4 ms·

> We’ve also introduced additional content safeguards for this experience, such as blocking prompts and generations in a wider range of categories. Oh good, ju

> We’ve also introduced additional content safeguards for this experience, such as blocking prompts and generations in a wider range of categories. Oh good, just what wasn't needed.

cft 3y ago
Is this for no-account users, or for everyone, including paid users? In the past, I thought that search engines censored content because of the advertiser's demands, but now that AI search switched to the paid model I think the censorship is purely ideological because these companies are based in San Francisco.
- ethanbond 3y ago
  So if you picked up the building and put it elsewhere it'd produce the same product but different politics?
- gbear605 3y ago
  The censorship is about not getting sued or regulated
- newaccount74 3y ago
  I think search engines block content because it is what users want. In the early internet, you'd almost always end up with a few pictures of naked girls in the search results, and many pages showed ads for porn sites. That's just not a good experience if you are at work and trying to get stuff done. Search engines have gotten much better, and in most cases explicit results are now only shown when you actively look for them.
Spivak 3y ago
I get that this sucks but what's the alternative? It would take legislation that will never pass to eliminate OpenAI's liability for the content that their model spits out and they're already under a microscope by regulators and the public.
- not2b 3y ago
  This doesn't have anything to do with their liability or regulations, it's what their paying customers want. A company that pays them to provide a chatbot for their business doesn't want the chatbot to say controversial things or tell people to commit suicide.
  - bilbo0s 3y ago
    Actually if it's about customers not paying them if they spit out "kill yourself" content, then I 100% understand that. These are businesses. Those servers cost money. That compute costs even more. The AI experts, (real AI experts by the way, not tensorflow and pytorch monkeys), cost big money as well. Someone better be paying for all that. So if the people who want offensive content are willing to pony up the dollars, then great. I've got no problem with giving them what they want. But if they want Barbie or Cocomelon to pay for the offensive content, then yeah, they can be safely ignored. Block as much as you like. (Or rather, block as much as Barbie wants blocked. Which is probably more than you like, but they're paying you handsomely for it.)
    - Y_Y 3y ago
      I may just be a pytorch monkey, but I cost more than a training server.
- nextaccountic 3y ago
  The alternative is running models locally
  - consumer451 3y ago
    This would be ideal, but what percentage of the world's population can afford the hardware required for that?
- francescovv 3y ago
  Alternative would be to have model itself aware of sensitive topics and and meaningfully engage with questions touching such topics with that awareness. It might be a while till that is feasible, though. Until then, "content safeguards" will continue to feel like overreaching, artificial stonewalls scattered across otherwise kinda-consistent space.

> We’ve also introduced additional content safeguards for this experience, such as blocking prompts and generations in a wider range of categories. Oh good, just what wasn't needed.

cft 3y ago

Is this for no-account users, or for everyone, including paid users? In the past, I thought that search engines censored content because of the advertiser's demands, but now that AI search switched to the paid model I think the censorship is purely ideological because these companies are based in San Francisco.

ethanbond 3y ago
So if you picked up the building and put it elsewhere it'd produce the same product but different politics?
gbear605 3y ago
The censorship is about not getting sued or regulated
newaccount74 3y ago
I think search engines block content because it is what users want. In the early internet, you'd almost always end up with a few pictures of naked girls in the search results, and many pages showed ads for porn sites. That's just not a good experience if you are at work and trying to get stuff done. Search engines have gotten much better, and in most cases explicit results are now only shown when you actively look for them.

Spivak 3y ago

I get that this sucks but what's the alternative? It would take legislation that will never pass to eliminate OpenAI's liability for the content that their model spits out and they're already under a microscope by regulators and the public.

not2b 3y ago
This doesn't have anything to do with their liability or regulations, it's what their paying customers want. A company that pays them to provide a chatbot for their business doesn't want the chatbot to say controversial things or tell people to commit suicide.
- bilbo0s 3y ago
  Actually if it's about customers not paying them if they spit out "kill yourself" content, then I 100% understand that. These are businesses. Those servers cost money. That compute costs even more. The AI experts, (real AI experts by the way, not tensorflow and pytorch monkeys), cost big money as well. Someone better be paying for all that. So if the people who want offensive content are willing to pony up the dollars, then great. I've got no problem with giving them what they want. But if they want Barbie or Cocomelon to pay for the offensive content, then yeah, they can be safely ignored. Block as much as you like. (Or rather, block as much as Barbie wants blocked. Which is probably more than you like, but they're paying you handsomely for it.)
  - Y_Y 3y ago
    I may just be a pytorch monkey, but I cost more than a training server.
nextaccountic 3y ago
The alternative is running models locally
- consumer451 3y ago
  This would be ideal, but what percentage of the world's population can afford the hardware required for that?
francescovv 3y ago
Alternative would be to have model itself aware of sensitive topics and and meaningfully engage with questions touching such topics with that awareness. It might be a while till that is feasible, though. Until then, "content safeguards" will continue to feel like overreaching, artificial stonewalls scattered across otherwise kinda-consistent space.