Thread: Neuroscientists reconstruct voice from internal dialogue

Telepathy machine reconstructs speech from brainwaves - health - 31 January 2012 - New Scientist

Runs rings around the pathetic attempts to reconstruct images. Which were really just a novelty with no real future in them.

I guess if they ever get to a point where this can be done easily, it's a massive +1 for the meditaters who can shut their brain up

Automatic dream journal entries! So much for my plan of using Morse Code.

Originally Posted by IAmCoder

Automatic dream journal entries! So much for my plan of using Morse Code.

Awwww shit.... how did I not think of that? lol
Could record it while sleeping.

I'm sure there's still another several years before it becomes useable.
So in the meantime your Morse code idea sounds great.
Have you gotten anywhere with that yet?

That's awesome. But could you imagine that kind of technology in the hands of a totalitarian state?

dat thought crime D:

Say what y'all will, I'm just going to be freaking out [in a good way] when technology can record the vocal track I hear in my head for my songs.
It gets really hard when I have to hit a high E-flat in real life, but I can hear that sound in my head just fine! ^_^

Originally Posted by tommo

Runs rings around the pathetic attempts to reconstruct images. Which were really just a novelty with no real future in them.

I find it disarmingly cute how you always feels compelled to insert some bizarre kernel of impenetrable absurdity into your posts.

These new technologies are all pretty cool, although none of them really represents any new understanding. I'm not sure of any practical benefits to this... speaking with my mouth was never much of an issue. I suppose for mutes it is very promising.

Originally Posted by Xei

I find it disarmingly cute how you always feels compelled to insert some bizarre kernel of impenetrable absurdity into your posts.

I just find it hilarious when you find any possible sentence in a person's post to baselessly argue with, without so much as mentioning a word of what the person actually said. That level of passive and vague insult takes skill, and is humorous on numerous levels.

Originally Posted by Xei

These new technologies are all pretty cool, although none of them really represents any new understanding. I'm not sure of any practical benefits to this... speaking with my mouth was never much of an issue. I suppose for mutes it is very promising.

Recording dream journals from in dreams would be one example, or for paralysed people, people with locked-in syndrome. Although they were all mentioned in the article and previous posts so I'm not surprised you were oblivious to that.

Originally Posted by Xei

These new technologies are all pretty cool, although none of them really represents any new understanding. I'm not sure of any practical benefits to this... speaking with my mouth was never much of an issue. I suppose for mutes it is very promising.

Every step that's taken towards true BCI represents a massive shift in the trajectory of the human species, with BCI being (in my opinion) an actual singularity in the sense that history, technology, and the species will be irrevocably changed.

Originally Posted by Xei

I find it disarmingly cute how you always feels compelled to insert some bizarre kernel of impenetrable absurdity into your posts.

These new technologies are all pretty cool, although none of them really represents any new understanding. I'm not sure of any practical benefits to this... speaking with my mouth was never much of an issue. I suppose for mutes it is very promising.

Are you kidding? There could be tons of practical benefits of this.

You do know what 'mute' means, right..?

IAmCoder already said that it'd make communication with the external world in lucid dreams much easier than previously, I didn't think saying it twice would add much to the conversation... sorry?

I don't tend to disagree that an understanding of neuroscience and developments in technology will greatly change our species, it's just that I don't see how this in particular is very important. It'd be kind of redundant learning how to speak again when just using your existing voicebox to speak, which can then be detected by a computer just as immediately, pretty much amounts to the same thing.

Originally Posted by Xei

You do know what 'mute' means, right..?

IAmCoder already said that it'd make communication with the external world in lucid dreams much easier than previously, I didn't think saying it twice would add much to the conversation... sorry?

You said you're "not sure of any practical benefits to this" which is a lot different from simply not mentioning something that someone else already said.
There's also the use that people could record their internal dialogue for a day and realise how insane they really are. It could also be used to shut off the internal dialogue which is entirely useless in most situations, if not all.

Originally Posted by Xei

I don't tend to disagree that an understanding of neuroscience and developments in technology will greatly change our species, it's just that I don't see how this in particular is very important. It'd be kind of redundant learning how to speak again when just using your existing voicebox to speak, which can then be detected by a computer just as immediately, pretty much amounts to the same thing.

We could get information from terrorists without torture. Just a thought.

Nice, can't watch the video where I am but would be interesting to see results on someone in a coma. Or used on someone seconds after passing away..

Originally Posted by ShadowOfSelf

Nice, can't watch the video where I am but would be interesting to see results on someone in a coma. Or used on someone seconds after passing away..

They wouldn't be able to use this device.

Sounds like they still got a little ways to go, because the sounds were just barely recognizable. I wouldn't underestimate how useful it would be for people who can not speak though. Your first reaction might be that it isn't many, but there are actually millions of people just in the US who could benefit from it.

Originally Posted by cmind

They wouldn't be able to use this device.

Why not?

Originally Posted by ShadowOfSelf

Why not?

Well, if it reconstructs internal dialogue, there has to be internal dialogue to reconstruct.

Originally Posted by snoop

Well, if it reconstructs internal dialogue, there has to be internal dialogue to reconstruct.

Well yeah, but do we know if people in comas have internal dialogue? Or is this the tech we need to find out.

Lol, there you go again. What are you talking about? You think people will willingly turn off their own cognitive faculties because they're 'useless'?

I can't find any decent bloody articles about the actual methodology as per usual, but it sounds like what they were doing was playing recordings of words to patients and trying to reconstruct the sound from brain activity. It is extremely dubious therefore whether this actually corresponds to 'mind reading' or not, because it is an unsolved question as to whether, or to what extent, human cognition (specifically high level, 'conceptual' cognition) is based on language, or to what extent thoughts in general, or even thoughts about words, produce as a by-product brain activity related to their physical enunciation.

Originally Posted by Xei

Lol, there you go again. What are you talking about? You think people will willingly turn off their own cognitive faculties because they're 'useless'?

Why not? Thinking is a LOT more efficient when you don't have to speak every measly thing you're thinking about.
So, there YOU do again, assuming people will not use this in a beneficial way.
You're seriously arguing the dumbest fucking things recently.

Originally Posted by Xei

I can't find any decent bloody articles about the actual methodology as per usual, but it sounds like what they were doing was playing recordings of words to patients and trying to reconstruct the sound from brain activity. It is extremely dubious therefore whether this actually corresponds to 'mind reading' or not, because it is an unsolved question as to whether, or to what extent, human cognition (specifically high level, 'conceptual' cognition) is based on language, or to what extent thoughts in general, or even thoughts about words, produce as a by-product brain activity related to their physical enunciation.

As far as the article I posted explains it, they "trained" an algorithm to figure out what the neurons were doing. And then used another program to convert that in to audio.
I don't think they were reconstructing and altering it purposely to make it sound more like the original sound.

I was thinking about that last bit too, when I say some words in my head, they are not exactly the same as spoken words.
Like sometimes "t" is hardly pronounced, if at all, same with "s", "ou", "r" and probably a few others.

Originally Posted by tommo

Why not? Thinking is a LOT more efficient when you don't have to speak every measly thing you're thinking about.
So, there YOU do again, assuming people will not use this in a beneficial way.
You're seriously arguing the dumbest fucking things recently.

Your statement is still just as impenetrably bizarre and unfounded, sorry. Cognitive neuroscience is nowhere near developed enough to tell if what you're saying even makes any sense, let alone is true.

The assumption that thought isn't actually based on language, the assumption that if it isn't then it's totally disposable, the assumption that a purely representational language faculty would somehow 'slow down' thought, again the assumption that such a faculty actually functions when not speaking... I don't have any clue why you think these are solid assumptions. But you seem convinced that what you're saying is totally well-founded and sensible.

Originally Posted by Xei

The assumption that thought isn't actually based on language, the assumption that if it isn't then it's totally disposable, the assumption that a purely representational language faculty would somehow 'slow down' thought, again the assumption that such a faculty actually functions when not speaking... I don't have any clue why you think these are solid assumptions. But you seem convinced that what you're saying is totally well-founded and sensible.

Yeah coz everyone doesn't do that on a daily basis or anything. It's one of the most founded things there is.

The inner dialogue is completely disposable. All you have to do is try it and see. Just stop talking in your head. You'll see it serves no positive use whatsoever. I couldn't give a flying fuck whether this has been proven by neuroscience. And it so obviously slows down thought because you have to "say" everything in your head after you actually think of it. If you think the internal dialogue is the actual thought.... I have no hope for you.

Originally Posted by tommo

Yeah coz everyone doesn't do that on a daily basis or anything. It's one of the most founded things there is. I couldn't give a flying fuck whether this has been proven by neuroscience. If you think the internal dialogue is the actual thought.... I have no hope for you.

I suggest you begin a career in cognitive science, because you appear to be in possession of proof of one of the most crucial and contentious issues in the field. As you clearly consider yourself far more intelligent and penetrating than the combined thousands of experimental psychologists, linguists, philosophers, and neuroscientists who actually spend their lives researching these issues, you would definitely be an invaluable asset.

The inner dialogue is completely disposable. All you have to do is try it and see. Just stop talking in your head. You'll see it serves no positive use whatsoever.

Oh wait, you're actually sub sophomoric and incapable of conceptualising an even remotely adequate thought experiment. Glaring flaws include

1. Equating a meditative state in which one nullifies one's own thought processes to a state involving active and spontaneous thought.
2. Failing to consider that one may not be conscious of the mechanisms underlying cognition.
3. More generally completely unaware of the extremely well established pitfalls of introspection in psychology.

And it so obviously slows down thought because you have to "say" everything in your head after you actually think of it.

4. Apparently adopting a rudimentary Cartesian theatre (!) in which potentially parallel mental faculties are clogged up in a centralising unit.

Originally Posted by Xei

I suggest you begin a career in cognitive science, because you appear to be in possession of proof of one of the most crucial and contentious issues in the field. As you clearly consider yourself far more intelligent and penetrating than the combined thousands of experimental psychologists, linguists, philosophers, and neuroscientists who actually spend their lives researching these issues, you would definitely be an invaluable asset.

I have no scientific studies proving it. I do have observation, the same which almost every human shares.

Originally Posted by Xei

Oh wait, you're actually sub sophomoric and incapable of conceptualising an even remotely adequate thought experiment. Glaring flaws include

1. Equating a meditative state in which one nullifies one's own thought processes to a state involving active and spontaneous thought.

No. When did I do this? What I'm saying is you can think so much more clearly when you just shut up.

Originally Posted by Xei

3. More generally completely unaware of the extremely well established pitfalls of introspection in psychology.

That's not at all what I'm talking about. What you linked to is about decisions and behaviour. i.e I did this because I felt this or because this happened etc.
Not objective observations about how much saying things in your head slows down thought.
Just think about when you're studying something, and you are just flowing perfectly, totally in to whatever the subject is. You aren't sitting there internally announcing everything you think.
Instead, you are just having the actual thoughts and doing. Everyone knows reading aloud is slower than just reading, right? (for the average person who hasn't trained to speak 123209 words a minute for a Guinness record). You're gonna pick on the logical jump for this, but, imagine how much faster the thinking process is when you're not talking in your head.

Originally Posted by Xei

4. Apparently adopting a rudimentary Cartesian theatre (!) in which potentially parallel mental faculties are clogged up in a centralising unit.

Nope.