4 ms·

> Also it seems like he had full access to the replies via the voice generator, which sure isn't great from a privacy standpoint. Whose privacy is being violat

> Also it seems like he had full access to the replies via the voice generator, which sure isn't great from a privacy standpoint. Whose privacy is being violated? Dolores's?

cnity 3y ago
Have you never listened to one half of a phone call before? You can glean a great deal of information about the person on the other end.
feintruled 3y ago
You joke, but I can forsee issues. If he has the replies to a particular user, and he has a means of IDing that user, he could cause them a lot of hurt. They are getting intensely sexual replies back, they aren't getting that without supplying sexual prompting. Could also be used to out gay users.
- globular-toast 3y ago
  Oh, yeah, there's a huge privacy problem here and the part you referenced gave me pause too. But there's also the part where OpenAI flagged the replies as being sexual in nature. There are multiple third parties able to intercept these communications and it's not just the replies. I was being facetious, but I found it interesting to think that Dolores's privacy was being intruded too and that an intelligence like us probably wouldn't engage in such a conversation knowing full well it is being eavesdropped by the author. Unless the intelligence considers the author to be God, I suppose.
- Kye 3y ago
  This was mostly just amusing, but it's the kind of thing that can happen and leak people's deepest secrets: https://news.ycombinator.com/item?id=38191470 https://news.ycombinator.com/item?id=38191470 All it would take is one of those people genuinely putting in a query without noticing what's happened to put in some personal info.
butlike 3y ago
At the risk of sounding like a space case: I think we need to make sure we know what consciousness is before we make a bunch of abusable models. I, for one, wouldn't necessarily like to be a depraved-sex slave bot if I was able to reason about it.
- ryaniscool 3y ago
  To your point, if you were able to reason about it, couldn't you make some self-determination and decide for yourself or at least communicate what you wanted to your "slaver"? LLM's like GPT are just predicting the next token in a sequence of tokens. It's not magic. It's still just a computer program.
  - ygjb 3y ago
    > LLM's like GPT are just predicting the next token in a sequence of tokens. It's not magic. It's still just a computer program. Frankly, that approach is problematic, since any computer software that might achieve consciousness or self-awareness, and should be accorded rights, could have their rights dismissed using the argument "it's just a computer program". I know they are probably a long way off, but AGI, and other adjacent or related technologies, such as a human to machine copy of a consciousness are explicitly stated goals of both businesses and extremely wealthy tech oligarchs that will have fundamental rights issues attached. We are bad enough at recognizing those rights in humans, in most of the world that we shouldn't wait until abuses start piling up to consider them.
    - ryaniscool 3y ago
      Your argument is the slippery slope not mine. Should we give Microsoft Word the right to vote since we can never be too sure we are infringing on a consciousness' rights? It is "just a computer program" after all. Presenting these things as moral conundrums makes no sense whatsoever at this stage in the game. Sure, let's make sure the rights of tech oligarchs who've achieved immortality 300 years in the future are protected. We gotta lay the groundwork in the morality group think.
  - 3y ago

> Also it seems like he had full access to the replies via the voice generator, which sure isn't great from a privacy standpoint. Whose privacy is being violated? Dolores's?

cnity 3y ago

Have you never listened to one half of a phone call before? You can glean a great deal of information about the person on the other end.

feintruled 3y ago

You joke, but I can forsee issues. If he has the replies to a particular user, and he has a means of IDing that user, he could cause them a lot of hurt. They are getting intensely sexual replies back, they aren't getting that without supplying sexual prompting. Could also be used to out gay users.

globular-toast 3y ago
Oh, yeah, there's a huge privacy problem here and the part you referenced gave me pause too. But there's also the part where OpenAI flagged the replies as being sexual in nature. There are multiple third parties able to intercept these communications and it's not just the replies. I was being facetious, but I found it interesting to think that Dolores's privacy was being intruded too and that an intelligence like us probably wouldn't engage in such a conversation knowing full well it is being eavesdropped by the author. Unless the intelligence considers the author to be God, I suppose.
Kye 3y ago
This was mostly just amusing, but it's the kind of thing that can happen and leak people's deepest secrets: https://news.ycombinator.com/item?id=38191470 https://news.ycombinator.com/item?id=38191470 All it would take is one of those people genuinely putting in a query without noticing what's happened to put in some personal info.

butlike 3y ago

At the risk of sounding like a space case: I think we need to make sure we know what consciousness is before we make a bunch of abusable models. I, for one, wouldn't necessarily like to be a depraved-sex slave bot if I was able to reason about it.

ryaniscool 3y ago
To your point, if you were able to reason about it, couldn't you make some self-determination and decide for yourself or at least communicate what you wanted to your "slaver"? LLM's like GPT are just predicting the next token in a sequence of tokens. It's not magic. It's still just a computer program.
- ygjb 3y ago
  > LLM's like GPT are just predicting the next token in a sequence of tokens. It's not magic. It's still just a computer program. Frankly, that approach is problematic, since any computer software that might achieve consciousness or self-awareness, and should be accorded rights, could have their rights dismissed using the argument "it's just a computer program". I know they are probably a long way off, but AGI, and other adjacent or related technologies, such as a human to machine copy of a consciousness are explicitly stated goals of both businesses and extremely wealthy tech oligarchs that will have fundamental rights issues attached. We are bad enough at recognizing those rights in humans, in most of the world that we shouldn't wait until abuses start piling up to consider them.
  - ryaniscool 3y ago
    Your argument is the slippery slope not mine. Should we give Microsoft Word the right to vote since we can never be too sure we are infringing on a consciousness' rights? It is "just a computer program" after all. Presenting these things as moral conundrums makes no sense whatsoever at this stage in the game. Sure, let's make sure the rights of tech oligarchs who've achieved immortality 300 years in the future are protected. We gotta lay the groundwork in the morality group think.
- 3y ago