About AI Safety Support

Origin: 

AI Safety Support (AISS) was created in mid 2020 by Linda Linsefors and JJ Hepburn. Originally AISS was just a brand and a website to promote and support a variety of AI safety field-building activities conducted by the two founders.  We deliberately did not make any long-term plans for AISS, but instead focused on filling the most urgent needs we noticed in the community while knowing that these needs may shift over time. 

Mission: 

Our mission is to reduce existential and catastrophic risks from AI by supporting everyone who wants to work on this problem. We want to give as many people as possible the chance to succeed.

Our guiding principles:

• Research is done by people. The best way to get excellent research is to take excellent care of current and future researchers. We believe in kindness and friendship, both because these things are instrumentally important to success and because we care about the individuals involved. 

• Helping instead of vetting. We will not waste your time by asking you to prove to us that you are promising enough. We will just get directly down to helping you. Because that is how we’ll both find out what you can do.

• Same-sidedness. If you want to save the world then we want to empower you, because your success is our success. You don’t even have to agree with us regarding the previous two principles. Even if you think that stuff about kindness is naive nonsense, and you only want to save the world for selfish reasons, we still offer to help you.

What we actually do (currently):

• Fiscal sponsorship. Setting up your own legal org and bank account is a lot of work. Becoming a registered charity is even more work. If you don't want to do this yourself AISS can help you.

• Various events (maybe?). AISS have experience running various AI safety event, especially online events. Most recently we organised the Virtual AI Safety Unconference (VAISU). This is something we would like to continue doing, but it's currently unclear if we have the funding to support this.

• Looking after the Alignment Slack (sort-of). Shortly after the founding of the Alignment Slack, AISS took over the stewardship of this space. In the early days we put in a lot of work to feed it with content. Now however, the slack has enough momentum to mainly just run itself. However, AISS is sill the main moderators, when some new channel connections needs approval, and that sort of things.

• One-on-one career advice. For a long time this was a corner stone of AISS. We talked to everyone who reached out to us, no applications needed. However due to lack of funding we could no longer maintain this service. Fortunately AI Safety Quest (AISQ) has picked up this mission. AISS is working with AISQ to pass on our experience in this area.

• What ever we notice need fixing that is within our power to fix.