I'm taking the decision to never use generative/transformative AI for interactions and communications with other humans. That includes email, chat, online comments, any content on this site. But also commit messages, bug tickets, pull request descriptions, decision docs created under my identity and where another human would be expecting to be interacting with me my words.

I do not want people to ever have to guess if prose written by myself has been edited or generated using AI tooling. The content will be 100% coming from the human Sam with all my human flaws and human oddities. If you see em-dashes they are mine — my keyboard layout has them in direct access. Any strange, uncanny phrasing is coming from me.

Human trust is precious and fragile, it is something I want to preserve in my life and introducing LLMs in human relationships is too damaging.

Like pretty much all other content in this website I'm writing that down for myself. I decided to publish it as a way to keep myself accountable, and hopefully inspire others to reflect and discuss that topic.

I will also eventually review my past content and add AI disclaimers where needed.