This is the repo for swyx's blog - Blog content is created in github issues, then posted on swyx.io as blog pages! Comment/watch to follow along my blog within GitHub
I last talked about my Latent Space adventures in April and last December. Even as a well regarded developer-part-time-creator, the Latent Space Newsletter + Pod has done much better than I usually do. Here are the stats as of today:
For posterity, here is the front page as currently stands:
and our top posts:
a big part of the thesis for using Substack instead of anything else is the built in recommendation network and it's obviously worked in growing subscribers:
altho we are still very much a net donor of subs:
celebrity guests of course work - we briefly broke into the US Tech Top 10 chart:
Reflections
I'd say I mostly failed to keep up with regular writing, altho I do combine much of my writing with podcast notes (eg the Code Interpreter post, where I've had multiple people compliment me in person for my extensive footnotes). The podcast thesis worked out in that having Alessio as cohost made it much more feasible in terms of financing, load balancing work/questions, having a sparring partner on questions/content strategy, and general accountability. He's taken up writing the show notes for many of our recent scheduled interviews, helped by smol-podcaster. I still retain editorial control over titling, and editing, and am mostly still sole author on the longform essays.
I've had some minor success with cross promo of podcasts:
However it's very very unclear how many subscribers they actually drive. Just adds to the social proof I guess.
On the Essay side I have not written much, but the one thing I had to do right I seem to have done well enough (the Rise of the AI Engineer), and we have rebranded both the pod/newsletter and are in the process of building a successful conference series accordingly.
Substack doesnt take podcasting seriously, and I know we have weaknesses in our LinkedIn and YouTube game. We briefly explored joining a podcast network, but demurred because of concerns about control and ad load. I'm still keen on it in future but want to get the timing and situation right.
What else would you like to know?
An aside on Topic Selection
I think the #1 job of the content creator, whether writing, speaking or podcasting, is Topic Selection. You must have an ear for what the Current Thing is, a mental model for what drives people to click/subscribe/share, an eye for a good, underrated story that can go viral if told well, some social/street credibility to book important guests.
But Topic Selection is #1.
FAQ from people
Are you editing/producing your podcast on your own or do you have an editor?
have editor for the studio recordings. twitter spaces are done on my own
how much are you paying per episode for the editor? wondering whether using Descript and doing it by myself for now is fine?
about $150-250. yes it is, good mic > good editing. the SMB7 and the DJI Mic for travelling are good enough.
do you have thoughts on doing a podcast with video vs without?
where possible, do it, youtube is extremely powerful platform. we only dont do it because we are lazy. however, most podcasts struggle on youtube bc they are not well edited/recorded/interesting. it may be a better strategy to just be "youtube first", and then strip out the audio for podcast, rather than be "audio first but with video".
category: note slug: latent-space-250k
I last talked about my Latent Space adventures in April and last December. Even as a well regarded developer-part-time-creator, the Latent Space Newsletter + Pod has done much better than I usually do. Here are the stats as of today:
and the pod is seeing a nice uptick today being on the HN frontpage all day:
For posterity, here is the front page as currently stands:
and our top posts:
a big part of the thesis for using Substack instead of anything else is the built in recommendation network and it's obviously worked in growing subscribers:
altho we are still very much a net donor of subs:
celebrity guests of course work - we briefly broke into the US Tech Top 10 chart:
Reflections
I'd say I mostly failed to keep up with regular writing, altho I do combine much of my writing with podcast notes (eg the Code Interpreter post, where I've had multiple people compliment me in person for my extensive footnotes). The podcast thesis worked out in that having Alessio as cohost made it much more feasible in terms of financing, load balancing work/questions, having a sparring partner on questions/content strategy, and general accountability. He's taken up writing the show notes for many of our recent scheduled interviews, helped by smol-podcaster. I still retain editorial control over titling, and editing, and am mostly still sole author on the longform essays.
I've had some minor success with cross promo of podcasts:
However it's very very unclear how many subscribers they actually drive. Just adds to the social proof I guess.
On the Essay side I have not written much, but the one thing I had to do right I seem to have done well enough (the Rise of the AI Engineer), and we have rebranded both the pod/newsletter and are in the process of building a successful conference series accordingly.
Substack doesnt take podcasting seriously, and I know we have weaknesses in our LinkedIn and YouTube game. We briefly explored joining a podcast network, but demurred because of concerns about control and ad load. I'm still keen on it in future but want to get the timing and situation right.
What else would you like to know?
An aside on Topic Selection
I think the #1 job of the content creator, whether writing, speaking or podcasting, is Topic Selection. You must have an ear for what the Current Thing is, a mental model for what drives people to click/subscribe/share, an eye for a good, underrated story that can go viral if told well, some social/street credibility to book important guests.
But Topic Selection is #1.
FAQ from people
have editor for the studio recordings. twitter spaces are done on my own
about $150-250. yes it is, good mic > good editing. the SMB7 and the DJI Mic for travelling are good enough.
where possible, do it, youtube is extremely powerful platform. we only dont do it because we are lazy. however, most podcasts struggle on youtube bc they are not well edited/recorded/interesting. it may be a better strategy to just be "youtube first", and then strip out the audio for podcast, rather than be "audio first but with video".
transistor is fine, but imo has stagnated for the past year. i ended up choosing substack because i think email distribution is important