Slashdot Mirror
Pandora CEO Roger Lynch Wants To Create the Podcast Genome Project (variety.com)
Janko Roettgers, reporting for Variety: Pandora's new CEO Roger Lynch has big plans for podcasts: Lynch told Variety on the sidelines of CES in Las Vegas Thursday that he wants to create "the equivalent of the podcast genome project" as the company plans to add many more podcasts to its catalog. Lynch, who joined Pandora as president and CEO in September, said that the company is working on a deep integration of podcasts that will allow users of the service to easily browse and discover new shows. Describing these efforts as a kind of podcast genome project is a nod to Pandora's Music Genome Project -- a massive database of dozens of musical attributes for every single song in the company's music library that is being used to compile stations and aid discovery. Pandora is also looking to offer podcasters monetization options that will be superior to the current state of podcast advertising. Currently, many podcasters still rely on ads that they read themselves on air, Lynch said. "It is not the most effective advertising model."
Let him!
I vaguely recall hearing about this company a long time ago, but I confess I've not heard of them since.
This sounds like a company trying desperately to remain part of the conversation, but I don't recall even seeing this mentioned in any news stories or seeing reference to them anywhere.
Or, all of the hep cats are using it and I'm completely unaware of it ... which I find true of most internet stuff these days.
It's amazing what not giving a fuck does for not being aware of these things.
Me, I have a huge music collection of MP3 ripped from the CDs I own, so I mostly just listen to that.
If their podcast algorithm works anything like their music algorithm, I'll pass... it gets worse over time and eventually becomes intolerable.
Which has more power: the hammer, or the anvil?
From http://www.pandora.com/restric... " Pandora is only available in the U.S. right now – but we are working on bringing our music service to other parts of the world. "
They have this message since they started offering service.
The cryptocurrency genome project: Blockchaindora!
That's the best/easiest way to approach the problem of listening to music, but what if you're trying to solve some other problem?
Suppose you wanted to automate suggestions to someone who doesn't have access to your files? Let's say you listen to Helloween all the time, but someone else listens to Gamma Ray all the time. Without using human brains how do you figure out that you and that other user are listening to similar-but-not-the-quite-the-same music and cross-suggest to each other?
Another problem: a guy who listens to Rob Halford a lot, bought a red widget. Another guy listens to Bruce Dickinson a lot but bought a blue widget. What color widget do you show in ads that you serve to Dio fans?
Advanced problem:
Person A, whose FBI file shows they live in Santa Fe NM has thumbs-upped "Moving Waves" by Focus and has a hotel reservation which ends today, in Durango CO.
Person B, who rated Jethro Tull's "Thick as a Brick" 5/5 and is a resident of Phoenix AZ just used his credit card to buy gasoline near San Diego CA.
Person C, whose phone's geo coordinates show he's almost always in Idaho, but whose phone now reports being in the state of Washington, keeps listening to "In Hearing Of" by Atomic Rooster over and over again though he hasn't formally rated it.
You are Jeff Sessions and you want to demonstrate your power and dick size to the public, by busting someone for interstate trafficking in supposedly-intrastate recreational marijuana, the scourge of society. Through careful circumvention of certain budgeting restrictions, you have amassed sufficient resources to set up one roadblock. Which of the three cars will have the most weed in it?
Good luck solving these problems with your puny MP3 collection.
to the usage of the term "genome" in this context
I do not believe in karma. "Funny"=-6. Do good and forbid evil. Yours, Oft-Offtopic Flamebaiting Troll.
I've been thinking about this for quite a while. Breaking this into genres should be pretty easy, dividing between non-fiction ("documentary" podcasts, audio books, etc, as further classified like books), fiction (non-music "entertainment" pieces, as further classified like movies), and presumably "other" for oddities that can't be classified that way (the fourth group is "music," which they've already got a good handle on).
For music, Pandora currently pays musicians and other experts to manually go in and rate numerous features for music because they're too hard to extract with machines; just like Uber and Lyft, the idea is that eventually this will be automated). For spoken word, this becomes pretty trivial. Cadence, tone, and similar metrics should be easy to get at least rudimentary metrics on, and the more important stuff (content!) should be a matter of state-of-the-art speech-to-text translation and then enter your favorite text parsing machine learning system(s).
Even categorizing into each genre should be really easy since the words can be easily extracted on nearly every title (when speech-to-text fails and there isn't already a transcript provided, perhaps there's room for some paid human transcription ... or perhaps that bars entry to the Pandora world, leaving the duty up to the content creators). Additionally, there's metadata on the listener's preferences for speakers' timing, cadence, style, and associations.
I've been thinking about this with respect to short-form standup comedy. Small quips that aren't knit into an overarching theme (the way you get in a larger standup routine), or that are knit into a common enough theme to weave together and create a custom comedy show on the spot. Think about "airline security" or "road rage" for example. A sophisticated ML system (with sufficient source material) designed to tell a story (not just a "playlist") with small standup segments would be worth a pretty penny.
Use my userscript to add story images to Slashdot. There's no going back.
Okay, why would I want to listen to podcasts on Pandora? They already have ads, and now I'd get Pandora's ads as well? Why not just keep pulling enclosure URLs out of RSS feeds and downloading those instead of adding a middleman?