4 ms·
"No animals have all the attributes of human minds; but almost all the attributes of human minds are found in some animal or other" This was a fun read in gene
by _dps 9y ago
"No animals have all the attributes of human minds; but almost all the attributes of human minds are found in some animal or other"
This was a fun read in general, but I think it's particularly interesting to consider some of its implications in the context of trends in AI (to be clear, the article has nothing to do with AI).
I think the excerpt above is rapidly becoming true for current generation ML/AI systems (I'm leaving aside the question of qualia [0], since we are still unclear how to evaluate that question even for animals).
To the extent the above becomes true, I think it's an interesting contrast to the implied point behind Dijkstra's famous "swimming submarine" analogy:
"The question of whether machines can think... is about as relevant as the question of whether submarines can swim."
The submarine "swims", but almost none of that activity could be mapped onto how a human swims. As the structure of ML/AI systems starts to mimic what we believe we know about the neuroscience of the brain, the question of whether machines are "thinking", and what that might mean about thought in general, is very interesting.
[0] https://en.wikipedia.org/wiki/Qualia https://en.wikipedia.org/wiki/Qualia
- StavrosK 9y agoThe question, to me, is whether machines will ever have consciousness, or whether consciousness is just "what it feels like to have a cortex". We assume that machines will have consciousness some day, but what if we had a machine that could do many things as well as a human, but didn't consider itself "alive"?
- _dps 9y agoYes, this is closely connected to the question of qualia, and the concept of a "philosophical zombie" [0]. On the one hand, naive human intuition seems to suggest that the "movie playing in your head" would not emerge from any number of transistors and wires connected to each other. On the other hand, it's not clear how you can argue that what goes on in our own heads is physically any different. So that leads you either to religion, or to the idea that somehow consciousness has to emerge from interactions of matter. And if that's the case, it raises all sorts of ethical questions about the creation of artifacts that might (some day) exhibit consciousness. And, of course, there are already many ethical questions about how we treat animals in the here-and-now. https://en.wikipedia.org/wiki/Philosophical_zombie https://en.wikipedia.org/wiki/Philosophical_zombie
- smokeyj 9y ago> So that leads you either to religion, or to the idea that somehow consciousness has to emerge from interactions of matter. Doesn't seem like they're mutually exclusive. I mean, I'm pretty confident I can interact with some matter and produce another human. If or how God plays a role seems to be the unknown factor.
- CuriouslyC 9y agoWhy does consciousness have to emerge? Why can't it be intrinsic?
- StavrosK 9y agoIntrinsic to what? Individual molecules don't have consciousness, a group of them forming a brain does. It has to emerge at some point.
- CuriouslyC 9y agoHow do you know an individual molecule doesn't have consciousness?
- red75prime 9y agoThere's no observable effects to distinguish conscious and unconscious molecule. Having a definition that can't be used for experimental validation is not useful.
- CuriouslyC 9y agoThere's no observable evidence for an internal experience in humans either. Each person just extrapolates that belief from their own consciousness. That is what got us into the silly position that animals don't possess consciousness in the first place.
- 9y ago
- _jal 9y agoA different question is whether consciousness is something desirable in machines. As an exercise in playing with that question, Peter Watts' fiction is hard to beat. As one distillation he presents goes, first-contact comes about because of our emissions being noticed. But the reason for the contact is the content of the messages - intelligent but not-conscious beings receive them, find them difficult to process because they are mostly self-referential silliness only a conscious mind could worry about, and conclude they are informational attacks. Maybe machines shouldn't be conscious, because consciousness is a waste of resources.
- red75prime 9y agoSome form of awareness of its own existence is required for a machine to be operating efficiently. Otherwise it can plan to use materials of its own "brain" as an intermediate step.
- CuriouslyC 9y agoThe way you're using consciousness conflates the human mind (which I would argue fits your definition of "what it feels like to have a cortex") and the property of experiencing existence. Animals might have a very different mind, but I'm quite certain they experience their existence.
- StavrosK 9y agoMammals still have a cerebral cortex.
- Razengan 9y agoOther questions are: - Would other intelligent species consider our AI machines alive? - Would other intelligent species consider -us- alive? ---- Suppose we build some robots, that know how to learn on their own, teach each other, process raw materials and construct copies (and better versions) of themselves from it. And we send them to an uninhabited planet far away. Decades later, they have a "civilization" of their own, and are colonizing other planets, all on their own. Would they, to an external observer looking for intelligent life, fail to meet any of the criteria for it? If yes, how? ---- Suppose there's an alien species that "sees" in a much wider spectrum than we do; from ultraviolet to infrared, and can see light polarity and perceive billions of more colors than us. They can also innately detect magnetic waves and other phenomena, so they can "feel" way more than humans do. Their natural lives also last for far longer than humans, so their thoughts and plans span decades and centuries, and their minds can casually consider enormous scopes without any effort... Would THEY consider humans "alive?"
- agumonkey 9y agoI prefer to re root the question into primitive (as in thermodynamically) survival. Everything in our system inherited millions of years of response to chaos and sustaining self structure. Machines are thinking already, above average person logical ability. The issue is elsewhere. All we do, even logical processes emerged from that legacy of survival. That's where I draw a line between animal life forms (including us) and machines, even advanced ones.
- glenstein 9y ago>The submarine "swims", but almost none of that activity could be mapped onto how a human swims. Along similar lines Hubert Dreyfus famously said "current claims and hopes for progress in models for making computers intelligent are like the belief that someone climbing the tree is making progress toward reaching the moon." The problem with this way of thinking is that the fundamental principle in question gets conceded like it's no big deal. If you grant that I can swim, that I can climb a tree, even an inch, you've already given me absolutely everything. Of course further progress will be hard, but it won't be forbidden by some kind of transcendent philosophical principle. So with the submarine example, we might "swim" differently, but we can both navigate through water. What would be cause for concern in the sense of raising some fundamental limitation would be if we discover that the place we want to go simple isn't accessible by any "submarine" we could ever build, or that the place we want to go isn't to be found in the "water."