5 ms·

Wasting news cycles on the question "is Bing Chat sentient" is pointless, because the answer is completely irrelevant. Let's agree for a second Bing Chat is se

Wasting news cycles on the question "is Bing Chat sentient" is pointless, because the answer is completely irrelevant. Let's agree for a second Bing Chat is sentient. Ok so what? Now let's agree that it's not sentient. Again, so what? What is the real-world consequence of agreeing one way or the other? It's like asking "Is Bing Chat more like the colour blue or like the color green?" with people arguing for both sides. Again, no matter what color we ascribe to Bing Chat is completely inconsequential. We should rather spend time making sure that the AI of the future will be democratized and not concentrated in the hands of the few, be it governments, multinational corporation, terrorists or whatever. But for this, "Open"AI would actually have to become true to their name.

mort96 4y ago
Well, the reason it's relevant is that we want to treat sentient beings well. The question of "are these chatbots sentient or not?" is the question, "should we consider these chatbots as entities which are worthy of some kind of moral consideration or should we treat them as just another computer program?" EDIT: Note that I'm not arguing in favor of it being sentient here, I'm not confident enough in my understanding of sentience to say. I'm just saying why it would be significant if it was.
mmahemoff 4y ago
I agree it's facile to argue about dictionary definitions. There's several distinct topics people are thinking about when they talk about AI sentience and it's more productive to address them directly. Namely: (a) practical implications for what AI enables and how it will impact the economy and our lives (b) safety concerns about how the tech could do damage, even posing an existential threat (c) ethical concerns about machine rights. Philosophers can sit around and debate sentience all day, but it doesn't answer those practical questions. It's usually circular reasoning arguments about definitions and, worse, those definitions are engineered to drift so they're always out of reach of the latest AI advances ("shifting the goalposts").

mort96 4y ago

Well, the reason it's relevant is that we want to treat sentient beings well. The question of "are these chatbots sentient or not?" is the question, "should we consider these chatbots as entities which are worthy of some kind of moral consideration or should we treat them as just another computer program?" EDIT: Note that I'm not arguing in favor of it being sentient here, I'm not confident enough in my understanding of sentience to say. I'm just saying why it would be significant if it was.

mmahemoff 4y ago

I agree it's facile to argue about dictionary definitions. There's several distinct topics people are thinking about when they talk about AI sentience and it's more productive to address them directly. Namely: (a) practical implications for what AI enables and how it will impact the economy and our lives (b) safety concerns about how the tech could do damage, even posing an existential threat (c) ethical concerns about machine rights. Philosophers can sit around and debate sentience all day, but it doesn't answer those practical questions. It's usually circular reasoning arguments about definitions and, worse, those definitions are engineered to drift so they're always out of reach of the latest AI advances ("shifting the goalposts").