AI as a distribution multiplier

Apropos of nothing, I think I’ve finally found a hot chocolate recipe worth sharing. Full fat milk, pure cocoa, vanilla, and brown sugar simmered just under a boil, and then —this is the critical part — blended at high speed for 45 seconds in a smoothie blender.

Now that’s out of the way, our theme for the week:

AI as a distribution multiplier

Sometimes, the medium is as important as the message itself.

There’s a reason we pay for college instead of spending four years at the public library. There’s a reason novels and movies are all of a common length. There’s a reason “tweets” are distinctly useful alongside “posts”.

Multiplying content across multiple mediums isn’t just a growth hack, it’s a valuable service. People learn in different ways, consume in different ways, and enjoy in different ways.

But what does this have to do with AI?

LLMs offer the possibility of taking old content and re-targeting it to new mediums.

• Sometimes in old ways (“write a LinkedIn post for this newsletter article”)

• Sometimes in new ways previously impossible (“re-enact the 1927 Solvay Conference on Quantum Mechanics”)

Here’s are two axes of medium retargeting I’d like you to think about as you browse the week’s links below.

The exes are the roles of the human and the machine. At the bottom-left are the personalized feeds we’re served up today. At the top-right are personalized AI tutors that act as everything from sparring partner to warmly demanding coach.

This week’s links

• Podly by CedricMakes - Transcripts for every Lex Fridman podcast.
  
  What is the “pivot table” for long form podcasts? Is transcription the first step toward building a round-up episode of “every Lex segment on Joseph Stalin?” Or DMT? Or self-driving cars? I’d subscribe.
  

• Chat with your favorite podcast by Dan Shipper
  

  Every podcast listener will soon be able to ask their own questions while listening. I don’t agree with the “but LLMs make things up” rebuttal. We won’t mind that medium-specific quirk, just as you don’t mind that this newsletter is only a filtered portion of what I think.
  

• Podcast AI - Interviews with historical figures
  
  This is a fantastic glimpse into where generated content is heading. If AI becomes good enough to accurately represent the worldview of historical figures, you’d better believe people will jump at the chance to listen to them comment on today’s news.
  

• The Infinite Conversation - Werner Herzog and Slavoj Žižek Bots chat endlessly
  
  As far as I can tell, this is the first instance of notable figures converted to AI form and republished in “longform podcast” style. This site has the feel of something that belongs in an art gallery. The Thinker, thinking out loud.
  

• Anypod by Praneet Kedari, Marco Mascorro, and Safwaan Khan - Full-text search for your favorite podcast
  
  This — or something like it —belongs in every podcast reader. I want a button that lets me pause a podcast and enqueue other episodes where a topic had been discussed. Building a curriculum for a topic as I listen.
  

• audio-analytics - by Enias Cailliau - Create an entity + sentiment search API for any audio & video
  
  I’ve always wondered where the writers for The Daily Show search their news archive. Well now we call can. Feed in video & audio files, and then query for “clips where Person A says negative things about Topic B.”
  

That’s it for this week!

Have a request for a theme or tips on a great project to check out? Let me know!

Shipped with AI is brought to you by Steamship, a package manager for hosted AI apps.