Hi everyone,

It’s been a fun couple of years. When I started this newsletter, I had the goal of tracking AI use in the public sector. I also created a reading list for others in my position (tech in city government).

It has been nearly 2 years since I left the public sector. I continue to engage with it, but not in the same ways.

For that reason, I will be retiring this newsletter, but it’s not goodbye forever.

1. I started an AI research and strategy lab: Future Ethics

2. The video from this recent webinar I participated in reflects my recent thoughts and work in ‘AI’ and ‘government’ (blog post summarizes some thoughts)

3. Follow the Partnership for Public Service for more useful content and programming in this space

4. As a fun side project, I built Link In Comments to help me stay on top of the massive amounts of reading in this area. You can try it too

Some parting thoughts:

I believe that the main difference between when I started the newsletter, and today, is that the conversation has shifted. It’s no longer, should I use AI? Or, what AI should I use?

I have conflicting thoughts on this. It continues to be my position that AI is but one tool out of many in ‘digital government’. Does it have the potential to ‘accelerate’ and ‘help’…? Yes. Does it also have the potential to do the opposite of that? Also yes.

I know many smart, thoughtful people who are in all levels of government around the world thinking about precisely this.

My earlier fears of ‘everyone is just going to use Microsoft’s AI for everything’ is partly right, but also it’s more nuanced than that.

There are now roles or teams set up to think about, design, potentially implement AI solutions in many levels of government. From my conversations with local and state officials, I feel thankful to know as a resident that there are people in public service who care deeply about the impact of AI on people and society.

However, I’m tentatively curious about what the role of Chief AI Officer, and how teams set up around this structure, will evolve Will these be governance-only role? Where do existing digital services team sit in this? Ultimately, who is accountable when a local or state government deploys an AI system that leads to discrimination in hiring, or prevents marginalized people from receiving benefits?

All of those are interesting questions, and I will continue to keep an eye on those through my role lecturing or giving talks on all things AI and government (if you’re like me to speak at something, contact me here).

But that’s not my full time focus anymore.

What I’m up to next

Since I left local government, I’ve been exploring the wonderfully chaotic and unstructured space of ‘a bit of academic / AI / ML research’ and ‘a bit of giving talks about bias’ and ‘actually running large, participatory evals’ (like the ones I worked on for the DoD and the Singapore government).

I’m really leaning into ‘combining my product development skills’ with ‘doing a lot of research’ and ‘trying to build products that can expand access and agency for marginalized people on AI and AI-related topics’.

My most recent project in this space was a collaborative evaluation application that can help community facilitators run a workshop where community members can evaluate how image-based generators may or may not cause harm or create bias and erasure to their specific communities.

I’m also working on ‘CityLangScoreCard’, a longer term piece of work that was informed by my time working on SF.gov and its translation systems. Many of the decisions that I helped put in place as former director of product, informed by Covid-19 era need to ‘get public health information right’, has also led to some academic research and open source tooling that I’m working on that will hopefully help others like me.

Lastly (this is a lot), I’m spending the next phase of my career in this area thinking about and working on issues around how AI impacts marginalized people, speakers of other languages, and people in the Global South.

If you’re interested in staying in touch, feel free to add me on LinkedIn. I am available for consulting in these areas.

If I write or set up another newsletter in the future, your participation will be strictly opt-in. Feel free to reply to this email and let me know you are interested, and I’ll keep you updated.