Initial Draft for AI Meta Policy - #129
Conversation
|
|
||
| --- | ||
|
|
||
| ## Versioning |
There was a problem hiding this comment.
Took some liberties here as well, if everyone's behind the idea of versioning I could add a decision log file to this folder
There was a problem hiding this comment.
@alea12 any thoughts on this? I could see versioning/maintenance being an extra hassle but I could go either way
There was a problem hiding this comment.
I like the idea of documenting decisions/changes/updates. To make this easy on whoever is writing the docs and proposing changes, I'd recommend consolidating into one log if possible, and potentially not requiring Major/Minor/Patch or version numbers. Personally, I think I'd struggle to decide if a proposed change was major or minor or patch, when I'm not sure there's much value from that sort of decision fatigue. Better to just document the change and keep moving.
Of course, deciding between Major/Minor/Patch might make it easier to decide who should be notified about a change. Like maybe all Major changes require an email to go out to all engineers? Or we could make it easy and just require all changes be distributed via email.
|
|
||
| After assessing the current policy against their engineers' experiences and against any external or vendor changes (e.g. licensing, pricing, data privacy updates), the pods will propose changes as needed, along with the associated rationale. | ||
|
|
||
| Final approval of these recommendations will come from either higher leadership, e.g. Garvita Kapur. |
There was a problem hiding this comment.
| Final approval of these recommendations will come from either higher leadership, e.g. Garvita Kapur. | |
| Final approval of these recommendations will come from Digital Engineering Leadership, i.e. Garvita Kapur. |
I wasn't sure what was meant by "either" - is there someone else who could approve besides Garvita? Sorry if I am misunderstanding.
There was a problem hiding this comment.
Oops that's a typo on my end! Not sure who else would approve besides Garvita... maybe the engineering directors?
| | Quarter | Rotating Pod Pair | | ||
| | :------ | :---------------- | | ||
| | Q1 | Research Catalog + Data Platforms & Access | | ||
| | Q2 | UPE + ISW | |
|
|
||
| ## Overview | ||
|
|
||
| This is a "meta" document — unlike the [AI working policy](./README.md) itself, this does **not** define how NYPL Digital should use AI in day-to-day engineering work. Instead, it answers the following questions: |
There was a problem hiding this comment.
The doc you're linking to seems to be called NYPL Engineering AI Guidelines. Maybe there was a conversation about this? But I think it would be helpful to use that name here for consistency and clarity.
|
|
||
| ## Rotating Stewardship | ||
|
|
||
| The AI working policy and its associated governance will be reviewed on a **quarterly basis**. Two engineering pods will be paired up to review and recommend changes to the policy, based on their teams' actual experiences with AI tooling. Not only does this encourage cross-pod collaboration, but it also gives technical leadership a chance to gather how and why tools like Gemini and CoPilot are used by engineers. |
There was a problem hiding this comment.
I really like this
5 participants
No description provided.