The financial and technological genius Elon Musk spent $44,000,000,000 on a social media site that he claimed was riddled with fake users, and which he tried to get out of buying when the market value of that website (thanks in part to his own criticism of it) dropped by an eleven-figure sum. Having been forced by the threat of a humiliating legal defeat to pay up for it, he is now engaged in ruining that website for many of its users.
This isn’t going down well with the advertisers that provide almost all of Twitter‘s revenue (or did, before some of the world’s largest companies, including Pfizer, General Motors and Mondelez, pulled their ads from the platform), or the banks that provided the finance for Musk to buy the site, who are now reportedly being offered around 60 cents on the dollar for that debt, which would represent a loss of roughly $5.2bn were it all sold at that level.
In his first email to all staff this week Musk warned that without a pivot away from advertising to subscriptions “there is a good chance Twitter will not survive the upcoming economic downturn”.
Readers who don’t play five-dimensional business chess at Musk’s level might have some doubts about a company that has been profitable for just two of the 16 years it’s been in business (2018 and 2019), which now has around $13.5bn in debt and which is facing dramatically reduced revenues going into a global recession. They might also worry that the mass sackings and resignations of thousands of its employees will create a situation in which the platform becomes unusable or unsafe.
Then again, it’s social media: the whole point is to waste time and money in quantities that would once have been unimaginable. Unsafe websites are all the rage! It looks a lot like TikTok shares your kids’ data with the Chinese government, but this hasn’t stopped it becoming the fastest growing social media platform.
This has been Musk’s argument for some time – that Twitter is valuable in another way, in that large numbers of influential people have put too much energy into it to leave. He may well be right. MPs have faced torrents of abuse on the platform for years, but almost all remain on it and most use it on a daily basis.
The only thing that would kill Twitter would be the sudden emergence of another platform offering a comparable but better experience. Fortunately, I have just invented one.
Musk’s problem is that he is trying to make his product more authentic, with a (currently very clumsy) focus on verification and subscription. But it’s not Twitter’s “real” users that are its true gold: it’s the bots.
After all, isn’t automation the future? Don’t we want software to take over the more onerous and depressing elements of daily life? If so, the unpaid work of vainly fighting an endless popularity contest with millions of people you’ll never meet is surely worth handing over to a computer. If you want to beat Elon Musk in taking over social media, you need to get out of the old world of user-generated content and let AI do the talking.
Let’s call the platform Gptr, because it would probably use some sort of generative pre-trained transformer (language software that can generate human-like text) and apparently it’s good to have a name that’s impossible to pronounce.
Here’s how it would work: you set up a profile and link your old social media accounts to it. The machine learning software goes through your old posts and starts automatically generating pictures and text. If you let it, it’ll post pictures you take on your phone or use GPS to spot when you’re in a nice restaurant and boast about it on your behalf. It will scan news articles and have public opinions about them that accord with your ethics. It will follow other people and promote things you’d probably like, if you weren’t busy doing something more fulfilling. It will construct an online persona that is as authentic as the one you would have made if you’d had to do it yourself.
You can check in and post as yourself whenever you like, of course, and the algorithms will prioritise this content to other real users. Or you can go outside and fall asleep in the sunshine under a nice big tree.
There are only upsides to this. For one, plausible deniability: the software can be blamed for every bad joke or ill-advised comment. For another, cybersecurity: AI can be more scrupulous about safeguarding the personal data that are a risk factor online. The Chinese security services won’t know if it’s you or your bot.
As a business, the big win is scalability. The reason all of Twitter’s users aren’t decamping to Mastodon is that they face weeks of tumbleweed as they find people they know and gradually build up a handful of followers.
Imagine how much more compelling it would be to be immediately welcomed by a network of 200 million bots. Your new profile would be followed, amplified and responded to from the moment you arrive. It’s just going to be a lot more rewarding, and let’s face it, these interactions would be every bit as authentic as the interactions you currently have with total strangers on social media.
And if you think advertisers would be put off, forget it – this isn’t a botnet, it’s Social Augmentation and it comes with huge commercial opportunities. Online personas can interact with brand content much more effectively than real people. In fact it’s likely that a lot of people, having automated their social media, would begin to accept automation of other aspects of online life, such as shopping.
This is easily a twelve-figure idea, and you should immediately take it to your nearest venture capitalist. Just don’t tell Elon Musk.
[See also: Whatever Mastodon is, it won’t replace Twitter]