Placeholder Image

字幕列表 影片播放

  • Thanks to LastPass for sponsoring this episode.

  • LastPass relieves the burden of remembering passwords.

  • Click the link in the description or stay to the end of the video to learn more.

  • [♪ INTRO]

  • If you've ever had a conversation with Siri or Alexa or your Google Home,

  • it might seem like machines are just a couple of steps away from being fully conscious beings;

  • things that actually experience the world.

  • But since we only have insight into our own experiences,

  • humans are the only things we know are conscious.

  • And the fact that we are is one of the most obvious things about us.

  • It's part of what makes us human.

  • Even so, figuring out exactly what consciousness is

  • and whether or not it could emerge in non-human things has stumped us for centuries.

  • Scientists can't even agree on whether or not it's possible

  • to study something that's so personal using the scientific method.

  • But some say that it's not just possible, it's necessary.

  • A theory of consciousness could help us figure out if a given coma patient is likely to recover.

  • It could influence the way we treat animals.

  • And it could even help us navigate our relationship with technology.

  • So, since the 1980s, researchers have been taking a crack at

  • figuring out consciousness from a scientific perspective.

  • For now, we're still nowhere near solving it, but we have learned some things

  • about what a theory of consciousness might eventually look like.

  • Except, first things first. We have a problem with figuring out consciousness:

  • it's that we have to define what it is.

  • It's hard to put a finger on exactly what it means.

  • For most purposes, you can think of consciousness

  • as the name for what it feels like to experience the world,

  • rather than just interact with it through inputs and outputs.

  • And, yes, every interaction starts with an input.

  • The things we see and smell and hear in our environment turn into electric signals

  • that travel through our brain, thanks to charged atoms moving in and out of neurons.

  • But somehow, the movement of those atoms

  • can create the experience of hearing music or smelling a rose.

  • And that experience? That is consciousness.

  • If you don't see how that's different from

  • a robot responding to an input with an output, think about this:

  • Your body responds to lots of stimuli without your conscious involvement.

  • For example, food makes your mouth water, and cold makes your hair stand up,

  • but you don't have to be conscious of those things for them to happen.

  • They just happen. Like a robot.

  • On the other hand, you're conscious of the way things smell or feel.

  • You're aware of what it feels like to be you.

  • And while your subconscious brain has countless processes happening in parallel,

  • your conscious experience is a single stream of events, almost like a story.

  • So you can think of consciousness as meaningexperienceorawareness.”

  • And scientists more or less agree up to that point. The question is where to go from there.

  • The study of consciousness as a scientific thing, as opposed to a philosophical one,

  • took off in the 1980s with Francis Crick, a neuroscientist

  • and, yeah, you've probably heard this name,

  • co-discoverer of the double-helix structure of DNA.

  • After he'd basically decoded life, he was ready to take on what he saw as

  • the next big scientific mystery: consciousness.

  • And he recruited a bunch of young researchers to get involved.

  • But Crick and his collaborators didn't set out

  • looking for the ultimate explanation of consciousness.

  • Instead, as a starting point, their goal was to find the part of the brain that gave rise to it.

  • I mean, it wasn't universally accepted that consciousness began and ended with the brain,

  • and it still isn't today, but that was most people's starting assumption.

  • This team called what they were looking for the neural correlates of consciousness.

  • And over the past few decades, they and other scientists have made some progress.

  • For instance, some research suggests that conscious experience

  • is rooted in a part of the brain called the posterior cortex.

  • But again, this is still a correlate of consciousness, not a cause.

  • So even if they do manage to narrow down the thing that gives rise to consciousness,

  • scientists still face the biggest question: how.

  • In an attempt to get to the bottom of that, two major ideas have emerged

  • about how we might study consciousness in a scientific way;

  • that is, with testable hypotheses and measurable results.

  • Neither one is an actual answer at this point,

  • but their goal is just to find ways to study this subjective experience using rigorous science.

  • So, they may be off-base, but at the very least,

  • they offer ways to approach the problem that are scientifically testable.

  • The older of those ideas is called the Global Workspace Theory.

  • It was developed by the neuroscientist Bernard Baars,

  • and it's been around in some form since 1982.

  • It suggests that there's a network of cells in your brain called a workspace

  • that's at the root of your conscious experiences.

  • At any given time, you've got all kinds of signals milling about in your brain,

  • but according to this theory, this workspace is kind of like the fan cam of your brain.

  • Any signal that happens to be processed by those cells

  • gets broadcast to the rest of your brain, and you become conscious of it.

  • This workspace is really limited, though, and it can only hold so much information at once.

  • Which could explain why information and ideas drift in and out of your consciousness so easily.

  • Like, you might be fixated on an annoying sound, and then forget all about it

  • as soon as someone involves you in an interesting conversation.

  • What we don't have is any proof that this idea is correct,

  • but there is some preliminary evidence for it.

  • For instance, in an experiment published in 2001, a team of researchers had

  • 15 participants look at words flashed on screens for 29 milliseconds each.

  • That is long enough for a word to become readable, but in some cases,

  • the experimenters used a technique called masking

  • to prevent the subjects from consciously registering the word.

  • They did that by flashing another image after the word,

  • which interrupted their conscious processing.

  • By imaging the participants' brains, the researchers were able to see that

  • the words that remained subconscious only produced a small amount of activity in the brain,

  • while the words that the participants became conscious of

  • triggered a whole flurry of activity in many different regions.

  • This wasn't necessarily proof of the Global Workspace Theory,

  • but it did give the experimenters a clear picture of the brain areas

  • involved in subconscious and conscious processing of the same signal.

  • And in later research, scientists have used that as a starting point

  • to try and figure out how messages that pass through this network

  • might get broadcast to the rest of the brain.

  • It is still far from a solution to consciousness, but this theory does a few things well.

  • For one, by suggesting that conscious processing is limited to a specific network,

  • it offers a possible explanation for why

  • our conscious brain is only capable of narrow streams of thought,

  • even though our brains can subconsciously process so many things in parallel.

  • But more importantly, the fact that it involved questions that were testable

  • gives us a way to study consciousness using the scientific method!

  • The theory still has its sticking points, though.

  • Like, it goes one step further than the neural correlates of consciousness,

  • but it still doesn't provide the ultimate how.

  • And then, there's also some resistance to the idea that consciousness is computational;

  • that it just comes about because of the way the brain is hooked up.

  • Because if that's true, it suggests that, in theory,

  • there's nothing keeping machines from gaining consciousness;

  • that it's all a question of having the right wiring.

  • And that would open up a whole different can of ethical worms.

  • But it is a testable hypothesis. And in science, that counts for a lot.

  • This isn't the only hypothesis out there, though,

  • and others take a totally different approach.

  • For example, in 2004, the neuroscientist Giulio Tononi proposed

  • what he called Integrated Information Theory, or IIT.

  • Since it's impossible to scientifically observe someone's personal, conscious experience,

  • even if we can see some of what it looks like in the brain,

  • Tononi crafted his theory by working backwards

  • from the few things that seem to be universally true about consciousness.

  • For one, we know that it's subjective,

  • meaning that you are the only person having your conscious experience.

  • No one else can step in and experience what the world is like for you.

  • Second, we know that the experience of consciousness isunified,”

  • meaning it can't be split into pieces.

  • Like, you are one thing; you cannot willfully split your sense of self into two selves.

  • You also can't, say, decide to only process certain kinds of information.

  • You can't wake up tomorrow and decide you're only gonna see the color blue,

  • or only smell nice things.

  • Although I think that would be a handy skill.

  • Tononi took these observations as starting points, along with the assumption that

  • consciousness somehow comes from our web of interconnected neurons.

  • Then he proposed that consciousness comes from

  • the amount of interconnectedness in a system,

  • or, in his words, the amount of integrated information.

  • IIT suggests that, essentially, the whole of all your neurons working together

  • amounts to more than the sum of its parts.

  • So, connected neurons can create an experience that individual ones cannot.

  • Again, not everyone is on board with this idea, and testing is still in an early stage.

  • But researchers have done some preliminary experiments that compare

  • the amount of connectivity in a brain to a person's level of consciousness.

  • For example, in one study, Tononi and his collaborators rounded up 11 volunteers

  • and used a magnetic pulse to deliver a burst of stimulation to neurons in their brains.

  • Then, they used sensors on the scalp to measure the amount of activity that pulse produced.

  • Next, they sedated the subjects and performed the same experiment.

  • The second time around, the pulse produced much less activity,

  • suggesting that there were fewer connections between neurons,

  • so that triggering one group of them

  • didn't set off the same chain reaction that it had the first time around.

  • It's still not possible to directly measure the amount of connectivity in the human brain,

  • but experiments like this can serve as a decent proxy;

  • a way of figuring out if there really is a link between integration and consciousness.

  • If there is, that implies that consciousness exists on a spectrum.

  • And when you play this theory out,

  • it implies that not only could machines become conscious,

  • but everything with any amount of interconnected information,

  • from a wasp to the internet, might already be a little bit conscious.

  • Like the Global Workspace Theory, IIT doesn't yet provide

  • any kind of satisfying answer to the problem of consciousness,

  • and the consequences also seriously challenge things

  • we instinctively believe to be true about the world.

  • So if it's true, we might have to think a little harder about how we interact with things.

  • Still, the idea of measuring interconnectedness

  • provides a path for exploring consciousness scientifically.

  • And, like with the Global Workspace Theory, that is what makes this idea stand out.

  • These hypotheses are still works in progress, but if either one proves true,

  • it could help us figure what has consciousness and what does not.

  • Like, if IIT is right, measuring the amount of interconnectedness in a brain

  • could help doctors and scientists decide how conscious a coma patient is.

  • Or it could help future computer scientists answer those same questions

  • about artificial intelligence programs.

  • On the other hand, if the Global Workspace Theory ends up being right,

  • and if scientists can identify that workspace, the fancam of the brain,

  • doctors could look for activity in that region to see if a patient has any signs of consciousness.

  • So far, neither one of those ideas is developed enough to be useful in any practical sense,

  • so for now, there's no foolproof way to identify consciousness.

  • And until we understand how it arises,

  • we probably won't be able to say what has it or not, either.

  • But the fact that we are conscious means we're going to keep being curious,

  • and scientists are gonna keep looking, as long as the problem is unsolved.

  • Thank you to LastPass for sponsoring this video.

  • LastPass saves you from having to remember your dozens or hundreds of passwords.

  • If you're consistently getting locked out of your accounts

  • or resetting passwords, LastPass can help.

  • With this app, you never have to write, remember, or reset your passwords.

  • You can store an unlimited number of them or share them with a trusted team member.

  • LastPass auto-fills your credentials in your browsers, making it fast and easy to log in.

  • You can also find the app on iOS and Android and sync your passwords across devices.

  • To start using LastPass today, check out the link in the description.

  • [♪ OUTRO]

Thanks to LastPass for sponsoring this episode.

字幕與單字

單字即點即查 點擊單字可以查詢單字解釋

B1 中級

科學如何理解意識? (How Science Is Trying to Understand Consciousness)

  • 6 0
    林宜悉 發佈於 2021 年 01 月 14 日
影片單字