Our unconscious is a mess – neuroscientist
Mind-reading is no longer some psychic power claimed by those with a special ‘gift’. Brain researchers are saying it will be very real very soon. So, will our thoughts cease to be our own? We ask neuroscientist John-Dylan Haynes, the director of the Berlin Center for Advanced Neuroimaging, to give us some expert insight.
Sophie Shevardnadze: Alright, so Professor Haynes, it's really great to have you with us today. So many interesting things to talk about. So computers will be soon able to read our minds and decipher our thoughts into text, kinda like the voice command apps we have right now. But thought is something so abstract – how can you even decipher something like that?
John Dylan Haynes: Well, first of all, I'd say that I'm not so optimistic that we're going to be able to translate our thoughts into text in the very near future. And let me explain why. The reason is that all our thoughts that we have are coded in our brain activity. So that is correct. So if you think about a cat, or a dog or a mouse, you will have a different pattern of brain activity for every one of these thoughts. But if you want this technology to enter day-to-day life, so that you, for example, sit at your desk and you think about a text you want to write, the information is in your brain, but there is no technical way that we can read out this information in everyday setting. There has been a lot of hype recently about, say high tech companies, claiming that they're gonna be able to do that in the near future. But the biggest problem hasn't been solved yet which is how can we get at all these details of brain activity in everyday settings.
SS: So it's not going to be like, instead of “Google, where's the nearest pharmacy?” when you dictate, it's now going to be like, I think “Google, where's the nearest pharmacy?” and it just reads my thoughts – this is not in the near future or –?
JH: In the media, there is a lot of confusion about these technologies. If I put you in a brain scanner, or if I have an EEG, a cap recording your electrical brain activity, I might be able to do something like that. But most technologies that are shown in this field, are technologies that you can’t take around with you on a day-to-day basis. They're not mobile, you can’t sit at your desk with an EEG cap and kind of think about a text and this gets translated into a text like a letter into your Word document.
SS: But is it possible, theoretically, I mean, look at computers, they used to be as big as this room when it was invented and now you know...
JH: Absolutely. So it's possible in principle because we know one thing from brain science, that is that every thought we have is associated with a unique signature pattern of brain activity. So if I think about one thing, for example, do the shopping, I have one very specific pattern of brain activity that only occurs if I think about doing the shopping, and every item on my shopping list will yet again have a different pattern of brain activity associated with it. So it's in the brain, there are two challenges, two main challenges. The one is, I need to be able to read out the details of your brain activity: I need to know ideally what every one of your 86 billion neurons is doing; I need to be able to record your brain activity with very high accuracy and ideally, in a way that is not damaging to you. So I don't want to open your skull, that would be unethical. So I can't really get the full blown details of the brain activity. So we have to rely on techniques like EEG, brain scanners, things like that, that don't have this very high resolution. And the second challenge is that I need to know the code. I need to know how the thought is encoded in your brain. And with that, I need your cooperation. So I need you to tell me what you're thinking and then I can look at the pattern of brain activity that you have while you're thinking this thought. Think about the hieroglyphs in ancient Egypt, you need a translation that tells you what these pictures mean. And we don't really have this kind of translation for brain signals yet.
SS: Where do you technically get that code? I mean, how do you get it?
JH: So there are different possibilities. So you can use the brute force approach. You can say ‘the sledgehammer approach’. The sledgehammer approach is, I put you in a brain scanner, and I'll let you think about a cat, a mouse and a dog and I record your patterns of brain activity and I train a computer to recognise these patterns of brain activity. And then I get you back in the scanner, and I can recognise every time you think about a cat, a dog or a mouse because the same pattern of brain activity occurs again. So the pattern of brain activity is an indicator, a proxy of your thought. It gets more difficult if you want to read out many more thoughts. So let's say I don't want to read just these three animals but let's say I want to take 50 different animals. I would have to learn for every one of these animals what your signature pattern of brain activity is. So my training phase for the computer is much longer, but I’ll also have to have you in the brain scanner longer than you'd have to be in the brain scanner for the first experiment because I have to learn more patterns of brain activity, and you can see immediately: if you take more and more possible thoughts you're trying to read out, this can get longer and longer. And in the extreme case, if you take a dictionary, with all the possible words you might want to read out, it could get really difficult, in fact, impossible to do this. So you have to find a different way. And the different way for doing this is to use mathematical models that understand kind of systematicity, so the systematics behind different thoughts. For example, cats and dogs and mice are mammals and they will have a different pattern of brain activity than, for example, lizards. And you can start learning the structure in this way as well.
SS: This code has to be individual for every being, right? It cannot be like a one-size-fits-all thing, right? Because when we ordinary people hear that there's a brain scanner that reads your thoughts, we automatically think there's just like magical capsule that you put a person inside and that sort of translates what that person is thinking from A to Z, that's like really far away from reality and may never happen, or it may happen?
JH: Well, that's what I call a universal mind-reading machine and I use an example to show what this would have to do. So imagine a stage trick, where you have someone who can supposedly read the thoughts of a person in the audience, which is a stage trick. But now you put a brain scanner on the stage and you take a random person from the audience and you say, ‘Please get in the brain scanner, and have a random thought, think about you whatever you want.’ And you record that and you should be able to tell what the person is thinking. So an arbitrary person having an arbitrary thought, that's a universal mind-reading machine and that is not available, I think, in the next 50 years or anything like that. And the reason is because we can have so many different thoughts and the way every brain codes information is quite different. But we can do simpler things with these technologies, where we don't have to be able to read out every possible thought, and also where the patterns of brain activity are more similar between people. So where we agree more people are more similar in the way they save the information in their brain.
SS: I was talking to your colleague, Susan Greenfield, maybe you know her, and she was talking about consciousness and why it is so hard to decipher consciousness is because you can't hack into people's feelings. So when we're talking about thoughts, is it kind of the same? Is it just as difficult to hack into people’s thoughts as it is into feelings, what we call consciousness, or easier?
JH: Well, with a brain scanner and a computer trying to decode what the person is thinking, you can aim at different things, you can aim at a person's conscious thoughts, what's going through my mind right now, and you can aim at unconscious processes, things like tendencies, you can also aim at emotions. So there are many different kinds of what's called technically ‘mental states’. So some kind of mind processes going on inside you that you can aim for and try and decipher. In fact, consciousness is easier to decode than unconscious processes because conscious processes have a wide effect in the brain. They don't just affect a small part of the brain, but a large part of the brain is affected when you think consciously about something. So there's more to go with the pattern is more clear, and it's wider spread and it's easier to read it out.
SS: Of course, but what I really want to know after you telling me this is, can a brain scanner eventually potentially tap into unconsciousness? I mean, imagine thousands of dollars that that would save all of us going to psychotherapists and psychoanalysts and unravelling all of our deepest fears and patterns.
JH: So I don't think that a brain scanner is the solution to every problem we have in psychology. I think we have to be careful. So, first of all, often people think that if you find something in the brain, and we have this idea that the brain is a mechanistic system, that means we can change things, it becomes more like a natural science mechanism that we can manipulate as well. But for most everyday applications, you don't need a brain scanner, and you don't need neuroscience, you need a psychologist. And the psychologist also changes your brain activity. If a psychologist asks you, ‘Did you have any difficult experiences when you were a child?’ and you trigger a recall process from someone's memory, that is a brain process that the psychologist is triggering, and going through this process might help you feel better. So you don't need a brain scanner for everything. Just even if the brain is the basis for all our thoughts, it doesn't mean that the brain scanner is the most efficient way of getting this kind of information.
SS: Professor Haynes, imagine if it’s possible to have a machine that would with highest possible accuracy decipher your unconsciousness? Because when you go to a psychologist or any kind of person, whether it's psychoanalysis or psychiatrist, he gives you his interpretation of what he thinks lies in your unconsciousness. But what if you had a machine that could accurately unravel what is in your unconsciousness?
JH: So I think the unconscious process – so I'm not a Freudian, I think Freud is very mistaken, and psychoanalysts are very mistaken about the idea of there being like this parallel universe in your mind of the unconscious, where all sorts of kind of interpretable things happen. I think the unconscious is just an amorphous mess, an unorganised mess of all sorts of bits and pieces and fragments. And I don't think it's possible to interpret the unconscious in this way in which we interpret other things. I think my analogy for this is people often try to interpret dreams. I think that's very misleading. For me, a dream is like, something that happens in the theatre between two acts. So let's just say you have one act, the curtain is open, something systematic happens, and you can interpret, then you close the curtain, there is an interlude. And what people are doing is they're moving all the props around on the stage, some of them go down to the basement and new props go on the stage. There's just a big mess. It's systematic but you can't interpret what's happening. It's not a theatre play that's happening there. So if you were to open the curtain during the interlude, you can't start interpreting that. It's just housekeeping, all sorts of weird processes. And I think that's the way the unconscious works and that’s the way dreams work. It's something intangible, chaotic, fragmented, that is very hard to interpret. So the question will be, what is the outcome of a machine that will decipher the unconscious? I think the outcome of such a machine would be it just gives you your pattern of brain activity. Thus, there is no simpler way of looking at this, I don't think you can see an unconscious conflict or something like that.
SS: Okay. So if we're just talking about a machine that gives you a pattern of your brain activity, and as we go on, it will get more and more accurate. Do I understand correctly?
SS: Brain scanner. Practical purposes: can we use that, for instance, to prevent crimes, to detect patterns of the criminal or intentions of malice?
SS: And if yes, can we also sort of outplay the brain scanner like we do with the lie detector when you just sort of command your emotions?
JH: That's a very important question, I think. When you come to the applications, you quickly end up in forensic or criminological situations. So the question is, ‘Can I prevent a crime from happening? Can I tell that someone is going to commit a crime?’ There is a famous movie called ‘Minority Report’, and they have a pre-crime division, which means that they try and they have these kind of telepathic mediums, and they can predict when a crime is going to happen, and they intervene, and they stop the crime from happening just before it happens. Now, the problem there is that people don't really commit that much. So, say, someone, for example, goes to the supermarket, and they want to steal something, and they’ve brought a bag where they can hide it in. But when they get to the supermarket, they might change their mind. So just because someone goes to the supermarket with the intention to steal something, it doesn't mean that they're actually going to pull it through when they get there. So it's very difficult to say that just because someone has a plan to do something, it's fully determined that they will definitely go and do it. That's the difficulty in that field. And another example, take, for example, the terrorist attacks in 2001. There are these famous pictures of two of the terrorists passing through, I think it's Portland airport, they're caught on security camera there. So what if you were to put these people in a brain scanner at that time, would you be able to find out what they're planning to do? It's quite difficult. And I'll tell you why it's difficult. They have a plan. But it might be difficult because they know that they're going to undergo a brain scan. So they might use a countermeasure. So they will just think about something else. So they might think about what they had for breakfast in the morning. So if you're not consciously thinking about something, attending something right at a point in time, it's very difficult to read that out. And also, you might accidentally falsely pick up people who are not terrorists. For example, someone might think, ‘I hope so nobody's blowing up a plane,’ or someone else thinks, ‘Oh, I'm going to this brain scanner test, and I am obsessive-compulsive and I can't prevent myself from thinking about blowing up a plane’ or something like that. The applications can get very difficult, if you want to be definitely sure that you can catch someone committing the crime, that you can make a very strong prediction. If, of course, you just want to have a suspicion and double-check or something like that, that will be a different situation.
SS: I remember that movie ‘Inception’, it used to be one of my favourites, it still is. This whole brain scanner thing potentially sounds like this is where it could lead us where you can put your thoughts into other people's thoughts. Because if you can decipher it, and you can decode it and hack into someone's brain, then probably you can also incept your thoughts in their heads, no?
JH: Yeah. So this is something that is very popular in science fiction movies. And there are two types this takes. One is reading out what a person is thinking, mental state decoding, mind reading; and the other is the manipulative side, writing something into a person's brain. I always compare this to a RECORD and PLAY button. So the RECORD button is reading out, and the PLAYBACK button is writing something into a person's brain. And the problem with the PLAYBACK, so sort of writing something into a person's brain, is we don't have any proper technology at the moment where we could write a pattern of brain activity, a specific pattern of brain activity into a person's brain. Now, the way in which information is stored in the brain about our thoughts is done in something like a mosaic. So every thought– Think about a mosaic of lots of little colour tiles, that's the way your brain stores information. And what we have is you can possibly use something like a sledgehammer to just kind of disrupt processing completely in all of the tiles. But the pattern that you need to do to create a specific thought that has to kind of recreate exactly the mosaic that you need for a person to think a specific thought and that's just not possible right at the moment. We don't have a technology that will allow us to do that.
SS: How does hypnosis work on a person's mind?
JH: Well, there are people who would claim that hypnosis doesn't even exist.
SS: Are you one of those?
JH: I'm one of those. Yeah.
SS: Okay, and then I'm thinking if this machine ever takes off and becomes broadly used and actually develops, and becomes more up-to-date or more accurate, then always the ethical side sort of kicks in because I was always under the impression that my thoughts are like my personal secret garden and no one can peep into them or look into them. Now all of a sudden, if we do have this kind of machine that reads your mind, or thoughts, more accurately, that could be widely used. So are my secrets gone forever? Well, what would happen if this machine goes into the wide use?
JH: Well, I think now we're talking about science fiction.
SS: Are we?
JH: What I mean is, “Let's not think about the technology now but what could it potentially do if it were to work better and better and better, and the ethical dimensions of that.” And I think this definitely raises the question of mental privacy: whether we should be given access to other people's thoughts, or if they should not be a possibility for someone to control the access that other people have to their thoughts. I'm used to my thoughts being private, and I don't want all my thoughts to become public. I don't want some people to know some things I'm thinking. So what if every one of us was running around, and we had a brain scanner hat, and there's a little TV screen, and everyone can read on the TV screen what we're currently thinking, would that be a better or a worse society? We think that would be bad because we're entitled to the idea of mental privacy. We're entitled to secrecy about our thoughts. There is no such thing as a thought crime. If you wish you could kill someone that's not a crime, it's a crime if you do kill someone. So something just happening in our mind is not something that is legally relevant if it has no consequence in the real world. So this would change the border, possibly, if we can now look into people's minds, read their thoughts. The question then is, what kind of consequence we could take from this? I think, you take the example of lying, it would prevent us from lying, we wouldn't be able to lie anymore if everything becomes transparent about what we're thinking. So is this bad or good? We tend to think that's bad. Because there might be a situation where, say, my wife has a new dress and she says, ‘Do you like my new dress?’ and I don't want to hurt her feelings and I say, ‘Yeah, it's good. I like it.’ But I might not think that. If she could see on my little cap that I'm thinking something different, that I'm actually lying, then this might hurt her feelings. So we want to be able to protect people’s feelings from the thoughts we have now. But at the same time, my wife knows that when she comes home, and she has been wearing a dress, and I say I like it, she can't fully trust it because she knows that I might also be nice to her in a case where I don't think it's such a nice dress. So there's always a little bit of a doubt. And there is a movement, and I don't follow this movement, but I just want to say that there is a movement where people think differently, it's the Radical Honesty movement. And they say, “We should always be completely honest about everything we do. And we should always be open and upfront about our thoughts.” I don't follow that principle. I think I'm not born for that. I don't think society is there yet. But in future, you don't know. Perhaps sometime in the future people will look back at us and think it's terrible for all the lies we've said in the past.
SS: Or not.
JH: Or not.
SS: My last question is a very neuroscientific question that has to do with the brain scanner but it's like a forever question. We're talking about our fundamental rights, like the privacy of your thoughts and then another one is free will. So we, ordinary people, believe that we're in command of our actions, and you know, I send the brain signal to lift my right hand, and then the brain sends the signal to my body and it lifts. But all the neuroscientists tell me the opposite, that the brain decides before I even know it, that I want to lift my hand and then I don't really understand where am I in all that? And is there such thing as free will or is it just an illusion?
JH: Well, first of all, it's your brain. It's not someone else's brain, it's your brain. So you have ownership over this brain process as well. You just don't aware of the process happening. And there's a second aspect to this, which is that despite the fact that something starts, you still have the possibility to a very late stage of stopping things. So you might not know what is triggering you to do something, but you can still stop the process. So if you realise you're doing something you don't want to do, you're starting to do something you don't want to do, you can still exert control until a very late stage.
SS: So like when you tell me, ‘Let's go have coffee,’ and I'm like, ‘Yeah, let's go have coffee but no, actually, I'm going to have a tea…’ Is this what’s just happened?
JH: Yes, exactly. All the decisions we make are realised by the brain, the brain is like the carrier of our decisions similar to the grooves on the CD being the carrier of the music, and it can't exist independent of the grooves on the CD. The thoughts we have are encoded in our brain and the patterns of activity in our brain. And they follow the principles of nature, they follow the laws of nature, and they're predictable.
SS: So free will is governed by instincts? Because a lot of people think it's governed by unconsciousness...
JH: We are not free to change the course of the world, things happen the way they happen. And nonetheless, the decisions we make are still our decisions. If I want something, I will try and do it. It's my thing. It's my wish to do that. So it's a way of expressing what I want. But I'm not in control of what I want, that’s just something that happens.
SS: So if everyone says that any thought that you have in your brain or your mind, or any action that you undertake, has to do with the electrical activity in your brain, if that's the case, that would mean that our brain also somehow works like a compass and our thoughts, our decisions are also based on magnetic fields and where the sun stands and how the moon stands. Is that somewhat true?
JH: Well, I'd say the electrical fields and the magnetic fields we have around us don't influence our thoughts very much. And I can tell you, there's a simple explanation why, because any electric field or magnetic field is very unlikely it has any effect on neurons. But if it does, it will do the same thing to all the neurons we have. But in order to specifically manipulate a thought you would have to change the activity in every neuron differently. You'd have to write something very specific into the brain, a very specific pattern and these diffuse processes or fields like electrical magnetic fields, they are not able to do something as specific as that.
SS: Professor Haynes, thank you very much for your insight. It's been a pleasure talking to you.
JH: Likewise. Thank you very much.