LLM disclaimers and philosophy
On this page, I will articulate how LLMs are used in my projects and on this website with the aim of being fully transparent. I believe that in the era of software 2.0 these kinds of disclaimers should be a standard part of digital output. Below text is my effort to put my money where my mouth is and provide that transparency.
As LLM tool usage increases across many fields, I also feel the need to contribute to the conversation of how they should be used. I will try to formulate a kind of AI-etiquette and advocate for it, while doing my best to follow that etiquette. Please feel free to point out my shortcomings, when you inevitably encounter them. To read more about my views on AI Ethics, check section Blog posts on AI Ethics to view my blog posts about it.
LLM disclaimer – transparency
Presenting work as yours that isn’t is a lie. Therefore, on this website I will do everything in my power to be transparent about where and how LLMs were used.
This website can be thought of as consisting of elements. There are videos, paragraphs of text, small web-app components and plots and graphs. Whenever something is done with assistance from an LLM, it will be clearly marked. The main goal is that everyone can tell immediately what was generated and what was not. Below is a list of all the ways LLM is used and signified on the site.
- Generated texts are encapsulated within blue boxes, where the box header clearly states the fact that they were generated. Generated texts are created with a prompt and even if I have edited them, the defining feature is that the original text draft is not my own.
- Generated plots and videos will include a clear tag signifying prompt driven work. For example, I’m learning to create animations for better communication of ideas, but those videos will be almost certainly workshopped with an LLM using prompts and skill-files. In the blog post How should LLMs be used?, I will discuss the philosophy of this in more detail.
- As I am not a full-stack-developer, and likely never will be, the website itself and all web-applications on this site are workshopped with an LLM. Whenever a project uses web-app elements, the text describing the project will explain how and why the LLM was used.
- In addition, each project and blog post will include at the end a summary disclaimer. If you are in doubt, please scroll to the bottom of the page to find out exactly how LLMs were used.
There are two reasons for transparency efforts. The first is the obvious fact that misleading people about your work constitutes a lie and does not promote fruitful and honest collaboration. The second is that being transparent about LLM usage effectively communicates how the person utilizes these tools and thus serves educational purposes. In this era of increasingly better software tools, gatekeeping and hiding their effective usage should not be the standard practice.
Blog posts on AI Ethics
- How should LLMs be used?
- Philosophy as conceptual engineering and responsibility
- Bilbo, ChatGPT and the elephant
Claude Opus 5.5 grammar fix suggestions accepted.