You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
Run a podcast for technical alignment research.
Interview alignment researchers to understand their research agenda.
To buy equipment ($350), pay for travel and a space ($60 per episode) and pay me for my time ($1000 per episode).
I am Algon, I was team lead for AIsafety.info, a project dedicated to sharing good models about AI safety. Before I ran the project, I was an editor there.
I have not run an alignment podcast before but I have run a long-standing reading group with interesting conversation with twitter people. We have covered Grand Futures, Anthropic Bias, The Strategy of Conflict, and Drexler's Nanosystems.
Socractic dialogue is a skill I've honed over the years, and alignment researchers I've interviewed have told me they appreciated my earnest desire to understand their models.
I turn out to be a boring interviewer, in which case I shutter the project. Which is why I'm asking for funding for four episodes.
As for getting guests, that shouldn't be an issue. Two alignment researchers have agreed to come on my podcast.
None for myself.