I love tinkering with new technologies and launching ideas.
Personal projects, advocacy experiments, and a 32-year-old role-playing community.
I've always been drawn to new tools and technologies — especially the ones that feel just a little bit ahead of their time, and I've never quite been able to resist the urge to try things for myself. That impulse runs deep: my dad was an entrepreneur, and I grew up watching him chase ideas and take risks! I think that shaped me more than I realized.
This page is where I keep track of the projects I've started outside my core consulting work — experiments, tools, and resources sparked by a mix of curiosity, admiration, and a deep love for what's possible when people build with purpose.
Some projects are polished, some are scrappy, but all come from that same drive to explore what's next and put it into practice.
Claude Code guardrails
Everyone I talk to at nonprofits is either using an AI coding assistant already or about to be. Almost none of them are security engineers, and the advice out there assumes they are. So I wrote down the handful of defaults I'd want a friend to have in place before they let an agent loose on their laptop.
The whole thing rests on one idea: everything the AI touches should be backed up somewhere it can't delete. You can't guarantee an agent never makes a mistake or gets talked into one — but you can make the worst day survivable. The repo has a drop-in CLAUDE.md, example permission and sandbox settings, an "out of its reach" checklist, and the research behind each choice so you can check my reasoning. MIT licensed, and honest about what it doesn't protect you from.
Freehold
I've been working with progressive nonprofits long enough to watch them get cornered by platforms they can't replace. Freehold is what happens when you sit with that frustration long enough — a framework for an internet where you create, publish, and actually own everything you share online. Your content lives on your own domain, hosted by whoever you trust, syndicated out to whatever networks you choose. No platform owning your audience. No algorithm rewriting what your followers see.
Not a company. Not a product. A proposal — and an invitation. The whitepaper is at freehold.is, published under CC BY 4.0. Quote it, fork it, build a piece of it.
Awesome ActionKit
ActionKit powers email and advocacy for most large progressive organizations, but the best tooling around it has always been scattered — buried in Slack threads, ClientCon sessions, and the occasional vendor blog post. New admins spend months rediscovering things other people figured out years ago.
Awesome ActionKit is the hub I wish had existed when I started. A curated, CC0 list of snippets, recipes, integrations, consultants, and guides — the things that actually work, collected in one place that's easy to find and easy to contribute back to. CampaignHelp seeded it with a handful of repos (redirect blocks, mailing blocks, a health check). Everyone else is invited.
Progressives for AI
Most of the progressive conversation about AI is a fight between people who think it's going to destroy everything and people who aren't paying attention. I don't think either of those is the useful place to stand. The technology is here, our side is going to be using it either way, and I'd rather we were good at it than caught flat-footed.
Progressives for AI is a newsletter and a discussion group for people who want to think seriously about what this stuff can do for organizing, advocacy, and the work of running an organization — without the doomerism and without the hype. If that's the conversation you've been looking for, come have it with us.
Progressive email suppression list
Every big progressive organization has quietly built the same file over the years: the list of email domains that never deliver. Typos of the major providers, throwaway addresses, spam traps that can get your whole sending domain blocked. The large orgs have one. The small orgs, who need it most, usually don't — and there's no reason for all of us to keep rebuilding it separately.
So I combined them. The internal lists people have been passing around for years, plus the big community-maintained sets, deduped into one clean file of roughly 66,000 domains that rebuilds itself nightly. There are ten-minute import guides for ActionKit, Action Network, and Listmonk. Released under CC0 — no attribution required, just take it.
It's purely defensive. The list exists so you can not mail these addresses. Please don't use it for anything else.
Ready for AOC
A grassroots supporter hub for people who want to see Alexandria Ocasio-Cortez run for president. The petition doubles as the email signup, and the whole thing is deliberately money-free — no donations, no merchandise, nothing that would drag it into political committee territory with the FEC.
An independent effort. Not affiliated with AOC, and not affiliated with any campaign committee.
HighSpeedRail.TV
American high-speed rail coverage is scattered across a few hundred YouTube channels, and most of the good stuff is made by people outside the US who are baffled that we don't have any. I wanted one place to watch it.
HighSpeedRail.TV is a curated video site for rail advocacy — new videos surfaced and reviewed weekly, with AI doing the first pass on what's worth including so the queue stays manageable. It's a small site about a big idea I'd like more Americans to see up close.
User Manuals
One of the greatest workplace tools I was introduced to at CREDO Action was the "user manual," which helps guide employees through an exercise in explaining to their team members what they need to be a whole person in the workplace.
So many of us struggle to manage needs and emotions that — if explained — might actually help our coworkers better coordinate so everyone can feel accepted and thrive in the workplace.
User manuals will not only help each person explain what they need, but also help everyone on the team understand each other better.
The Transparent Hiring Pledge
Hiring processes are often opaque and frustrating — especially for job seekers in the progressive space who expect more from movement employers. After years of seeing friends and colleagues burned by "ghost jobs," misleading postings, and silence after interviews, I created the Transparent Hiring Pledge to push for change. It's a simple commitment: organizations pledge to adopt a handful of clear, respectful practices — like posting salary ranges, closing job listings when filled, and communicating outcomes to all applicants.
I believe transparency in hiring is more than a best practice — it's a signal of respect. When organizations are honest and responsive in how they recruit, they help build trust and equity into the very first steps of a working relationship.
Four Days with AI
After reading about how a lot of companies are hoping to make workers more efficient with Artificial Intelligence, it dawned on me that we're already headed in the wrong direction with this technology that fascinates and delights me on a daily basis: Instead of taking the gains from Artificial Intelligence tools to finally — FINALLY — start cutting back on the work week and give back some rest and rejuvenation to workers, many startups and techbros are looking for ways to expand OKRs and make workers produce even more!
The four-day work week has been a growing movement over the past few years, and I feel like it's a good time to try and combine both the four-day work week movement with the gains of Artificial Intelligence, as they start to ramp up, to wrestle back the work week from employers.
We need a movement that will remind everyone in society that we aren't made to work ever-increasing amounts under ever straining pace indefinitely.
It's time to embrace Artificial Intelligence and use it to help us take more time off.
Tracking progress on "Machines of Loving Grace" predictions
The future of AI is unfolding fast — and it's easy to lose track of what's hype, what's real, and what's still just theory. When Anthropic CEO Dario Amodei published "Machines of Loving Grace," it stood out as one of the most thoughtful, concrete predictions about where AI is headed in the coming years. But thoughtful predictions still need accountability. That's why I created the AI Progress Tracker — to help document, fact-check, and annotate the progress we're actually making toward the world Amodei describes.
This isn't a fan site or a hit piece. It's a living record. By mapping each of Amodei's milestones and claims to real-world developments, the AI Progress Tracker offers a way to stay grounded in reality — whether you're excited about what's coming or deeply skeptical.
StarBase 118 PBEM RPG
That's a lot of letters, right? PBEM stands for "Play By Email," and RPG stands for "Role-Playing Game."
In 1994, on AOL, I co-founded a Star Trek role-playing game soon after learning about them in chat rooms. I was a Star Trek fanatic, as a kid, and I loved the utopian vision of the future mixed with the sense of cool, calm competency that everyone on the show gave off amidst the vast interplay of species in the galaxy.
That game grew into a community that I still have a hard time believing exists to this day, and which I've led for over 30 years.
Over the years I've watched people go through many phases of life, meet fellow members and get married, have kids, and even divorce. It's been the privilege of a lifetime to lead and watch all of this happen.
Beyond all that, the game provided me with a testing ground for leadership, community, and technology like perhaps nothing else could have.