Metric: "New Newcomer"

[emmairwin]

Based on this article, and Justin Wheeler's experiences of how new contributions and assessment of contribution is changing, lets consider what a metrick might look like:

I think like with use, and feedback - creating user stories is important to identifying the metric, and data that can help inform a decision.

I think the question for this metric is : How do we encourage newcomers in an era where 'first contributions' are much bigger, more complex and not-always human' (but please propose better ones!(

https://blog.stdlib.io/ai-and-the-invisible-newcomer-in-open-source/

[deepanshutiwari27]

Instead of measuring the extent to which an individual’s first contribution is made, what we require is a newcomer progression index, which involves understanding how people learn to move from first meeting through to first contribution and thereafter becoming regular contributors. like tracking
Discovery | First issue comment
Engagement | Discussion participation
Contribution | PR submitted
Integration | PR merged
Persistence | Returns within 90 days
Trust | Given review responsibilities
Stewardship | Mentors another contributor

[emmairwin]

Trust is important in future, how do we know what work comes from the person and which from an LLM
how does a maintainer process large complex contributions from new comers

[justwheel]

Discussed in 2026-06-24 AI Alignment WG meeting.

---

We briefly touched back on this toward the end of today's meeting, and @emmairwin tagged me to share some of why this article was so interesting to me! The article in question spoke toward how the newcomer experience is changing in new communities, and how new users are often engaging with LLMs first instead of going into a community, asking questions and getting help, and a few of those folks being converted into helping others, and even going deeper into maintainer responsibilities.

The article in question describes how it is becoming harder to recognize new contributors engaging in a community for the first time. It describes how the role of explicit invitations as more important than ever, to create specific in-roads for new contributors to go from passive user to engaged contributor.

Another angle from the article, which may or may not fit in, is this idea of proof-of-work. This part focuses more on trust and how a new contributor needs to earn/gain trust when using LLMs that they have actually done independent, thoughtful, critical thinking behind the change, and not just the first output that the LLM spits out. Having proof-of-work protections in place help maintainers have better balance for maintainers who need to address the rise of sloppy contributions driven by LLMs.

These are thoughts in digestion, but curious for what others think here too.