3 ms·

I’ll note that Google itself originally started where it was mostly serving its users. Many social media sites were like that. The Xbox Home Screen used to be l

I’ll note that Google itself originally started where it was mostly serving its users. Many social media sites were like that. The Xbox Home Screen used to be like that. They captured a huge market, figured out how to integrate ads, and then forced them on their users. AI made or used by ad companies might follow a similar path. I think they’ll mostly try to show you ads outside the AI itself, though. Maybe on the page or in the app where you interact with the AI. Or it stays as bait for their ad-driven services.

kazinator 3y ago
It will be interesting to see how AI gets deliberately enshittified, and whether those strategies will work. I have a feeling (and I hope I'm right) that people will not tolerate bullshit AI interactions as much as they tolerated crap web search results. My reasoning is due to these differences. - Under web searching, the Ad results (sponsored links) are delineated. You can filter those out or skip them, and you know there is real search content after that. - In the reach search content, there is spam. But there is a rhetoric that at least used to be plausible that the spam is not the search engine's fault. The search engine indexing and ranking follow some algorithm that is not highly intelligent, and is being gamed by the SEO bad guys. Not only did people make that excuse, but they knew that if they sift through the long tail of the search result, they may be able to find what they are looking for; i.e. that it's there, just outranked. I suspect, people won't be willing to apologetic excuses on behalf of AI. If it's bullshitting them, they will end the conversation and go to another AI. Google built an advantage in the search space through massive infrastructure. In AI, it's possible to wield market power through hardware also. All else being equal, people will go for the free, highly available, fast AI, compared to slow, paid, downtime-riddled AI. There will probably be a battle among AIs for integration into applications and services. That's where a particular AI will be able to get a stranglehold on the users: AI integrated into apps for which there are no easy substitutes, and where you can't choose another AI.

kazinator 3y ago

It will be interesting to see how AI gets deliberately enshittified, and whether those strategies will work. I have a feeling (and I hope I'm right) that people will not tolerate bullshit AI interactions as much as they tolerated crap web search results. My reasoning is due to these differences. - Under web searching, the Ad results (sponsored links) are delineated. You can filter those out or skip them, and you know there is real search content after that. - In the reach search content, there is spam. But there is a rhetoric that at least used to be plausible that the spam is not the search engine's fault. The search engine indexing and ranking follow some algorithm that is not highly intelligent, and is being gamed by the SEO bad guys. Not only did people make that excuse, but they knew that if they sift through the long tail of the search result, they may be able to find what they are looking for; i.e. that it's there, just outranked. I suspect, people won't be willing to apologetic excuses on behalf of AI. If it's bullshitting them, they will end the conversation and go to another AI. Google built an advantage in the search space through massive infrastructure. In AI, it's possible to wield market power through hardware also. All else being equal, people will go for the free, highly available, fast AI, compared to slow, paid, downtime-riddled AI. There will probably be a battle among AIs for integration into applications and services. That's where a particular AI will be able to get a stranglehold on the users: AI integrated into apps for which there are no easy substitutes, and where you can't choose another AI.