Project summary

p(Doom)1 is a nightmarishly hard, weird side project where players attempt to battle an inexorable rise in, you guessed it, p(Doom). I took the advice 'write what you know' literally.

The game is about the challenge facing real AI safety researchers, grantmakers and philanthropists, wrapped up in a way that hopefully communicates the challenge while mitigating the actual despair. You can't win; you can only buy time. There's no victory condition. Your score is how long you hold P(Doom) back before a run ends.

I started designing it out of a confluence of my interests (game design), gaming background (Magic: the Gathering, PoE, Tarkov, Civ, X:COM, sim games), professional background (project management specialising in governance), and career (operations work in Global Catastrophic Risks).

The early friends-and-family alpha is out at pdoom1.com. It's getting patched daily currently, and will move to a weekly cycle as of Friday 31 July.

What are this project's goals? How will you achieve them?

- Get the game in front of more eyeballs by improving quality of UX, UI, and infrastructure.

- Turn that into community engagement at a higher level of background technical understanding -- people learning a bit more about AI safety as a side effect of an engagingly hard game, the same way Young_Me learned a bit about history by playing Civilization.

- Underneath it is, I suppose, my real theory of change: people voluntarily adopting good actions generates the space for better actions to come about. Small and middle-power players affect things through norm- and market-setting. This feels like a step in that direction.

- Secret goal: watch people argue about how p(Doom) is calculated.

How will this funding be used?

- Fund development so I can continue to justify spending some of my time on the game.

- Pay a developer to take over leading some elements of gameplay.

- Pay human artists to uplift and steer the artistic direction. I'm using purely AI generated assets right now and would like to be paying artists instead.

- Pay for assets and hosting.

- Support any community that emerges out of people enjoying the game.

- Maintain and update the open historical AI-safety dataset the game runs on. It lives in a standalone public repo, is broadly useful to researchers, journalists and developers, and is designed to be available for other people's games and tools to draw from -- with maintenance and update schedules built into the dev cycle.

Who is on your team? What's your track record on similar projects?

Me, presently. I designed a board game about Rugrats trivia when I was 8 -- my brother and I enjoyed playing it, though my parents sucked at answering the tricky questions.

I played, then judged, then did coverage for Magic: the Gathering, and during that time interviewed hundreds of players, many pro-level, who went on to lead and design games themselves. Many had opinions. We talked a lot about game design and balance, and getting a good understanding of what makes depth and complexity work across rulesets.

Professionally I do operations work in Global Catastrophic Risks with a governance-focused project-management background. I know the pains I am inflicting on players.

What are the most likely causes and outcomes if this project fails?

- People might take AI safety less seriously due to this game being low-status or failing in reputationally damaging ways.

- People might see this and be discouraged from making a better game.

- The game might contribute to ideas that are dangerous, disparaging or discouraging of good actions, like giving a false impression that some capabilities are maintainable or that we can swerve into safety effort later, rather than taking actions NOW.

-I've designed against that last one deliberately: the game is explicit that we're not saying or implying labs really do the truly weird stuff that takes place later in play, and the underlying dataset stays strictly factual and is not predictive. Hopefully this also lessens the likelihood of someone lawyering me.

How much money have you raised in the last 12 months, and from where?

None. I have put in a fair amount of my own time over the last year and a few grand of my own savings into it until now.