in

Can One Person Make 9 Figure Games? An Intro to The AI-Generated Creativity World

What allows a single person to create a game on the same level as one created by hundreds of experts with a budget of over $100 million? It’s possible because of the advances in game development technology and tools available today.

A decade ago, it would have been impossible for an independent developer to create a game of this size and scope. But now, with the right tools and technology, it’s not only possible – we’ll probably see many more games like Elden Ring in the future powered by one indie dev.

The left image is from Disco Diffusion 5

Why are AAA games so expensive to produce? The answer usually has to do with content: a lot of work goes into designing, building, and perfecting the game. However, game-makers may be more like composers than brick-layers in the future. Most of the above will be made possible by an emerging field called Computational Creativity.

With this technology, game-makers can create rich and complex content more efficiently, making AAA games more affordable for everyone – from indie developers to the AAA game studio. Imagine being a storyteller capable of creating a gigantic 8-9 figure game (encompassing 1000+ hours) for less than six or even five figures?

According to OpenMind BBVA,

Computational creativity is the study of building software that exhibits behavior that would be deemed creative in humans. Such creative software can be used for autonomous creative tasks, such as inventing mathematical theories, writing poems, painting pictures, and composing music. However, computational creativity studies also enable us to understand human creativity and to produce programs for creative people to use, where the software acts as a creative collaborator rather than a mere tool. Historically, it has been difficult for society to come to terms with machines that purport to be intelligent and even more difficult to admit that they might be creative.

Synthetic media is going to be future topic discussed.

Elden Ring Concept Art and Design COMPLETED in 2 minutes?

Composable frameworks and artificial intelligence are increasingly crucial in enhancing human creativity. For example, AI-powered tools like the Disco Diffusion software can generate unique artwork with little human input.

This tool is just one example of how AI is changing the landscape of creativity and enhancing human potential. This online tool runs within Google Colab, which allows for executing cloud computing-powered GPU engines running AI algos. Here is a small example showing 25 monsters I created using this tool (over 6 hours of computing time required):

Created with Disco Diffusion

The monsters in Elden Ring are far more exciting and impressive than anything a computer could create with minimal input from me. They’re so good that it’s hard to imagine how they could get any better. However, given that technology is continuously improving, it’s not hard to imagine that the monsters in Elden Ring will only get more impressive and realistic over time. For now, add a bit of artwork design expertise to the above, and you will have elementary design built to be READY to GO.

Creating ART from Ideaz

This is another Disco Diffusion creation, experimenting with the theme of sunflowers and destroyed tanks. Again, we see the ability of the AI model to not only reproduce the artistic style of something like Elden Ring but also to fuse ideas which never been combined before. This results in a truly unique and exciting composition.

Ravaged war machines being reclaimed by nature.

Some people see AI as a threat to human jobs and creativity. They worry that machines will eventually take over many aspects of our lives. While I understand these concerns, I believe there is still a vital role for humans in the AI world. We are the ones who come up with the ideas that the machines then execute. Without us, the machines would be nothing more than expensive tools. Human minds don’t paint with paintbrushes, only through ideaz.

Awesome chat on generative art

As computationally-generated paintings become more commonplace, the best artisanal human-crafted pictures may become scarce and, therefore, more valuable. However, “painting with ideas” exponentially increases the rate of human output – shifting work towards ideation, iteration, and composition. This could lead to a future where machines create the best paintings, and human-created images are only valuable due to their rarity.

Art + Design Evolves Into 3D Modeling

You’ll need to generate 3D models from digital paintings to create multiple immersive worlds. A 3D model defines the object’s shape and size in virtual space. This paper demonstrates how to learn the properties of 3D models and create new versions based on those properties.

This new learning model teaches computers how to generate a 3D object from an image. This will help speed up the process of building game worlds, allowing for novel 3D models to be created quickly and easily from photos.

Fascinating video showing 3d models created utilizing images

We are seeing a trend of older technologies, often seen only in research papers and Python scripts, being productized. This is making them more accessible and user-friendly.

AI-Powered Music Videos and Concerts?

Yannic Kilcher used two different learning models, CLIP and BigGAN, to generate a music video along with some lyrics he had written. The lyrics and song components were primarily meaningless, but they contained many words that the image-generation models found interesting to display.

Storytelling with AI – What about Fiction books generated based upon personalized user interest?

A language model like GPT-3 can interpret and generate text that sounds very human. For example, here is a story developed by AIDungeon:

AIDungeon

The Google Pathways model is even more advanced than GPT-3, with over 540 billion parameters. It can follow a chain of reasoning (cause and effect), give explanations to jokes, performs pattern recognition, simplify Q&A/AMA sessions focused on scientific knowledge, and summarize overly complicated texts. As more parameters are added into the mix for DL models such as this, they will demonstrate an increasingly deep “understanding” of language.

This is an excellent tool for explaining jokes. It can take complex ideas and break them down into much simpler concepts. Software like this is also conducive to understanding complex topics.

Code Generation With Da Vinci Models and More

OpenAI Davinci can be used to generate code for a Wordle-type game. This was done by Andrew Mayne using a few lines of input. The game is quite simple and does not require much effort to play. All you need is a bit of creativity and imagination. Enjoy!

Here is a powerful tool anybody can use to engage in pair programming with a very advanced DL-powered engine by Microsoft (powered by GitHub code). You can write in natural language and have the machine generate the code for you (Make sure the code has built-in security QA for added power)

AI-generated Virtual Worlds

Promethean AI aspires to help you create entire worlds by comprehending stories and generating 3D models in response to language prompts. With this technology, your imagination will be the only limit to creating new virtual realities. Check this out:

A system like Promethean is powerful because it is composable; it uses AI to generate objects and settings, but its true strength lies in its ability to iterate, aggregate, and tweak ideas. This kind of scaffolding for creativity allows a single person to have a much greater output than they would otherwise achieve.

Minecraft, Roblox, and other similar frameworks have already demonstrated the potential of this approach. These are all beautiful examples of a dev center for the metaverse – MUCH more will be discussed in terms of NVIDIA’s vision for the Omniverse

The Future 🙂

As AI continues to develop and evolve, so will its contributions to computational creativity. With frameworks that facilitate the connection of different pieces of creative content, and AI powering a multitude of iterative content creation processes, we can expect an exponential increase in output from both individuals and small teams. Thanks to platforms like Minecraft and Roblox that enable composability, people can craft experiences never commercialized before. The next eon of creativity will be characterized by the ability to rapidly iterate ideas with AI assistance which will bring forth experiences we can’t even possibly begin to imagine.

Taryn Smith from the AI Generated Art Group on FB

What do you think?

781 Points
Upvote Downvote

Written by Miles Price

Testing new editing system