Project summary

AI Safety Events is an organization supporting the realization of events beneficial to the field and community of AI safety. The goal of AISE is to reduce global catastrophic and existential risks from advanced AI by building the field and AI safety research community.

Project goals

AISE's current high-level objectives are:

• Organizing events at ML/AI conferences, hackathons, retreats, etc

• Providing event operations services for organizations and researchers

• Tracking all events in the field of AI safety and offering a calendar and newsletter

• Becoming a coordination/Schelling point of event organization in the community

How will this funding be used?

For the second half of 2023, we are seeking a 25K-100K grant.

Funding will be used for worker salaries, funding our own events (venues, catering, logistics), subsidizing attendees and offering stipends to speakers, tooling such Zoom, Notion, etc, training material such as the ‘Art of Gathering’ course, and registering the org legally in Canada.

Conditional on minimal funding, the median outcome for the second half of 2023 is:

• Organize the AI Safety Unconference NeurIPS 2023 even more successfully than previous years

• Hire an operations generalist 

• Setup a global AI safety events tracker and related newsletter

• Build a new website
  

The 10%-best percentile outcome:

• Start preparing for new conference events in 2023 at ICML, ICLR, etc

• Proactively support other efforts such as conference workshops and online events

• Organize a retreat for independent researchers

• A long-tail of smaller goals

• (possibly) AI Safety + Crypto private event

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

- ML Safety Social at NeurIPS 2022

- AI Safety Unconference at NeurIPS 2022

- AI Safety Unconference at NeurIPS 2019

- AI Safety Unconference at NeurIPS 2018

Reference: https://aisafetyevents.org/events/

In July 2023, the Center for AI Safety did a cost-effectiveness analysis of the kind of events of , including specifically of the ML Safety Social we helped realize last year. https://forum.effectivealtruism.org/posts/7kFPFYQSY7ZttoveS/cost-effectiveness-of-professional-field-building-programs

See some testimonials: https://aisafetyevents.org/testimonials/

How could this project be actively harmful?

Poor execution of the project could lead to wasted time of crucial people.

What other funding is this person or project getting?

We are currently operating with minimal self-funding, in limited capability.