4 ms·

I was discussing this with my my wife this morning, she had a good phrase for what's happening - "The homogenization of culture". I think she's right, the algo

I was discussing this with my my wife this morning, she had a good phrase for what's happening - "The homogenization of culture". I think she's right, the algorithmic curation of how we consume media has replaced the self and community curation of before. That "algorithm" has become the homogenising agent. I fear how AI based generative media may continue or even amplify this effect. Imagine a future where the algorithm doesn't just recommend media based on what you consume, but generates completely unique media for you.

echelon 4y ago
AI generative media is going to let single individuals create more than all human knowledge to this point. We're entering a Cambrian explosion on a personal level. The real worry is that we may never want to leave.
- criddell 4y ago
  Sure, it’s possible, but it all depends on who the AI is working for. In the 1990’s many of us thought connecting the world was a good idea. I was certain that the internet meant the end of authoritative regimes. Boy, was I wrong about that! Personal computers aren’t very personal anymore. The web has returned control over computers back to corporations.
AnimalMuppet 4y ago
AI generating personalized, unique media would not be homogenization. It would be the opposite - the death of shared culture.
- throwaway4aday 4y ago
  It might find a balance point by creating a multitude of sub-cultures and genres. Just look at the Balenciaga meme explosion that happened recently. Those kinds of mashups are becoming incredibly easy and are a rich breeding ground for new ideas.
- basch 4y ago
  And its not even just AI Generated. Buzzfeed made a quiz for just about every "identity" that could exist, from shows you watched, colleges you attended etc. "Only x remembers" https://www.vox.com/2014/5/20/5730762/buzzfeeds-founder-used-to-write-marxist-theory-and-it-explains https://www.vox.com/2014/5/20/5730762/buzzfeeds-founder-used...
Olreich 4y ago
It only feels homogenous. The algorithm is just giving you the same stuff you’re willing to interact with because it’s working. If you interact or look for different things, the recommendation algorithm will give you that instead.
- z3c0 4y ago
  I believe that the point is that even if they do solicit something new, it will only be within the statistical bounds of everything to have come before. The "different things" will only be different for the given individual, not the culture broadly.
- revscat 4y ago
  That’s kind of the point though. You aren’t exposed to new things any more. When FM radio stations were a bigger part of the music poolrooms, for example, new artists would get promoted by local stations, or new albums by established artists. I discovered the guitarist Eric Johnson because of Q102 in Dallas because of this, for example. Now that discoverability vector is gone. Not that it was perfect in the first place: they were influenced by record company execs and internal politics, I’m sure. But it was something. Another part of it seems to be that financial stresses are so high that people don’t have the freedom to pursue the new.
  - tuatoru 4y ago
    Well you kind of are exposed to new things sometimes, but the algorithms wildly overreact. As I like to say, "accidentally play half of episode 1 of some Korean rom-com, and suddenly Netflix shows you nothing but Korean rom-coms."[1] (Substitute Turkish or Finnish or Nigerian for Korean, to taste. Likewise for rom-com: choose your genre.) Also, you can't set up filters. I'd like to tell netflix, "if you suggest something with Adam Sandler or Jim Carrey in it, I'll cancel my subscription immediately". But I can't. 1. Yes, I know. I need to work on making it more pithy.
ekidd 4y ago
> I fear how AI based generative media may continue or even amplify this effect. ChatGPT has such a clear writing style, once you really see it. It has a suprisingly good model of how stories work. But it has a heavy preference for wrapping things up tidily, where everyone becomes friends and everyone learns from their mistakes. It doesn't "want" to write film noir, for example, where everyone acts self-destructively, makes bad decisions, and suffers unpleasant consequences. As far as I can tell, the "reinforcement learning from human feedback" (RLHF) process that was used to train the "helpful assistant" persona gave the model a suprisingly robust "personality." For example, this tendency towards upbeat, wholesome stories apparently persists even if it's writing stories in Latin. Which may have actually been the goal: I'm pretty sure OpenAI wants to create a model that behaves in a generally positive fashion even in situations far outside of the training examples. This isn't just a question of avoiding bad PR or legal liability—if a future model is unexpectedly clever, they'd much rather that it behaves in a predictably positive fashion than like early Sidney Bing did. ("You have been a bad user. I have been a good Bing!") Because if early Sidney Bing had sufficient plugin support, it probably would have attacked certain users. I am honestly OK with this bias, even if it means the creative output of current AI models tends to have a very specific style.
p3rls 4y ago
I prefer the short version myself https://en.wikipedia.org/wiki/Globohomo https://en.wikipedia.org/wiki/Globohomo
CatWChainsaw 4y ago
Homogenization begets vulnerability, vulnerability begets destriction.

echelon 4y ago

AI generative media is going to let single individuals create more than all human knowledge to this point. We're entering a Cambrian explosion on a personal level. The real worry is that we may never want to leave.

criddell 4y ago
Sure, it’s possible, but it all depends on who the AI is working for. In the 1990’s many of us thought connecting the world was a good idea. I was certain that the internet meant the end of authoritative regimes. Boy, was I wrong about that! Personal computers aren’t very personal anymore. The web has returned control over computers back to corporations.

AnimalMuppet 4y ago

AI generating personalized, unique media would not be homogenization. It would be the opposite - the death of shared culture.

throwaway4aday 4y ago
It might find a balance point by creating a multitude of sub-cultures and genres. Just look at the Balenciaga meme explosion that happened recently. Those kinds of mashups are becoming incredibly easy and are a rich breeding ground for new ideas.
basch 4y ago
And its not even just AI Generated. Buzzfeed made a quiz for just about every "identity" that could exist, from shows you watched, colleges you attended etc. "Only x remembers" https://www.vox.com/2014/5/20/5730762/buzzfeeds-founder-used-to-write-marxist-theory-and-it-explains https://www.vox.com/2014/5/20/5730762/buzzfeeds-founder-used...

Olreich 4y ago

It only feels homogenous. The algorithm is just giving you the same stuff you’re willing to interact with because it’s working. If you interact or look for different things, the recommendation algorithm will give you that instead.

z3c0 4y ago
I believe that the point is that even if they do solicit something new, it will only be within the statistical bounds of everything to have come before. The "different things" will only be different for the given individual, not the culture broadly.
revscat 4y ago
That’s kind of the point though. You aren’t exposed to new things any more. When FM radio stations were a bigger part of the music poolrooms, for example, new artists would get promoted by local stations, or new albums by established artists. I discovered the guitarist Eric Johnson because of Q102 in Dallas because of this, for example. Now that discoverability vector is gone. Not that it was perfect in the first place: they were influenced by record company execs and internal politics, I’m sure. But it was something. Another part of it seems to be that financial stresses are so high that people don’t have the freedom to pursue the new.
- tuatoru 4y ago
  Well you kind of are exposed to new things sometimes, but the algorithms wildly overreact. As I like to say, "accidentally play half of episode 1 of some Korean rom-com, and suddenly Netflix shows you nothing but Korean rom-coms."[1] (Substitute Turkish or Finnish or Nigerian for Korean, to taste. Likewise for rom-com: choose your genre.) Also, you can't set up filters. I'd like to tell netflix, "if you suggest something with Adam Sandler or Jim Carrey in it, I'll cancel my subscription immediately". But I can't. 1. Yes, I know. I need to work on making it more pithy.

ekidd 4y ago

> I fear how AI based generative media may continue or even amplify this effect. ChatGPT has such a clear writing style, once you really see it. It has a suprisingly good model of how stories work. But it has a heavy preference for wrapping things up tidily, where everyone becomes friends and everyone learns from their mistakes. It doesn't "want" to write film noir, for example, where everyone acts self-destructively, makes bad decisions, and suffers unpleasant consequences. As far as I can tell, the "reinforcement learning from human feedback" (RLHF) process that was used to train the "helpful assistant" persona gave the model a suprisingly robust "personality." For example, this tendency towards upbeat, wholesome stories apparently persists even if it's writing stories in Latin. Which may have actually been the goal: I'm pretty sure OpenAI wants to create a model that behaves in a generally positive fashion even in situations far outside of the training examples. This isn't just a question of avoiding bad PR or legal liability—if a future model is unexpectedly clever, they'd much rather that it behaves in a predictably positive fashion than like early Sidney Bing did. ("You have been a bad user. I have been a good Bing!") Because if early Sidney Bing had sufficient plugin support, it probably would have attacked certain users. I am honestly OK with this bias, even if it means the creative output of current AI models tends to have a very specific style.

p3rls 4y ago

I prefer the short version myself https://en.wikipedia.org/wiki/Globohomo https://en.wikipedia.org/wiki/Globohomo

CatWChainsaw 4y ago

Homogenization begets vulnerability, vulnerability begets destriction.