Yarn Spinner team refuses to integrate AI based on labor rights and development ethics. They criticize current technology for becoming a tool for corporate layoffs, deviating from the original intention of helping creators.
(Background summary: 5 lines of code driving Silicon Valley crazy! Aussie sheep herder breaks open AI programming singularity)
(Additional background: OpenAI reportedly spent 3 billion USD to acquire “automatic programming magic Windsurf,” aiming for the Vibe Coding business opportunity)

Table of Contents

• Past

• Future

• Frequently Asked Questions (FAQ)

As the most influential open-source narrative scripting tool in the indie game community, Yarn Spinner’s stance is crucial for many developers. In this official statement, the team explains sincerely and firmly why they choose to refuse adding generative AI to their tool.

They not only approach from a technical perspective but also deeply reflect on labor rights, development ethics, and “the essence of game development.” This may offer some reflection amid today’s industry obsession with technology trends.

Below is the full Chinese content translated by Dongqu:

We are often asked questions about AI. Will we add it to Yarn Spinner? Do we use it ourselves? What are our views on it? These are very fair questions. It’s time to document all of them.

Yarn Spinner does not use the current technology called “AI.” Our product does not have generative AI features, we do not use code generation tools to build it, and we do not accept contributions that contain generated content as we know it. Let’s talk about why.

TL;DR: Tools created by AI companies hurt people, and we do not want to support this behavior.

Past

First, let’s talk about some history. Our background includes quite a bit of work related to AI and machine learning (we shouldn’t conflate these two terms, but since everyone does, we will alternate usage).

We have held related lectures for game developers and non-programmers; written small machine learning bots for game scripting; conducted research and academic work. We even authored a book on using machine learning in games, mainly about procedural animation. It’s a series of very interesting and worth-exploring techniques, and we have indeed explored them.

When we started university, neural networks and deep learning (the core technology behind most AI products today) were still slow and difficult to operate. By the time we finished our PhDs, the situation had changed. Tools like TensorFlow made everything simple and fun, and the accessibility of GPUs allowed those without big tech budgets to train and infer. For a long time, we were genuinely excited about this potential.

Then, things started to change.

It’s hard to pinpoint exactly when. Maybe it’s always been like this, just we didn’t notice. But by the end of 2020, it became very clear: the AI we liked was not the kind that tech companies were interested in.

They increasingly favored generative images, chatbots that help you write things, and “summaries” of art rather than enabling people to “experience” art. Efforts to address known issues (such as deepening cultural biases, difficulty in achieving certainty or interpretability) were devalued and ignored. Researchers and developers raising concerns were dismissed.

Since then, the situation has only worsened.

If you look at what AI companies are marketing now, it’s not what we want. When you peel back all their words, the tools they make are essentially either for laying off workers or demanding more output without hiring more staff. That’s the “problem” AI companies want to solve.

Any other achievements they realize are just accidental surprises on the path to “laying off your friends and colleagues as much as possible.”

In this era of especially difficult re-employment, unemployment, and even threats to life, AI has become a “tool for layoffs.” We do not want to be part of it. Until the problem is addressed, we will not use AI in our work nor integrate it into Yarn Spinner for others to use.

We do not want to support companies that make these tools, nor normalize their behavior. So, we do not use them.

Future

We occasionally see comments that sound like “fait accompli”: “If you don’t adopt AI, you will be left behind,” or its relatives: “Everyone is using it”… We disagree.

Regardless of our views on AI, this is not the proper way to develop. It’s “tool-driven development.” The goal should never be “we use this tool,” but rather “how can we help you make better games?”

Great games are born from people’s passion for a certain idea and pushing it into reality. This usually means “reducing” rather than “adding.” Changing ideas, maintaining health for yourself and colleagues, being willing to adapt and accept feedback. Good tools should also do this.

We keep asking ourselves: “How can this help make better games?” and follow the answers. The process of exploration is important, and most of the time we find that many ideas cannot withstand even a little scrutiny. We prefer to have a few refined features that solve real problems rather than a bunch of junk for marketing copy.

We are proud of Yarn Spinner. We don’t think its widespread use in so many games is just a coincidence. Our development process is effective, and we keep adding new features. If some features don’t meet developers’ needs, we modify or remove them.

We constantly discuss potential ideas and methods with internal teams, other game developers, and even non-developers. We will continue to ask: “How can this help make better games?” and release features that pass the test.

Who knows, maybe the world will change, and we can revisit machine learning someday.

Frequently Asked Questions (FAQ)

Why do you only care about people being laid off? I heard AI is also bad in “other aspects”! AI (especially the companies that make it) has many issues. Some are potential, even hypothetical concerns; some are very real and happening right before our eyes. Some problems are much more serious than “being laid off.” Between writing this blog and publishing it, even worse issues have emerged. Even if the labor rights issues around AI suddenly disappear, many problems remain before we can safely use it. But focusing on one argument at a time is more effective. Labor rights issues are fixable and should be resisted. Once that’s resolved, we look at the next problem.

Why don’t you “properly” develop machine learning so that no one gets hurt? Considering our background and experience, we might develop a set of AI tools we believe are beneficial, ethical, and do not fund the companies we oppose. But there are two problems: first, these things take a lot of time to produce, and as we said, most ideas don’t pass initial exploration. Balancing creative exploration while building new models to test is very difficult. Second, although we can make our own tools, most people cannot. If they see us using a certain technology and want to try it, they will ultimately support those companies we oppose. We don’t want this behavior to become normal, so we must lead by not using them.

My boss requires me to use AI at work, am I part of the problem? Finding and keeping a job is essential for survival, and recent pressures make this even harder. If you can oppose it, try to do so. But no one will blame you for wanting to keep your job.

Will you ban people who use AI from using Yarn Spinner? No. Although we hope you don’t use it, we understand this is “our” bottom line, not yours. We will still advocate against these tools and worry about the harm they cause. You need to realize that if you use them, you are financially and socially supporting those despicable companies that do evil. They will use your support to push their agenda. If these tools help you, we sincerely are glad, but please stop using them.

I really like using AI, and my company hasn’t laid anyone off? Such comments often come from programmers. Unfortunately, because the word “AI” is now so confusing, the same concerns still apply. Your adoption helps promote the companies that make these tools. Others see you using it, and may force other studio members to use it or promote it elsewhere. From our observations, layoffs and overwork often follow. If it’s not happening to you and your colleagues, that’s good, but you are still helping this happen elsewhere. And as we said, even if labor issues are solved tomorrow, many other problems remain. Besides layoffs, there are many other concerns.

Are you just a bunch of AI-hating fanatics or Luddites? No. We are simply angry at the people who make these things. AI and machine learning have huge potential, but they are being squandered on “making already very vile rich people richer and more vile.” We still care about technological development because we hope to explore it again someday. But for now, the people pushing these tools are not those we want to give money or support to.