Our Principles
We are interested in what other people have to say. It's an interest driven by the fact that you are, like us, human. Understanding your view of the world helps shape our own. Who wrote something and why they wrote something matters.
Even if human written text is indistinguishable from computer generated text, there is intrinsic value in listening to what people have to say. The expression of a point of view is imbued with value by the person who holds that view.
The sheer scale with which AI generated content is being created puts us on a path to an internet where actual human ideas are hard to find in a sea of machine generated content.
We created OurWork as a place where writers and readers can come together in the knowledge that what is posted here, however bad it may be, is written by people.
We also offer a sharable certificate that enables writers of all kinds to demonstrate how they made it. Read more about how we do this.
Our Work is based on these principles:
- Human thought matters
- Writing is thinking
- The real threat is scale
Human thought matters
Or more to the point, the human origin of ideas matters. Whether it comes to statements of fact, opinions, analysis, or anything else. The text is useless without someone to take responsibility for it; without someone willing to stand by that text to say that they took time to think about it and believe it's worth a read.
This extends to writing: if you've taken the time to write an idea down, it means there is at lease one human being (you) that believes this to be worth the effort.
Writing is thinking
Human though is not linear. We iterate, correct mistakes, and change our minds. This is often reflected in the history of a document. Our writing reflects our thinking.
At least for now, LLMs do not work that way. This means we can still take proof of the process of writing as proof of thinking. And by analyzing the process of writing we can determine the human origin of text.
The real threat is scale
We see a great risk that the work and the views of humans could be drowned out by automatically generated content simply as a result of the amount of AI content.
Generating content at scale also makes it significantly easier to manipulate public opinion and create a false sense that certain views are held by many.
From this perspective, the real goal is not to create a perfect AI detector. The goal is to eliminate the scale factor of AI.
There are certainly ways to try to fake the process of human writing and it takes continuous effort to combat this. If faking human writing is more effort than just writing something yourself, it mitigates the worst threats of AI.
Regarding the use of AI in general
We do not need to take a broad position on the use of AI to feel strongly that there is value in human work.
Some have argued that all use of LLMs is unethical because of how they were developed and their environmental impact. The reality is also that many people would risk their jobs if they do not follow employer mandates to use AI. Not everyone can afford to take a principled stand and people should do what makes sense in their personal circumstances.
Our position at present is that if AI was used in the creation of text that is shared with other, that text should be labeled. This enables everyone to make their own choices.