Some approaches to solving alignment go through teaching ML systems about alignment and getting research assistance from them.[1] Training ML systems needs data, but we might not have enough alignment research to sufficiently fine tune our models, and we might miss out on many concepts which have not been written up. Furthermore, training on the final outputs (AF posts, papers, etc) might be less good at capturing the thought processes which go into hashing out an idea or poking holes in proposals which would be the most useful for a research assistant to be skilled at.
It might be significantly beneficial to capture many of the conversations between researchers, and use them to expand our dataset of alignment content to train models on. Additionally, some researchers may be fine with having their some of their conversations available to the public, in case people want to do a deep dive into their models and research approaches.
The two parts of the system which I’m currently imagining addressing this are:
Clear instructions for setting up a tool which captures audio from calls automatically (either a general tool or platform-specific advice), and makes it as easy as possible to send the right calls to the dataset platform.[2]
Me too! Also, if this or any of your other projects needs a domain, AED’s https://ea.domains/ might have a good match to offer. I’m also happy to host it on
Maybe a little late, but here is an android app that does recordings, you can contribute directly on the github.
Other potential project ideas that can help with this are:
An iPhone app
A discord bot like Craig that’s a little more streamlined in that after it finishes recording it automatically sends the file without more need from the user.
Some approaches to solving alignment go through teaching ML systems about alignment and getting research assistance from them.[1] Training ML systems needs data, but we might not have enough alignment research to sufficiently fine tune our models, and we might miss out on many concepts which have not been written up. Furthermore, training on the final outputs (AF posts, papers, etc) might be less good at capturing the thought processes which go into hashing out an idea or poking holes in proposals which would be the most useful for a research assistant to be skilled at.
It might be significantly beneficial to capture many of the conversations between researchers, and use them to expand our dataset of alignment content to train models on. Additionally, some researchers may be fine with having their some of their conversations available to the public, in case people want to do a deep dive into their models and research approaches.
The two parts of the system which I’m currently imagining addressing this are:
An email where audio files can be sent, automatically run through Whsiper, and added to the alignment dataset github.
Clear instructions for setting up a tool which captures audio from calls automatically (either a general tool or platform-specific advice), and makes it as easy as possible to send the right calls to the dataset platform.[2]
Ought’s Elicit is the prime example.
I hear OBS might be a good tool for this.
Thank you for providing this outline, plex. I hope we get good engagement with this project.
Me too! Also, if this or any of your other projects needs a domain, AED’s https://ea.domains/ might have a good match to offer. I’m also happy to host it on
Maybe a little late, but here is an android app that does recordings, you can contribute directly on the github.
Other potential project ideas that can help with this are:
An iPhone app
A discord bot like Craig that’s a little more streamlined in that after it finishes recording it automatically sends the file without more need from the user.
Something similar for zoom + meet + teams