I’ve had a drop off of non EA friends who are willing to listen to ai episodes. I think the ai stuff is repetitive. I liked the podcast more and recommended it more before the shift to focus on AI and I am an AI professional interested in safety. I think the priority should be interesting conversations.
Separately, most podcast listeners I know subscribe to podcasts. There is an effect where publishing too much may cause people to unsubscribe, when it starts to feel like spam. I doubt the main thing preventing further expansion is quantity. I don’t think that a third host would meaningfully help things.
With respect to numbers of episodes: - We’re considering potential new hosts publishing on new feeds. They might try a different format, or they might try the same format but a different target audience— Over the last ~year, most of our growth has been on youtube, rather than audio feeds. As far as we can tell, on youtube it’s very uncommon to find new episodes directly via a subscription, and relatedly there seems to much less of an effect of reduction in engagement if you have episodes more than once per week.
This is all pretty tentative. For example, we might find that the audiences we can reach on youtube turn out to be much worse than via other means. That might push us towards trying to increase the quality rather than quantity of episodes. But it would also help enable us to improve quality if we had more capacity—Rob has historically felt the need to rush to get episodes out, and could usefully spend decidedly more time researching / editing episodes.
I’ve had a drop off of non EA friends who are willing to listen to ai episodes. I think the ai stuff is repetitive. I liked the podcast more and recommended it more before the shift to focus on AI and I am an AI professional interested in safety. I think the priority should be interesting conversations.
Separately, most podcast listeners I know subscribe to podcasts. There is an effect where publishing too much may cause people to unsubscribe, when it starts to feel like spam. I doubt the main thing preventing further expansion is quantity. I don’t think that a third host would meaningfully help things.
With respect to numbers of episodes:
- We’re considering potential new hosts publishing on new feeds. They might try a different format, or they might try the same format but a different target audience—
Over the last ~year, most of our growth has been on youtube, rather than audio feeds. As far as we can tell, on youtube it’s very uncommon to find new episodes directly via a subscription, and relatedly there seems to much less of an effect of reduction in engagement if you have episodes more than once per week.
This is all pretty tentative. For example, we might find that the audiences we can reach on youtube turn out to be much worse than via other means. That might push us towards trying to increase the quality rather than quantity of episodes. But it would also help enable us to improve quality if we had more capacity—Rob has historically felt the need to rush to get episodes out, and could usefully spend decidedly more time researching / editing episodes.
Growth coming from YouTube recommendations makes sense. In that case I agree that more episodes is not a bad thing.
Thanks for the feedback!