Stop making sense: the impossible social network
Plus the trouble with obituaries written by AI
By now, we’re all used to Apple having a product launch in September at which New Things are unveiled which become the focus of huge media attention, and go on to garner huge usage, revenues and lasting success. The iPhone announcement has become a marker of the technology seasons, indicating that summer is over and that we’re in for three months of news grind. And when Apple announces a move into a new sector, it indicates that the time is ripe to exploit a burgeoning niche that’s ready to explode. Examples include the iPod, iPad, Airpods and of course the iPhone itself.
But on Wednesday September 1st 2010, Steve Jobs himself introduced a new Apple product which rapidly garnered millions of enthusiastic users, aimed at a sector which seemed to be red-hot, and yet which in two years was put on ice, its existence scrubbed from the company’s site. That was Ping.
The video of Jobs introducing Ping is on YouTube; it’s just over nine minutes. He begins by saying
Today, we're announcing something really cool. And we call it Ping. What Ping is, is it's a social network for music. It's sort of like Facebook and Twitter meet iTunes. It's not Facebook, it's not Twitter, it's something else that we've come up with: it's a social network all about music, and it's built right in to iTunes. So you can follow your favourite artists and friends, and you can discover what music they're talking about listening to and downloading.
The emphasis was on downloading, of course, because at that time streaming was not the dominant mode for (legally) listening to music online. Though Spotify had started in 2006, and in 2010 began offering its free tier in the UK, there wouldn’t be an offering in the US until July 2011. Napster, in its legal form, was struggling to gain acceptance. People’s mental models hadn’t adapted to a streaming model for music, despite the huge popularity of radio; music was still thought of as something you had in your possession.
Jobs continued his demonstration, sketching out what Ping could do and how you’d use it:
So here's a post by one of your friends that you're following, here's a post by an artist—Lady Gaga—that you're following, another friend right here—you can find people by just typing in their names, and if if they've registered for Ping and they've held their hand up and said you can follow me, you can sign up to follow them with just a simple click. And you get a custom top 10 chart of songs and albums that are customized to what only the people you follow are downloading from iTunes, just for you. And it's really cool.
So there were artist pages and fan pages, and you could tell your favourite artist what you thought of them, and comment on their fabulous photos. You’d also be told about upcoming gigs by any artist you were following. Ping, Jobs pointed out, was open right away to more than 160 million iTunes users in 23 countries from day one. Wrapping up, he noted:
Ping is not just available on your computer. It's available on your iPhone and it's available on your iPod touch. It's showing up right in the iTunes Store. There's a new button right in the middle called Ping, you push it and you get your recent activity right on your phone or your iPod. It's amazing. So, Ping. It's a social network for music created by Apple.
That was the Wednesday. On the Thursday, as Wikipedia recounts
Twenty-four hours after Ping was launched to the public,[20] reports of the service being flooded with spam were published. The fraudsters would create an iTunes profile and post links to a number of online scams, including ones that promised "free iPhones" or "free iPads" in exchange for filling out online surveys. For the most part, these suspicious links were being posted in the comments sections of popular artists on Ping, such as Britney Spears, Lady Gaga, Katy Perry and U2, all of whom were among the recommended accounts listed on the Ping homepage.
Other problems, such as fake accounts, also blossomed, demonstrating forcefully to Apple that there’s a lot more to running a social network than just enabling people to connect with one another. Even so, within a couple of days a million people had activated their Ping account. (Some of them might even have not been spammers.)
Jobs however had a separate problem: though he had touted (and shown off) Facebook integration, Mark Zuckerberg decided to play hardball: though Apple wanted people to be able in effect to transfer their social graph over from Facebook to Ping, Zuckerberg wanted payment of some kind. Jobs found the terms too “onerous”, and despite attempts to charm the young Facebook founder over dinner, failed to get his way.
Yet iTunes was, at the time, a behemoth. With 160 million users, many of them in the US, and social networking exploding in use as smartphones became commonplace (2010 was the inflexion year when penetration really took off), you’d think that Ping should have stood a chance simply on its own terms. A social network where you get to see your friends’ musical tastes, and share your thoughts about what they’re listening to, and to interact with (or at least stan) your favourite bands. What went wrong?
I don’t think that it’s necessarily that Apple couldn’t run a social network, though of course the spam problem shows that it was unprepared for the task. I think it’s quite different, and can be answered by considering this: how many friends do you have who you like specifically because of your shared love of a particular artist or kind of music? I’d bet the number is pretty low, unless you work in the music business.
Bizarrely, considering how central music is to our culture, you just can’t build a many-to-many social network centred around music. Though it feels like the sort of thing that would be easy to do (people like music!) and even to monetise (people want to know when their favourite bands have new albums out and are playing gigs!), the underlying challenge is far bigger than that of Facebook or Twitter.
The problem is that our tastes in music are far more nuanced than, say, sports or politics. Even a seemingly trivial question like ranking the albums of a group that you and someone else likes can lead to wild disagreement. Try ranking three Arctic Monkeys’ albums: the first one, Whatever People Say I Am (That’s What I’m Not); AM; Tranquility Base Hotel and Casino. (And if you’ve never heard any of them, look them up.) Same band, same singer and lyricist, three totally different approaches.
Now consider trying to shoehorn your Facebook social graph onto that potential for disagreement, and discovering that the person who you really like for their sense of humour is a massive [insert artist you can’t abide (ayca)] fan. Would you be able to see them in the same light again? Sure, they’d make those jokes, but you’d also know they’re an ayca fan.
Music is complicated because our liking for songs, and albums, and artists, is so multi-dimensional. Sometimes our love of a song is because of its Proustian ability to transport us back to another time and place (as the Elbow song describes). The Tricky song “Aftermath” (which is on as I write) uses the sound effect of a Motorola pager alert—which always takes me straight back to my days on The Independent, when we reporters wore them on our belts, and would get buzzed with the message CALL NEWSDESK with exhausting regularity. Music we care about anchors us.
But that multi-dimensionality makes social networking around music impossible. If you want a way to think of your own musical tastes, think of three artists you like. For the first artist, plot their songs/albums on a 0-1 continuous X/Y scale, and join the dots. Great: a line. Maybe a bit up and down. For the second artist, do the same, but imagine the X/Y scale at right angles to the first one (extending toward or away from you). Now join the lines: you’ve got a plane. Now do the same for the third artist, and join that line with the plane: you’ve got a surface. Maybe it looks like this:
(In which case, congratulations on your vivid 3D imagination.) Now imagine you keep doing this for all the artists and songs that you like, adding dimension after dimension and mapping them. And next imagine you do this for all the artists and songs you like. You can’t imagine it, but a computer could compile a matrix that would be a representation of how much you like those songs. It would have to do a lot of guessing, based on how often you played the songs, and Liked them, and perhaps shared them with friends. Mathematically, it’s not hard.
The difficult bit comes when you try to find someone who has even vaguely similar musical taste to you. The computer has to generate those matrices and then try to find a good fit between them. Yes, yes, machine learning, but just try to consider the challenge: Person A likes the Arctic Monkeys’ AM. Now, do you recommend they next listen to Whatever People Say or Tranquility Base? Or perhaps neither of those? Your music library, like you, is unique.
But, you say, my sense of humour, my politics, my favourite football team, the topics I find important in life—all those are just as variable as music. Which is true, and yet on social networks we tend to follow people, and we frequently tune out things they say that we don’t find interesting. By contrast a social network based around music is like going to a huge cocktail party where the music’s very loud and you hate every second track.
There’s a counterpoint to all of the above argument in the form of MySpace, which thrived on music—most famously, for me at least, in bringing Lily Allen’s first album to notice. When Ping launched, quite a few commentators thought Apple was trying to scoop up users from the then-dying network. Certainly, when social networking was on the rise, MySpace was an important space for finding music, as this article at Stereogum relates:
at Myspace’s height — roughly 2005 through 2008 — the website changed the way artists and fans found each other and how record labels and buzz-seeking blogs found fresh meat. Artists like Panic! At The Disco, Arctic Monkeys, Soulja Boy, Lily Allen, and Colbie Caillat would become pop stars in part because of their presence on the site, whereas artists such as Los Campesinos! or Nicole Atkins would eventually settle into cult careers after navigating through the sudden, unexpected attention the site could often generate.
There’s plenty in the article (paywalled) about artists who bootstrapped themselves to fame through the place where Tom Was Your First Friend—artists including, wouldn’t you know it, the Arctic Monkeys. However I don’t class MySpace as a social network organised around music in the way that Ping was. MySpace was organised principally around friends, and their myriad of shared interests; in the end it tried to get people interested in so many things at once that it collapsed in a black hole of horrendous design.
Ping, meanwhile, never took off. Even by the end of 2010 it was being described as one of the year’s biggest flops. Apple discontinued it just over two years after the launch, at the end of September 2012. By then Steve Jobs was dead, and Tim Cook was firmly in charge. Asked about Ping in the summer of 2012, Cook observed that “the customer voted and said 'this isn't something that I want to put a lot of energy into'.” Apple decided the same. And so the idea of the music-based social network died. Nobody, not even Apple, can make the matrices overlap.
Glimpses of the AI tsunami
(Of the what? Read here. And then the update.)
• AI video language translation with lip movements. Post-processed, but people are trying it and are amazed. Now your CEO doesn’t need to learn languages to do that promotion video for all the other countries.
• Stability AI announced a text-to-CD quality audio generation system called Stable Audio. Ars Technica observes: “Before long, similar technology may challenge musicians for their jobs.”
• Microsoft’s MSN publishes garbled AI article calling deceased NBA player “useless”. Don’t let your AI systems write obits, people.
• You can buy Social Warming in paperback, hardback or ebook via One World Publications, or order it through your friendly local bookstore. Or listen to me read it on Audible.
You could also sign up for The Overspill, a daily list of links with short extracts and brief commentary on things I find interesting in tech, science, medicine, politics and any other topic that takes my fancy.
• I’m the proposed Class Representative for a lawsuit against Google in the UK on behalf of publishers. If you sold open display ads in the UK after 2014, you might be a member of the class. Read more at Googleadclaim.co.uk. (Or see the press release.)
• Back next week! Or leave a comment here, or in the Substack chat, or Substack Notes, or write it in a letter and put it in a bottle so that The Police write a song about it after it falls through a wormhole and goes back in time.
I'm reminded that MySpace (bought by Murdoch for $500 million, sold for $25 million to a marketing company) forms the basis of the largest collection of marketing data outside of Google and Facebook (and it's continually updated with new marketing data every day).