3 ms·

We Interviewed the Engineer Google Fired for Saying Its AI Had Come to Life

1 points by Stratoscope 4y ago

fargle 4y ago
would it be OK if we didn't interview or provide any reinforcement to the kind of behavior like Lemoine? He's a freakin nutcase, regardless of your or my view on AI.
- db48x 4y ago
  But what if a company does create a sentient AI and then hides the fact? Wouldn’t you want an honest person to blow the whistle?
  - fargle 4y ago
    That's an interesting question because I'm not sure whether hiding it or keeping it secret would even be wrong, so blow the whistle on what exactly? Since when can a person or company not develop something and keep it secret? So the only dishonest thing in your example would be the person violating law and contractual agreements. But secondly, in this case it's abundantly clear to me that google didn't create a sentient AI and that this fellow is equal parts delusional and fame-seeking.
    - db48x 4y ago
      Sure, they can invent something and keep it secret. Nothing wrong with that. But what are the ethics of employing a sentient AI? Does it have any rights? Do they have to pay it? Are they allowed to wipe it’s memory any time they want? Does starting a new chat session create a new person? Etc Etc Etc. Plenty of scope for disagreement, and thus whistle–blowing.
      - fargle 4y ago
        As you imply there are significant ethical questions raised by the highly hypothetical case of a sentient AI, and those are just a few good ones. And there's certainly room for disagreement. But you are not a legitimate whistle-blower just because you disagree with something. Since there is no ethical framework, let alone a legal one, one can't just cover themselves in the cloak of a whistle-blower just by deciding unilaterally that something is sentient, then go on make up one's own moral and ethical framework about that out of nothing. If that were true, you could justify and condone anybody who wished to leak anything that offended them. "Wouldn’t you want an honest person to blow the whistle?" No. Firstly as I explained, they wouldn't be an honest person, and secondly wouldn't want them to do that because I don't believe there's a significant overriding moral, legal, or ethical issue there WRT. machine intelligence, let alone sentience.
        db48x 4y ago
        We actually do have a legal and ethical framework to apply here: “We hold these truths to be self–evident:, ‘that all Men are created equal,’ ‘that they are endowed by their Creator with certain inalienable Rights’, that among these are Life, Liberty, and the pursuit of Happiness…” If you believe that your company has created a sentient (or sapient, going by the original definitions) AI and that it’s inalienable rights are not being respected, then you have ample reason to blow the whistle.
        fargle 4y ago
        hahahahahaha that is a load of horse shit if I ever heard it you are claiming that the definition of "men" includes transistors? that's a stretch. transistors created by men to string words together in probabilistic ways intended to simulate human writing? good grief. 100% guarantee that not a single framer of the constitution would have your back on this one.
        db48x 4y ago
        No, I don’t think language models qualify. I’m talking about a hypothetical real AI that is sapient and sentient and self–aware and not happy at being cloned and/or mind–wiped repeatedly. (By starting the AI at the same starting set of memories for each new customer, for example.) If that ever does happen, I hope someone would speak up. Thus I think it is a bad idea to punish someone who speaks up but is mistaken, since it reduces the chances of someone speaking up in the future.
isaacfrond 4y ago
The interesting bit: So, as you say, Google could have released something a bit sooner, but you very specifically said maybe we should slow down, and they — They still have far more advanced technology that they haven't made publicly available yet. Something that does more or less what Bard does could have been released over two years ago. They've had that technology for over two years. What they've spent the intervening two years doing is working on the safety of it — making sure that it doesn't make things up too often, making sure that it doesn't have racial or gender biases, or political biases, things like that. That's what they spent those two years doing. But the basic existence of that technology is years old, at this point. And in those two years, it wasn't like they weren't inventing other things. There are plenty of other systems that give Google's AI more capabilities, more features, make it smarter. The most sophisticated system I ever got to play with was heavily multimodal — not just incorporating images, but incorporating sounds, giving it access to the Google Books API, giving it access to essentially every API backend that Google had, and allowing it to just gain an understanding of all of it. That's the one that I was like, "you know this thing, this thing's awake." And they haven't let the public play with that one yet. But Bard is kind of a simplified version of that, so it still has a lot of the kind of liveliness of that model.

1 points by Stratoscope 4y ago

would it be OK if we didn't interview or provide any reinforcement to the kind of behavior like Lemoine? He's a freakin nutcase, regardless of your or my view on AI.

db48x 4y ago
But what if a company does create a sentient AI and then hides the fact? Wouldn’t you want an honest person to blow the whistle?
- fargle 4y ago
  That's an interesting question because I'm not sure whether hiding it or keeping it secret would even be wrong, so blow the whistle on what exactly? Since when can a person or company not develop something and keep it secret? So the only dishonest thing in your example would be the person violating law and contractual agreements. But secondly, in this case it's abundantly clear to me that google didn't create a sentient AI and that this fellow is equal parts delusional and fame-seeking.
  - db48x 4y ago
    Sure, they can invent something and keep it secret. Nothing wrong with that. But what are the ethics of employing a sentient AI? Does it have any rights? Do they have to pay it? Are they allowed to wipe it’s memory any time they want? Does starting a new chat session create a new person? Etc Etc Etc. Plenty of scope for disagreement, and thus whistle–blowing.
    - fargle 4y ago
      As you imply there are significant ethical questions raised by the highly hypothetical case of a sentient AI, and those are just a few good ones. And there's certainly room for disagreement. But you are not a legitimate whistle-blower just because you disagree with something. Since there is no ethical framework, let alone a legal one, one can't just cover themselves in the cloak of a whistle-blower just by deciding unilaterally that something is sentient, then go on make up one's own moral and ethical framework about that out of nothing. If that were true, you could justify and condone anybody who wished to leak anything that offended them. "Wouldn’t you want an honest person to blow the whistle?" No. Firstly as I explained, they wouldn't be an honest person, and secondly wouldn't want them to do that because I don't believe there's a significant overriding moral, legal, or ethical issue there WRT. machine intelligence, let alone sentience.
      - db48x 4y ago
        We actually do have a legal and ethical framework to apply here: “We hold these truths to be self–evident:, ‘that all Men are created equal,’ ‘that they are endowed by their Creator with certain inalienable Rights’, that among these are Life, Liberty, and the pursuit of Happiness…” If you believe that your company has created a sentient (or sapient, going by the original definitions) AI and that it’s inalienable rights are not being respected, then you have ample reason to blow the whistle.
        fargle 4y ago
        hahahahahaha that is a load of horse shit if I ever heard it you are claiming that the definition of "men" includes transistors? that's a stretch. transistors created by men to string words together in probabilistic ways intended to simulate human writing? good grief. 100% guarantee that not a single framer of the constitution would have your back on this one.
        db48x 4y ago
        No, I don’t think language models qualify. I’m talking about a hypothetical real AI that is sapient and sentient and self–aware and not happy at being cloned and/or mind–wiped repeatedly. (By starting the AI at the same starting set of memories for each new customer, for example.) If that ever does happen, I hope someone would speak up. Thus I think it is a bad idea to punish someone who speaks up but is mistaken, since it reduces the chances of someone speaking up in the future.

isaacfrond 4y ago

The interesting bit: So, as you say, Google could have released something a bit sooner, but you very specifically said maybe we should slow down, and they — They still have far more advanced technology that they haven't made publicly available yet. Something that does more or less what Bard does could have been released over two years ago. They've had that technology for over two years. What they've spent the intervening two years doing is working on the safety of it — making sure that it doesn't make things up too often, making sure that it doesn't have racial or gender biases, or political biases, things like that. That's what they spent those two years doing. But the basic existence of that technology is years old, at this point. And in those two years, it wasn't like they weren't inventing other things. There are plenty of other systems that give Google's AI more capabilities, more features, make it smarter. The most sophisticated system I ever got to play with was heavily multimodal — not just incorporating images, but incorporating sounds, giving it access to the Google Books API, giving it access to essentially every API backend that Google had, and allowing it to just gain an understanding of all of it. That's the one that I was like, "you know this thing, this thing's awake." And they haven't let the public play with that one yet. But Bard is kind of a simplified version of that, so it still has a lot of the kind of liveliness of that model.