The Week: Media, modelling and brand raiders
Links! As well!
Welcome, as always, to this week’s free edition of The Terminal. Here’s what I’ve published in the last week:
A look at MediaLab, a strange holding company currently engaged in a flurry of buyouts of depressed media brands. [$]
If you’re not a subscriber and you’d like to be, smash the button below:
Look on my works, ye mighty, and despair
I’ve been following with great interest the Ozy saga of the last week, which kicked off with Ben Smith’s column in the NYT.
I was honestly barely even familiar with the brand before it hit the headlines, which probably speaks to the central narrative here: a media company which claims to be very popular and widely consumed but ends up being largely a confidence trick played on investors and advertisers. From CNN:
Beyond Ozy, this story is about digital media spin; business models that don't pass a simple smell test; and unbelievable web metrics that people pretend to believe until forced to face facts. Gawker's Jenny G. Zhang said it best: "It is generally safe to assume that most once-buzzy digital media upstarts are houses of cards built on shaky valuations and inflated audience numbers that no one should invest in expecting huge returns (not us, though, we're different)."
This has been an ongoing conversation. A 2018 feature in New York magazine asked the question of how much the internet was fake, and concluded the answer was “a lot of it”. Platforms like Facebook have been repeatedly hit by scandals over inflating audience numbers and massaging metrics to look good for advertisers.
Tim Hwang’s book Subprime Attention Crisis goes further, arguing that the basic building blocks of the ad-supported web — everything from impressions to ‘engagement’ — are largely black boxes with arbitrary definitions, fomenting a possible collapse with plenty of parallels to the 2008 global financial crisis.
New subscription models (like Substack and this very newsletter) are an effort to escape from that escalating series of feints and magic tricks. Whether it’ll actually be sustainable into the future is an open question. I suspect there’ll be some serious bumps in the road, especially once we’re clear of the pandemic and people start wondering if they were crazy to subscribe to all this stuff during lockdowns.
Trust the model
I wrote about this last week, but it’s kind of interesting to see the country’s ‘government by modelling’ further fall to pieces as we hurtle towards reopening and some of the more doom-laden predictions haven’t played out. In the above interview, there seems to be a concession that modelling is as much about precipitating or encouraging certain favoured outcomes rather than providing neutral guidance.
Despite recent unrest over vaccine mandates, Australia and its states and territories have remained pretty steadfastly technocratic throughout the whole pandemic; acting very cautiously and always having Covid modelling to provide some objective basis for any plans. The phrase ‘trust the science’ has been sapped of all meaning, but it certainly seems like most levels of government have been ‘trusting the science’ throughout, even if the science itself is deeply contested in various ways.
What that has led to is a strange case of duelling models, where citizens of varying political persuasions who intuitively support various approaches to managing the pandemic fall back on their preferred modelling to support their program. There’s absolutely no reason why a normal non-expert should be arguing over the Burnet modelling versus the Doherty modelling versus the OzSAGE modelling — and yet this is the peculiar situation we find ourselves in, especially online.
Anyway, maybe it’s a win for the technocrats that so many Australians seem inclined to this way of thinking about governance.
Smash and grab
I wrote about this in more detail in a subscriber-only post this week, but I think it bears mentioning here because it’s such an interesting and under-covered story. This week, a company named MediaLab bought Imgur, which was once the default image-hosting solution for Reddit as well as a thriving (if insufferable) online community in its own right. The same company also bought Genius (formerly Rap Genius) for a paltry $80 million1 a couple of weeks ago.
A tour of MediaLab’s website does not contain much info about what it is or what it does other than being a “holding company of consumer internet brands.” But follow the rabbit hole down and you’ll land on the website for AssemblyExchange, the company’s owned and operated programmatic ad network. There you’ll see the company’s other brands, which include one time luminaries like Kik, Whisper and WorldStarHipHop. Along with Genius and Imgur, it’s a real rogue’s gallery of platforms that were briefly popular in the 2010s before being outmoded or made irrelevant.
Launched by Whisper founder Michael Heyward, MediaLab’s business model is simple: it picks up depressed media and tech brands that are past their use-by date for a steal, slashes their operations to the bone, and then blasts programmatic advertising at whatever users are still hanging around in the mostly abandoned carcass of the platform.
In my longer post I described it as like an aircraft boneyard for 2010s media brands. It’s a weird vibe. Circa 2014 or so, Rap Genius was something you heard about a lot – it was something cultural moment for a minute there. Now it’s being stripped for parts by a new kind of corporate vulture. It’s like the dark end of the miserable ‘there’s an app for that!’ period of the last decade.
This episode of Darknet Diaries from earlier this year goes into some detail about MediaLab’s administration of Kik, a messaging platform once popular with American teens and now possessing an intractable child porn problem. (‘Administration’ is a very loose term, given they don’t seem to do much of it at all.)
The Facebook files (again)
The Wall Street Journal’s Facebook Files investigation continued, and I think this tweet is right: the things that have really stuck in the broader discourse is the material about Instagram’s malign influence on teenagers and their mental health.
I often talk in this newsletter about the fact the definition of a ‘digital native’ is altered with each generation, with zoomer online life already being alien to many millennials. There seems to be an overwhelming sense that its always the next generation is the one that is too online, too disconnected from the real world, always on their damn phones, etc. (Let’s put aside the fact that it is the boomers who have indisputably become the most pudding-brained on contact with social media.)
I think child safety and mental health on social media is a real cipher for broader anxieties about what the internet and networked society is doing to the culture, and the rapidly accelerating gulf between generations as a result. Anyway, it’s interesting that these stories do seem to hit hardest (moreso than, say, huge data breaches or surveillance.)
This revelation also got some airplay, I think because the concept of ‘leveraging playdates’ activates the fight-or-flight impulse:
In a document from 2019, the company considered whether it could “leverage playdates” to get more tweens on Facebook, specifically by getting them to use Messenger Kids app while hanging out with their friends, The Journal reported.
One of the documents The Journal reviewed said Facebook’s goal was “messaging primacy with US Tweens,” which it said would have the knock-on effect of “winning with Teens.”
Presented without comment
This week’s reading
A look at how lame Facebook memes might be corralled as a data harvesting operation by… someone. I’m not entirely convinced (and the article asks more questions than it answers) but it’s plainly obvious something weird is going on there.
Liked this in The New Yorker: “On The Internet, We’re Always Famous”.
This about the “radicalising role” of Telegram in the German elections. Didn’t really seem to move the needle in the end.
This on dinosaur-hunting cowboys in the American west, who are turning ancient fossils into a “hot new asset class”.
I’ve wanted to read something about Miami becoming a centre of energy for cryptocurrency enthusiasts and a strange new generation of tech guys, and this was pretty good.
More and more people talking about the ‘metaverse’. This time it’s Epic Games CEO Tim Sweeney, who shares a narrative marginally different from Mark Zuckerberg and Facebook. But still, nobody pushing this seems to articulate a vision that doesn’t involve more seamless ways to interact with brands:
The simplest way to define the metaverse is as an evolution of how users interact with brands, intellectual properties and each other on the Internet. The metaverse, to Sweeney, would be an expansive, digitized communal space where users can mingle freely with brands and one another in ways that permit self-expression and spark joy. It would be a kind of online playground where users could join friends to play a multiplayer game like Epic’s “Fortnite” one moment, watch a movie via Netflix the next and then bring their friends to test drive a new car that’s crafted exactly the same in the real world as it would be in this virtual one. It would not be, Sweeney said, the manicured, ad-laden news feed presented by platforms like Facebook.
Interesting report in BuzzFeed featuring declassified State Department investigations of Havana syndrome.
This did the rounds extensively this week, but anyway: the Yahoo News investigation into CIA plans to kill Julian Assange.
Great podcast episode about stablecoins and central bank digital currencies, and how they might be regulated.
Absolutely loved this interview with Steely Dan’s Donald Fagen.
A weird but compelling story about building little horror dollhouses.
Paltry compared to many tech valuations in these frothy times, I stress. I personally would love $80 million.
"I think child safety and mental health on social media is a real cipher for broader anxieties about what the internet and networked society is doing to the culture, and the rapidly accelerating gulf between generations as a result. Anyway, it’s interesting that these stories do seem to hit hardest (moreso than, say, huge data breaches or surveillance.)"
Love to hear you expand on this :))