because of increasing pressure to ban tiktok and pull it from app stores/etc to keep China from building profiles/models targeting westerners en-masse.
Everyone wants to be set up to be the American Tiktok when the plug gets pulled on the chinese one. And the plug is already being pulled by a number of countries and states and organizations.
It's an interesting one though because obviously we have ad profiles in the west, but should an "adversary state" be allowed to collect and warehouse that data, knowing it's potentially a national security threat for personalized targeting of natsec security compromises/etc?
Should western companies be allowed to collect this data either? But they are the "good ones" and only use it for advertising (and letting the three-letter agencies buy it, of course). But what basis do you use to ban it? And is it really any safer to have it, when the chinese are pretty good at hacking these data warehouses, and the security is often incredibly sketchy? They got the plans to the F-35, the federal Office of Personnel Management database, the RSA key-secret database... if you store it, they'll ultimately get it eventually too.
What if instead of the classical "this user is a gambler and in financial trouble, have an agent call them" targeting you build a personal weight-model on what things trigger engagement for that specific user? What if they show you OANN until you're so enraged you go attack the capitol? What if they troll you on War Thunder forums until you leak classified documents? And you can potentially do that en-masse, use AI to generate a personalized "havoc profile" of what content they should surface at this exact point to make you the most socially disruptive at some key moment, or the things they most need an intel leak on, and the AI can use that profile to surface the content that most effectively winds you personally up.
I'd expect it's probably possible (not sure if it exists) to do engagement/stimulation (arousal) and sentiment analysis in realtime. So you can know that if they're usually scrolling X when they're upset, that you should push content Y to push their buttons most effectively. Like you can literally micro-target people at their weakest moments with the content that most effectively pushes their buttons to accomplish whatever goal.
Essentially - what if instead of "AI girlfriend" it was "AI best-friend with links to extremism", and he's gently radicalizing you a little bit every time you're scrolling on the shitter? That's the product for Tiktok.
And while most people are not going to get wound up to the extent of "attack the capitol/leak classified documents"... statistically like all advertising, it works, some people will stochastically be stimulated to action. And even at a more pedestrian level, it's very easy to ramp up polarization and internal tension, like Russia has been doing during elections/etc. But you can automate that shit at a scale and precision of targeting far beyond traditional 1:1-scale intelligence work.
Selling more diapers and kleenex is baby shit, Tiktok is the point of the spear on true infowars shit, where everyone can be probed for their personal hot-buttons and then press them automatically, en-masse while you're scrolling. And we just use that to sell more McDonald's. Using a nuke to kill a mosquito.
It's a weird time we live in, this stuff essentially parasitizes our cultural values of free speech and free information information exchange and turns it into an attack vector. That's been what advertising is all along (it's the art of convincing people to buy shit they don't need/make decisions they wouldn't have otherwise made) but we are starting to go past the "benign toxoplasmosis" to something that can be weaponized for political ends.
Anyway that's why the security apparatus is getting wound up about tiktok. You absolutely will not get a security clearance if you're anywhere near tiktok, you will not get tiktok anywhere near a government computer, and lately countries are just banning it outright because it's too dangerous.
And reddit sees that and says "hey, we can be Tiktok for westoids and fill that gap". There is competition of course - Snap and Insta are really in a better place to pivot to this. But Reddit has been pushing towards being a content-drip for a while, that's been the entire point of the "new-experience" and first-party app, and this is just them making the leap. And that means third-party clients have gotta go.
That's the information I was missing, thanks! I had heard some complaints about TikTok a while ago but wasn't aware that things had progressed to the point where it's going to get banned.
I could definitely see a slim chance of being new TikTok having higher expected value than continuing on as old Reddit if TikTok is removed from the picture.
https://www.nytimes.com/article/tiktok-ban.html
Everyone wants to be set up to be the American Tiktok when the plug gets pulled on the chinese one. And the plug is already being pulled by a number of countries and states and organizations.
It's an interesting one though because obviously we have ad profiles in the west, but should an "adversary state" be allowed to collect and warehouse that data, knowing it's potentially a national security threat for personalized targeting of natsec security compromises/etc?
Should western companies be allowed to collect this data either? But they are the "good ones" and only use it for advertising (and letting the three-letter agencies buy it, of course). But what basis do you use to ban it? And is it really any safer to have it, when the chinese are pretty good at hacking these data warehouses, and the security is often incredibly sketchy? They got the plans to the F-35, the federal Office of Personnel Management database, the RSA key-secret database... if you store it, they'll ultimately get it eventually too.
What if instead of the classical "this user is a gambler and in financial trouble, have an agent call them" targeting you build a personal weight-model on what things trigger engagement for that specific user? What if they show you OANN until you're so enraged you go attack the capitol? What if they troll you on War Thunder forums until you leak classified documents? And you can potentially do that en-masse, use AI to generate a personalized "havoc profile" of what content they should surface at this exact point to make you the most socially disruptive at some key moment, or the things they most need an intel leak on, and the AI can use that profile to surface the content that most effectively winds you personally up.
I'd expect it's probably possible (not sure if it exists) to do engagement/stimulation (arousal) and sentiment analysis in realtime. So you can know that if they're usually scrolling X when they're upset, that you should push content Y to push their buttons most effectively. Like you can literally micro-target people at their weakest moments with the content that most effectively pushes their buttons to accomplish whatever goal.
Essentially - what if instead of "AI girlfriend" it was "AI best-friend with links to extremism", and he's gently radicalizing you a little bit every time you're scrolling on the shitter? That's the product for Tiktok.
And while most people are not going to get wound up to the extent of "attack the capitol/leak classified documents"... statistically like all advertising, it works, some people will stochastically be stimulated to action. And even at a more pedestrian level, it's very easy to ramp up polarization and internal tension, like Russia has been doing during elections/etc. But you can automate that shit at a scale and precision of targeting far beyond traditional 1:1-scale intelligence work.
Selling more diapers and kleenex is baby shit, Tiktok is the point of the spear on true infowars shit, where everyone can be probed for their personal hot-buttons and then press them automatically, en-masse while you're scrolling. And we just use that to sell more McDonald's. Using a nuke to kill a mosquito.
It's a weird time we live in, this stuff essentially parasitizes our cultural values of free speech and free information information exchange and turns it into an attack vector. That's been what advertising is all along (it's the art of convincing people to buy shit they don't need/make decisions they wouldn't have otherwise made) but we are starting to go past the "benign toxoplasmosis" to something that can be weaponized for political ends.
Anyway that's why the security apparatus is getting wound up about tiktok. You absolutely will not get a security clearance if you're anywhere near tiktok, you will not get tiktok anywhere near a government computer, and lately countries are just banning it outright because it's too dangerous.
And reddit sees that and says "hey, we can be Tiktok for westoids and fill that gap". There is competition of course - Snap and Insta are really in a better place to pivot to this. But Reddit has been pushing towards being a content-drip for a while, that's been the entire point of the "new-experience" and first-party app, and this is just them making the leap. And that means third-party clients have gotta go.