
A world ravaged by climate change. Your father vanished in CenCon Valley when you were a child. Eighteen years later, you set out to find answers.
OUR GEN AI STATEMENT
Generative AI, both a curse and a blessing. We recognize multiple truths, including its capability as a tool, its reckless consumption of human creation, and its puppet masters' dubious intentions. How do we balance leveraging a tool while also understanding its potential for harm? Looking at history might give us some insight.
Learned Language Models (LLM) in the form we understand today have been around since 2017. However, they are based off of theories developed in the 1950s! It was simply that we did not have the compute power nor the vast amount of digital data to practically apply those theories, until today. It is true, the digital data being fed stands on the shoulders of the internet. But we should recognize that many of the technologies that we take for granted today, like the internet as we know it (TCP/IP) and even GPS, were originally developed for the US Department of Defense. Always on the pretext of war.
As language models accelerated their capability with the mass consumption of human created data, whether legal or not, moral or not, our technology is often born alongside darkness. The ubiquitous mobile phone in our pocket, depends upon precious metals mined through questionable practices and assembled in dystopian factories. The chips that power those devices are manufactured at fabrication plants where metallurgy recipes are closely guarded and labor is exploited. The very food we eat is farmed by exploited labor as well. Our participation in Western society seems to deem all of these uncomfortable truths as necessary. We hear the familiar argument that these create opportunities to lift others out of poverty where an industry would not exist otherwise. But does it excuse our willingful use of Gen AI? Maybe not, especially given the excessive forced rollout of this technology through the reckless mass construction of data centers. But what about the slide rules? Those poor slide rules, replaced by pocket calculators. Will humans share the same fate of being replaced by Gen AI? No, not all of us, at least. But we have to be dilligent, accept it as a tool. However, don't conflate capability with competence. Art is still uniquely human no matter what capabilities it can grant you. With competence comes understanding and skill, time devoted to improving ones competency so that they can be capable. This is where a human will always win out and it is also why technical competency can't be lost either.
Maybe we're just making excuses so call us a hypocrite if you like. After all, we are American citizens who oppose many of the country's policies, especially those of the Trump administration and any corporate-centric politics. Yet we still live and work here, exploiting this opportunity of being born here. We see our use of Gen AI the same way: something we have reservations about but work within. We can recognize the benefits while also being honest about our criticisms. At the end of the day, we want to make things better in this country for the working person. That's why we want to create games and tell stories through it. We see it as an opportunity to both express ourselves and raise awareness about certain societal themes, balancing a dark and serious tone with silly moments that give some room to breathe and maybe laugh.
How Do We Use Gen AI?
Simply, sprite generation and coding.
For starters, we don't use it for any story development, dialogue, ideation on game mechanics, music, audio or sound effects. When it is used for sprite generation, typically a hand drawn reference image is supplied. Occassionally we reference pictures of local architecture. Often we leverage non-digital concept art from friends (some who are very talented traditional artists) or ourselves (crudely drawn) to generate a pixel sprite version of the art. Then we take it into Aseprite and clean it up by hand.
For coding, it is used extensively. I (Jacob Moretti), have been a professional Software Engineer for over a decade. Everything built is architected in thought out systems, often treating the AI agent just as I would a colleague if I were leading a project. With requirements and specifications, then making decisions on trade-offs as they come up. It certainly helps to be able to read the code and produce a mental model of how the systems interact. This isn't vibe coding. I'm building frameworks that I understand for the long haul.