Reddit employees will be able to easily know that more than 1000 subs went dark. They’re also probably well aware about how ad revenue works on the platform.
This message wasn’t really aimed at the employees. It was aimed at the MBAs and bankers who are preparing the IPO. It was meant for investors to say “everything is fine and this is normal…”
Specifically infinite/increasing-growth capitalism.
They need to grow profits/value. These social media tech companies have about saturated potential user base. People are foreseeing a slow-down in ad budgets, and user-data is becoming a commodity. These realities, plus the need to drive up profits, leaves these companies with “no choice” but to seek more and more user-hostile features and rent-seeking behavior to satisfy investors. This was probably a long time coming, and was sorta delayed by valuation gains via an overpriced market. Now the investor love affair with tech has cooled a bit, and they’re asking more for hard-profit on the balance sheets.
The other thing that is ripping through social media platforms is a kind of ‘outrage’ or FOMO among leadership that LLM builders like OpenAI are eating lunch at their table by consuming their user content for training. This feeds into the previous point because they think that their users content can be a product of its own beyond attracting more users or being leveraged by ads. I am not so sure about this because these companies have already gotten their ‘lunch’ if they wanted it, and I am dubious about the value of a model built on data as provided by Reddit’s finest minds, and these platforms have patently failed to stop disinformation bots (which will get worse from these LLMs). But who knows.
Something to keep in mind is that these companies aren’t concerned with total profit or revenue or anything like that - it’s all about the percentage. I suspect in the short term, these AI-articles will look very profitable. Networking effects, consumer habits, and SEO will carry the day for a time.
But what always screws these MBA types is the inability to recognize the specific natures of their business and the second order effects. Not all costs are representable on a spread-sheet.
Basically, the second order to me really boils down to this: AI generated content isn’t really a ‘brand’. Good writing shops tend to build a following with their writers and expectations with their editors. The writing, investigative, and editorial bent of a house is essentially what makes a shop. See The Economist and The New Yorker as examples. In other places, a lot of niche shops are selling personality as much as product with youtube, podcasts, and others.
this means there is no real ‘value add’ someone like an AI shop can provide. You are throwing yourselves down the hole of becoming a pure commodity, and as every business major knows, being a commodity sucks. Short term profitable, but literally no one cares about where a mass produced nail comes from and its a race to the bottom of price.
So, as time goes on, with the barrier for entry being incredibly low, every bill and joe who fancies themselves an SEO wizard has no reason to not jump in, so your competition rises and your ability to charge some value for (ads?) drops a lot. But that’s the tip of the iceberg. Many of the companies that would occupy this brandless, commodity-filling space are way better positioned to make a run at it than the GAMURS Groups of the world. Microsoft’s Bing chat and (probably soon to follow Bard) will whip your ass in the long-game. Why search Bing to get an AI article from the Escapist when Bing will do it for me? I really doubt anything churned out by an AI with some edits will be that much better per convenience.
This whole could easily collapse in on itself. Like a lot of people in the AI space, I’m interested to watch what happens when AI begins to consume and be built on its own content.