- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
Let me tell you about a Microsoft AI researcher […] who recently spent quite a lot of time considering whether the legendary Microsoft real time strategy game Age of Empires II is conscious, and built a basic neural network within the video game using digital goats to prove his point. […]
De Wynter built an LLM within AoEII using goats. “The point of the paper is to formally show that we anthropomorphise too readily, and that sometimes the claims we make with regards to LLM capabilities are too strong,” he told 404 Media. “It’s not an easy task, given that ‘human-like attributes’ is a bit of an abstract term.”



Wait, really? Can you give several examples of this, then? Sentience involves an organized, (hopefully at least sometimes) self-aware response to one’s environment. For example, the chance for something to become sentient is never increased by a lack of stimuli, right?
A person suffering from pseudocoma or “locked-in syndrome” is as close a real-world approximation I can think of. With that as a starting point we can arrive at my conclusion via thought experiment and create as many examples as you’d like.
Think of a brain in a vat. Maybe it’s your brain. You’re awake but you have no sensory input and, to your knowledge, no way of asserting your existence to the outside world.
It’s a hellish thought. It demonstrates the importance of recognizing that one self’s connection to the external is not equivelant with the self though.
Now digitize that mind. We now have an artificial intelligence. Disembodied. No sense organs to speak of. A thinking mind alone in a void.
Some variations on the Chinese Room demonstrate how thick a quagmire this discussion can be
Agreed! 😄
My point is not that an organism might find an evolutionary niche which requires no sense organs and yet develop sentience (though the universe is so vast that improbable is not impossible).
My point is that our measures and models cannot be the thing itself. Reponse to stimuli as you describe are mere measures we use to categorize minds based on our current models. Even if we believe our means are valid and useful and provide access to some level of truth: we cannot reasonably claim they allow total surety.