Mangrove: Matchmaking for AI Safety

Joining a peer group to upskill into AI safety makes people more likely to put in the effort to become useful to the field. However it rarely happens. Many AGI-pilled people outside of safety hubs default to reading alone, or worse, going about life as usual.

This is because it is hard to find peers with shared interest, availability, and motivation to meet on a recurring basis. We hypothesize that if the friction of finding peers with these three traits were lowered, more people would gain the motivation to upskill autodidactically, enabling qualified talent to grow faster than the expansion rate of fellowships like MATS, Generator, or Horizon.

While the odds that there are people in your local area who share both your availability and your specific AI safety interests are low, the odds that a good fit exists globally is near certain. So how can we connect these suitable people together online?

Mangrove is that connection layer. Members share their interests and paint their weekly availability, then post project ideas or apply to existing ones. When enough people commit to an idea and their schedules overlap, Mangrove forms the team and provisions everything it needs to start: a private Slack, workspace, and suggested meeting times. Finding the right peers goes from months of coffee chatting to minutes via the software.

• 62 member alpha pilot with positive feedback- including 6 members asking if their peers could get access.
• 14 projects ideas posted, of which 7 have kicked off as group projects, with one already submitting research to a UN Forum.
• A functional platform with native project management software, provided to all users for free (AI meeting notes, automated todo tracking, scheduling, and more).