Hacker News Reader

Comment by azakai | Hacker News Reader

As someone with a computational neuroscience background, I am not sure I agree that those are the right authorities to decide on consciousness in animals,

> cognitive neuroscientists, neuropharmacologiests, neurophysiologists, neuroanatomist and computational neuroscientists

Don't get me wrong, it's great to have their input on this topic. But to not include philosophers of mind for example seems very wrong. In particular because consciousness is very hard to define. The definition there involves "affective states" and "intentional behaviors" which in themselves are also extremely hard to define.

edit: Also, how about psychologists studying animal behavior, ethologists, etc.

View on HN

This article is as absurd as if it had said: "It's official, scientists have discovered the meaning of life!"

The gall of these scientists to claim that not only can they definitively say what consciousness is, but that they can prove that something has it is astonishing.

> The gall of these scientists to claim that not only can they definitively say what consciousness is,

That isn't what they say. The author of the article is editorializing, and the HN title needs to be changed. (puts up mod alert spotlight).

What they say is what consciousness is not, that is, it does not require a complex brain, that's not a prerequisite. They don't say very much about what it is, just its prerequisites. Buried in their supporting arguments, they state that "birds appear to offer" "parallel evolution" of "near-human-like levels of consciousness". As a non-biologist I am not entirely sure what exactly is meant by that, but it is not within a hundred miles of a "proof".

Yeah, should've known the article would grossly misrepresent the claims of the scientists.

Agreed. To my mind, this is one of the things that makes it so hard to REALLY achieve artificial intelligence - we don't even know exactly what we're supposed to be creating. It seems to me that CS types who are actually working on AI (as opposed to sophisticated machine learning) are going to need some help from philosophers to even fully define the problem.

As 'CS type' actually working on ML, I disagree. I'd rather prefer some help from hardware guys, that would allow me to run a neuro network with 100 billion neurons really really fast.

If it's your long term dream of fully replicating humans, it would be helpful to read Kant and similar philosophers to have a clear understanding of what consciousness is.

If it's not and the goal is just replicating and accelerating certain cognitive abilities, then it might not be super helpful and possibly disheartening (disheartening because of how much we lack full understanding of consciousness due to limitations of only being able to understand it via our own consciousness).

My undergraduate degree was in philosophy, with a concentration in philosophy of mind, and while I can't recommend highly enough taking the time to read through and learn from the sundry perspectives and takes we've had on consciousness throughout history, be warned that that way lies confusion and dragons.

Kant certainly didn't know what consciousness was. He had some instructive things to say, and massively helped frame the debate for people who came later, but even present-day philosophers who have the benefit of modern science and technology to help in their investigation are no closer to pinning down how it is that meat can feel.

Absolutely not. After 2500 years, Western philosophy has yet to achieve broad consensus on anything at all. Philosophers don't have a clue what consciousness is - they can barely even describe it without using utterly ambiguous words like "qualia" or "sense datum". The only way we'll ever be able to know anything about consciousness besides what's going on inside our own heads is for us to directly experience what's going on inside someone else's (or some animal's).

Not until the Singularity, basically. I'm not going to hold my breath for that one.

> If it's your long term dream of fully replicating humans, it would be helpful to read Kant and similar philosophers to have a clear understanding of what consciousness is.

From a scientific perspective, this elevates philosophy beyond anything reasonable. Philosophers have many things to say, but they tend not to be the kind of people who are trained to ask, "Is there an empirical, falsifiable test of my claim?"

Throw all the hardware you want at it. Without philosophy you have no way to answer the question: what does it mean? Do you think you can just simulate 100 billion neurons and then ask the computer?

> what does it mean?

Assuming that it means something to begin with...

The question of whether a computer can think is no more interesting than whether a submarine can swim

Call it what you want, if it works, that's all that matters. Also, presuming that a "thinking" brain must be like a human brain is naive.

http://fcmconference.org/img/FCMCProgram.pdf

Have a look at the pannel of the conference where the declaration was published and signed...

Even if they are not philosophers, (human or animal) psychologists or ethologists, the work closely with some.

Among these are some of the smartest people I met in my life (and I never met Stephen Hawing).

Edit: Two of the panelists are animal psychologists, actually.

I did take a look. They are all brilliant. But the declaration mentions only neuroscience-related specialties, not philosophy or (non-neuro-) psychology.

Again, not a criticism, I have the utmost respect for them. But for such a declaration - which I 100% support - it would have been nicer to have a more rounded panel.

Reply on news.ycombinator.com