As I defined my new role, I carefully considered the title. The Fedora community sentiment toward LLM-gen-AI is deeply divided: some rally against it, while others push for rapid adoption without wider community consensus. I needed a title that signaled a neutral, balanced approach. My mandate at Red Hat is to support upstream projects in adopting AI, focused primarily on Fedora. Working in the “AI” space at Red Hat is fascinating because I am exposed to diverse ways that customers use and deploy innovative open source technology. Additionally, Red Hat provides real value by supporting customers in their “hybrid AI” journeys. Many of my Red Hat colleagues consistently push for more Free Software and open source answers for customers and enterprises building infrastructure to support AI inference, local models, and more. Red Hat has a responsibility to innovate when new technology opportunities emerge that its customers are acting upon. With this in mind, I am more convinced that LLM-gen-AI is something that open source maintainers and contributors can leverage for real workloads. There are opportunities to solve real problems and routine maintenance tasks for complex projects. LLM-gen-AI used right can support maintainers in automating boring, cyclical work so they can focus more on the exciting work of innovation and focused engineering efforts within their projects. Or even going outside and spending time offline.

However, I distinguish "AI alignment" from "AI adoption." "AI adoption" communicates a pre-defined, non-negotiable stance where the goal is simply to increase usage. "AI alignment," by contrast, communicates that LLM-gen-AI use exists on a spectrum.

My intent in Fedora is not to insist, but to negotiate and compromise. I want to align how our community uses these tools with our existing values, norms, and culture.

My approach to "AI alignment" is influenced by the CHAOSS Project. I co-chair the CHAOSS AI Alignment Working Group with Emma Irwin and Coraline Ada Ehmke. Recently, Emma, Adrian Edwards, and I presented at FOSSY 2026 on this topic in greater detail. This experience frames my definition of "alignment" as I move forward in my new role.

Traditionally, "AI alignment" is a term used by model builders to describe processes where communities have little influence. As LLM-gen-AI grows in widespread use, the power gap between those model builders and the communities they impact will widen. This creates an unsustainable dynamic in the safety and well-being of our communities with these new tools.

We need more than just "AI adoption". We need bidirectional feedback loops. Free Software communities deserve a seat at the table. This is not just for iterating on model development, but for defining how we integrate LLM-gen-AI tools sustainably and responsibly.

To achieve this, we (i.e., open source community citizens) need a stronger value proposition for engagement with model builders. Whether commercial or altruistic, we must persuade model builders that a community-driven approach is a critical advantage. If we refuse to engage, we risk Free Software values and culture being shut out of conversations entirely. So, I believe it is better to be an advisor than a bystander. Advising is its own form of open source contribution. Therefore, I lend my support toward the wider notion that "AI alignment" fosters two-way conversations that ensure model builders actually listen to the communities their work impacts.

September 2026 is my first full month in this role. While there is still much to define, a few priorities have emerged. Here is where I will be focusing my initial energy:

I am both excited and nervous to work with fellow Fedorans on these initiatives. I know there are strongly-held opinions on both sides of the LLM-gen-AI debate. ( An understatement! ) I accepted this role because I believe participation is more constructive than standing on the sidelines. My goal is to navigate this work in alignment with Fedora’s Four Foundations: Freedom, Friends, Features, First.

Since March 2026, I invested a lot of time into migrating my blog to a new publishing system and improving the website user interface. However, now that my site is fully functioning on a technical level, it is refreshing to begin writing content again here. There is more to come from me in this space. Expect new content about my work in Fedora and the LLM-gen-AI space, and other open source and personal items too.