I still don't get it. why would they need an AI model to inflate user counts like this?
There are much simpler ways to create users than spending $100 of millions on gpus. The technology of batch inserting rows into a db, and AI chat bots have existed for a decade.
I’ll explain, but keep in mind that this is an exercise in cynicism, since none of us really know what’s going on.
Currently many tech company stock prices are driven by their potential future with regards to AI. Investors are looking for AI plays, hoping that they’ll generate large returns as Web two companies did.
Meta wants to remain a leader in tech. In order to do that And keep commanding higher and higher stock prices, they need to invest heavily into AI. They would invest into AI with no practical reason as long as it looked good to investors.
Meta is investing heavily into AI without a clear monetization plan, so they need to find a use for it or some “Proof” that it drives the metrics that meta correlates with value.
To that end creating a ton of AI driven engagement is a no-brainer. It shows that the company is forward and has an AI plan and their numbers are growing both things that investors like to see.
The millions spent on GPUs will return billions of dollars shareholder value.
I understand that AI can encourage users to engage more with the platform... but the original idea was that FB would count AI as a user, not that humans would interact more with the AI.
If FB was trying to mislead about their user counts, they don't have to use AI to do that.
Doesn't that strategy only work for a temporary amount of time? Surely investors will ask questions when Facebook claims that it has 100 billion active users?
Then next step is to not show anything human created to other humans - because human created content needs moderation, people swear at each other.
Imagine how pleasant and disinfected content will be there, every company will be happy to pay for ads because there will be no offensive comments and no images of your aunt. All the content perfectly aligned with what people want to see and there somewhere in between ads.
Kids in 10 years or 15 won’t even know that in past people actually could post something on the internet- they will think it was all like TV for us that only rich and handsome people get to post on the internet.
Who wouldn’t buy all those things that happen to be presented by perfect people in perfect world that you don’t have real access to but you can read on internet or have a part of that world by buying some crap they advertise.
It already happens but FB/Meta doesn’t have control over influencers - they can create their own now and deplatform all real ones because they won’t need them. Companies love that as it won’t turn out some influencer that they bought ads with died overdosing drugs somewhere on the street.
Why would they count an ai model as a monthly active user?