Mark Cavendish crashed out of Le Tour on stage eight. It was a minor little bump and shuffle in the peloton that saw him tumble and break a collarbone. If you didn’t know, this was going to be his last tour before retirement in which he hoped to eek ahead of the 34 stage wins he shares with Eddy Merckx to become the world record holder. Since he returned a few years back it’s always been edgy. Cavendish is one of the last few hotheads in the sport; unlike the current crop of media-trained, elusive professionals, there was always a lairy threatening-ness to him, like an overly-effusive new ‘friend’ in the club. You could never tell if he was suddenly going to flip and start an argument or fight.
Underneath it all is the fact that, even when you’re a ‘legend,’ by the time you’ve passed prime, you’re basically just kept around for your PR value for the team and your legendary status. Your whole career and role in the team is no longer structured around promise or potential, or even a solid, recognised position, but the popular memory. In the past, that’s visibly and audibly weighted heavy on him, where he’s seemed tense and stressed, desperate to prove he’s still valuable. This tour was different though; he was cheery, relaxed, playful. This was his last tour and he only needed one victory to have ticked off his life’s ambition and retire happy but with a little tumble that’s been taken away.
It’s doubly sad because Cav is only a little bit older than me. Elite-level cycling is obviously a different ball-game (pardon the mixed metaphors) than hot reckons and whatever it is I think I’m doing, but knowing someone his age is considered ‘veteran,’ ‘grandee,’ and ready for retirement does dig a little. Anyway, he may yet get another chance, it’s just a case of whether he wants it or not.
An algorithm for viral
So I did sign up for Threads because I am a person with a career and a morbid curiosity in the world. But last week I thought about the death of social networks as a hardware and energy problem: You’re storing huge amounts of media on these huge server stacks costing massive overhead on the off-chance that 0.00001% of it goes viral and propels a bunch of ads into a couple of million people’s eyes. The input/output formula just doesn’t work. The dream of converting eyeballs into $$$ has always been promissory; that one day someone would crack the perfect formula and balance the scales and so the investment had, until recently, kept flowing.
Now generative AI comes along with the promise that you don’t need to store petabytes of dog videos and reactions to train crossings; you can just generate content on-the-fly and microtarget it. And, undoubtedly, inevitably, any day now, some AI startup is going to claim to have cracked the formula for vitality. They will claim to have the algorithm that can read the zeitgeist, plug into the news, scan the trends, identify the audience, read the airwaves man, and perfectly position a perfectly generated viral clip, image or tweet.
And it actually doesn’t matter that that’s not true; that virality, like art, is elusive and gestalt; not reducible to measurable and modellable qualities because tech is promissory, it’s always what it’s going to do, not what it actually does. However, what the crack Ai viral algorithm will do, will be a megamagnet for VCs now disillusioned with a social network innovation cycle which is just stealing and micro-slicing interfaces form each other in an an attempt to recapture the honeymoon period of when users saw them as a benefit rather than a necessary bother.
(Or as Bruce Sterling once said; ‘all tech companies eventually just become airlines; they’re a pain but you don’t have a choice.’)
Threads, as so many have said, is a bit of a mess. The Facebook-y stuff is good; the interface and experience of being captured by a cultish belief in abstract notions of ‘community’ and ‘connectivity.’ But the content is brittle; just awash with uninteresting nonsense from brands, celebrities and influence. Again, as many have commented, the benefit they have is a captive audience of people already on Instagram; what they don’t have is a reason to use it. Like the Labour party, simply being ‘not Twitter’ (not Tory) may not be enough to get it over the line.
So let’s think about it the right way round; Instagram and TikTok, the current big social networks media catapults are products of an advertising and content algo-logic that emphasises scale, multiplied by attention, to the non-zero factor of virality. AI is all about scale and the use of that to pinpoint that non-zero chance of virality. In that context, is the most profitable next model another waterfall of gibberish a la Twitter? Probably not, why have the noise when you can map the perfect path for a user to get to the ad you want. (Or at least you say you can and you built a platform around that promissory idea to sell to investors.)
A metaphor riffing on Bruce: In another conversation I compared social networks to Zeppelins; a small, tragi-comedy blip in the history of flight which over a hundred years settled around the mid-horror of budget airlines. Now it’s getting all hyped about hydrogen fuel to save or right-way a flawed and destructive model and tech bro evangelists tell us we’ll all be in personally autonomous flying taxis by 2045. Meanwhile, lots of us wonder why we don’t just have better trains.
I guess Threads is the speculation on hydrogen fuel, generative AI the autonomous flying vehicles. None of it works, but somethings are more palatable than others.
Short Stuff
- Dave Karpf again that Threads’ problem will be in mistaking Twitter for a public arena when it was really for journalists, academics and other liberal elites. It’s a good theory. Twitter always had a huge divergence in number of sign ups v active users with 10% of user doing 90-odd% of posting. There’s no new incentive for people to use the same form of interaction (save, perhaps the influencers and brands who are better served at Zuck’s house.)
- Gary Marcus launched the Center for the Advancement of Trustworthy AI which is too long a title and too forgettable. THe idea of ‘AI governance in-a-box’ is smart though and draws on philanthropy to my point above.
- Interesting paper delineating myth from metaphor in Causal Layered Analysis (if that’s your kind jam).
I think I’m drifting away from Are.na. I need somewhere to store reckons, signals, images and tools and Are.na is just too much of a moodboard to give me the cross-functionality I need. I’ve been using raindrop.io to save and sort things but as far as I know there’s no public face for it so there’s no way of sharing what I’m saving/reading.
Anyway, love you, bye.