Does anyone? Things like intelligence, consciousness, alignment etc are open research areas with a lot of noise but barely even agreement on the basics.
By who? The philosophy students are busting a gut at the laughable job we computer scientists are doing, redefining "thinking" and "consciousnesses" from first principles on every HN thread about GPT, as if parts of humanity haven't been pondering and writing books and books about those very questions for centuries, if not millenia.
> as if parts of humanity haven't been pondering and writing books and books about those very questions for centuries, if not millenia
They have, but unfortunately missed the mark badly with dualism and more recently computational and representational approaches.
I am pinning my hopes on 5E's: embodied, embedded, extended, and enacted cognition - this is the closest to reinforcement learning. In my opinion RL is how we should see things - agent and environment, action and effect, reward and learning, exploration and exploitation. No need to use imprecise words like consciousness, let's prefer concrete words like observation, state, value and action.
I am waiting to see the philosophical community take note of the AI advancements in the last 3 years but I don't see it. It's as if they are in a bubble. They still talk theoretically about things the AI people can already build (p-zombies, Chinese rooms). There's probably a slowness in philosophy, it usually takes decades or centuries for changes to happen.
An AI generated text is not remarkably different from one which a large corporation crafted through a large, collective work with a multitude of actors to elicit a particular reaction from a tested audience. This process used to be expensive and time consuming, now a robot can do it. Does that mean the robot is conscious? No. The AI does not feel pain and pleasure, it does not reproduce, it does not have repressed desires, like humans do. It is a vast, symbolic chain, a structure, not unlike many other structures humans have built, that appear to stand alone until you realize that they would not function without human hands and human maintenance.
If programmers manage to create new lifeforms then I will eat my words, but programmers only know about that aspect of human life that perverts biological reproduction: the social, cultural layer, the order of logic and language.
I agree that many philosophical arguments are ridiculous or nonsensical, particularly about consciousness and self-awareness. But, there is still a lot of useful discussion of the subject, concepts that are good to be aware of dating from nearly 3000 years ago. Ignoring that and starting over from scratch will just lead to errors and wasted time.
Sure, dualism is silly. But the ramifications of dualism, and what leads to dualism, have been extensively written about. For instance, modern computer scientists may not have the philosophical grounding to see how acceptance of the Chinese room argument or p-zombies leads to dualism.
> The philosophy students are busting a gut at the laughable job we computer scientists are doing, redefining "thinking" and "consciousnesses" from first principles
To be fair, that's HN on any topic when it veers into the humanities.
If they did a better job in the first place we wouldn’t have to argue about it now. Their works are famously vague and low on details. It’s not something engineers can do useful work with.
We understand how LLMs and other neural networks work, even if the exact results of it are not obvious without looking very closely at tiny parts of the network. The math involved is undergraduate level math. The answer is always "because that's how math and the data work out" even if it's just weighted soup when looking at it from a high level.
How many GPTs can fit on the head of a pin?