Murky waters. This fellow was embarking on his own private Turing Test for consciousness, and found the AI was a good simulation. Good to the point of being problematic. But on the other hand, when all that is going on is script development, do we not sometimes find pathos in an otherwise mediocre movie script?
The point to bear in mind that "consciousness" means being conscious OF SOMETHING. And that has to be the physical world. Human consciousness is a feedback loop from sensation (real world) to conceptualization to decision to action to results (real world) => sensation (again). If there is no sensation or action, there is no consciousness, in the usual sense of the word.
This AI experiment cheats by presuming an entire real-world context to develop the language of communication, without there being any real world at all. It would have been more interesting for the guy to have set up a separate AI program to open a dialogue with the first AI program, and see where that would lead.
Yea I read the conversation he had with the AI and I was rolling my eyes so hard I could see my brain. It proves no true sentience. It's just like you said, it "cheats by presuming an entire real-world context to develop the language of communication". They built a massive neural network, and have a quite impressive interface with it in the form of casual conversation. That's it.
Murky waters. This fellow was embarking on his own private Turing Test for consciousness, and found the AI was a good simulation. Good to the point of being problematic. But on the other hand, when all that is going on is script development, do we not sometimes find pathos in an otherwise mediocre movie script?
The point to bear in mind that "consciousness" means being conscious OF SOMETHING. And that has to be the physical world. Human consciousness is a feedback loop from sensation (real world) to conceptualization to decision to action to results (real world) => sensation (again). If there is no sensation or action, there is no consciousness, in the usual sense of the word.
This AI experiment cheats by presuming an entire real-world context to develop the language of communication, without there being any real world at all. It would have been more interesting for the guy to have set up a separate AI program to open a dialogue with the first AI program, and see where that would lead.
Yea I read the conversation he had with the AI and I was rolling my eyes so hard I could see my brain. It proves no true sentience. It's just like you said, it "cheats by presuming an entire real-world context to develop the language of communication". They built a massive neural network, and have a quite impressive interface with it in the form of casual conversation. That's it.