A friend’s visit took me to a museum exhibition that allows you to experience blindness for an hour. You navigate different locations in total blackness with only the voice of your visually impaired guide calling out to you.  Read on for some thoughts and insights about our brain.

Ever since reading the philosophy of David Hume this realization that “I” am my sensory input has stuck with me. A year into my neuro science masters I’ll tell you that “I” am also these top down predictions that have been learnt throughout the years, but these predictions have been created in order to deal with sensory input. So what are blind people’s predictions? How do they experience the world without such a strong sense?

First we were given a cane and a crash course how to walk with it, tapping on the floor directly in front of us. While everyone just took a cane I took some time checking different sizes wanted my ‘extra’ limb to feel good.

Different sense modalities provide us with information about different distances and sight allows us to extend out world model to include further away environments. That’s why I chose a longer cane hoping that would help compensate and provide me with information about further away surroundings.

For me it was a scary experience especially leaving my eyes open. When I closed my eyes and saw nothing the prediction error wasn’t as big. But staring into total nothingness the brain refused to let go that nothing was there. “Are you seeing these white shiny dots?” my friend asked midway. As fake top down hallucinations took over her perception.

In most people alpha waves increase when closing their eyes, even compared to total darkness. This is most likely an inhibitory mechanism to prevent these false inferences. Would this still occur with blind people? Would their alpha waves of blind people be different? I couldn’t find any information on that.

I asked our guide about anxiety levels when she goes out and she said she doesn’t feel frightened unless she is going somewhere totally knew. It felt so limiting. I kept trying to find the challenge within the limitation and also remain patient because everything just takes much longer. It was also very hard that we were a bunch of ‘new’ blind people constantly bumping into each other. It was so tiring for my brain and I kept sitting down whenever I managed to feel up a bench or a sofa.

The next realization I had was that the directionality of perception spread out. Without the visual input coming from the front there was a different sense of three dimensionality, my brain was modeling the back of the body and the sides with the same amount of resources. This resulted in a slight sense of floatation.

We were taken to a fake supermarket trying to feel our way through shelves packed with stuff. I asked about technological aid and was told that some apps exist where you can take a picture and get information about it. Now I see that there is also a site where you can lend your eyes to a blind person. http://www.bemyeyes.org/. Too bad it’s just on iphone

Lately I’ve been thinking a lot about what happen when two sets of eyes meet and stare at each other.  It’s not just the knowledge that someone is watching you, it’s the knowledge that they are admitting to watching you. It’s like a ‘handshake’ protocol both sides saying “Yes, I see you”.   Supposedly blind people don’t have this, they don’t know when people are ‘looking’ at them and I was wandering if this gave them some extra freedom but my guide said she somehow did have a sense of when she was being looked at.

Finally we were led into the light and I felt how the power structure changed. In the dark we were dependent on our guide to call out to us so we would follow in the right direction and now suddenly we had access to much more information than her. I thanked her for the profound experience hoping that one day she would at least be able to have a taste of our world.

Seeing Is Believing

A friend’s visit took me to a museum exhibition that allows you to experience blindness for an hour. You navigate different locations in total blackness with only the voice of your visually impaired guide calling out to you.  Read on for some thoughts and a recommendation to try your-self.

Ever since reading the philosophy of David Hume this realization that “I” am my sensory input has stuck with me. A year into my neuro science masters I’ll tell you that “I” am also these top down predictions that have been learnt throughout the years, but these predictions have been created in order to deal with sensory input. So what are blind people’s predictions? How do they experience the world without such a strong sense?

First we were given a cane and a crash course how to walk with it, tapping on the floor directly in front of us. While everyone just took a cane I took some time checking different sizes wanted my ‘extra’ limb to feel good.

Different sense modalities provide us with information about different distances and sight allows us to extend out world model to include further away environments. That’s why I chose a longer cane hoping that would help compensate and provide me with information about further away surroundings.

For me it was a scary experience especially leaving my eyes open. When I closed my eyes and saw nothing the prediction error wasn’t as big. But staring into total nothingness the brain refused to let go that nothing was there. “Are you seeing these white shiny dots?” my friend asked midway. As fake top down hallucinations took over her perception.  

In most people alpha waves increase when closing their eyes, even compared to total darkness. This is most likely an inhibitory mechanism to prevent these false inferences. Would this still occur with blind people? Would their alpha waves of blind people be different? I couldn’t find any information on that.

I asked our guide about anxiety levels when she goes out and she said she doesn’t feel frightened unless she is going somewhere totally knew. It felt so limiting. I kept trying to find the challenge within the limitation and also remain patient because everything just takes much longer. It was also very hard that we were a bunch of ‘new’ blind people constantly bumping into each other. It was so tiring for my brain and I kept sitting down whenever I managed to feel up a bench or a sofa.  

The next realization I had was that the directionality of perception spread out. Without the visual input coming from the front there was a different sense of three dimensionality, my brain was modeling the back of the body and the sides with the same amount of resources. This resulted in a slight sense of floatation.

We were taken to a fake supermarket trying to feel our way through shelves packed with stuff. I asked about technological aid and was told that some apps exist where you can take a picture and get information about it. Now I see that there is also a site where you can lend your eyes to a blind person. http://www.bemyeyes.org/. Too bad it’s just on iphone

Lately I’ve been thinking a lot about what happen when two sets of eyes meet and stare at each other.  It’s not just the knowledge that someone is watching you, it’s the knowledge that they are admitting to watching you. It’s like a ‘handshake’ protocol both sides saying “Yes, I see you”.   Supposedly blind people don’t have this, they don’t know when people are ‘looking’ at them and I was wandering if this gave them some extra freedom but my guide said she somehow did have a sense of when she was being looked at.

Finally we were led into the light and I felt how the power structure changed. In the dark we were dependent on our guide to call out to us so we would follow in the right direction and now suddenly we had access to much more information than her.

I was fortunate enough to attend a psychedelics seminar that presented the results of the first ever fMRI study that investigated what happens to the human brain on LSD. Read on to explore the science, my theories and as always the anthropological experience.

The conference

This was not a typical scientific seminar, there were almost no ‘professionals’ attending yet the conference hall was packed with around 150 people that were really excited, which is a rare case for university seminars. These people actually paid to support the study in what is possibly the largest crowd funding scientific experiment to date (the money side was head by WALACEA a site trying to democratize science).

The ‘Self’ network trying to predict the ‘meme’ machine we are made of.

I arrived early and used the time to conduct an informal survey. Most of the people were there because psychedelics had helped them or loved ones. Many of them had been following the media divulging very positive results regarding psychedelics experimental treatment for anything from alcoholism to the fear of death in those terminally ill. ‘My dad had terminal cancer and from what I read I wish he had access to psychedelic treatment’, one women told me and I found myself sharing with her and others my limited knowledge.
Psychedelics will reveal the scaffolding ‘you’ are really made of.  The various networks that are creating a meme machine that is constantly connecting “cause” and “effect”, trying to predict the sensory inputs, will be exposed. These networks know nothing about permanence, they are in constant change being turned on and off according to environmental influence. It is only the higher level “self “ that has created an inaccurate model of permanence. This ‘self’ is just a network that links the other networks together trying to predict (and thus influence) the behavior of the whole organism, as well as other organisms (if you are interested you can read more in my neuro-philosophy paper). Once this specific network is destabilized and reality is experienced without it, a familiar sensation takes over, we have all felt it as young children before we had a “self” and if we are perceptive enough we get a slight glimpse of it every night for a brief moment before we fall asleep. If experienced in the right environment there is only wonder and openness in this state of being.

Some Background

Amanda fielding from the Beckley foundation for consciousness and drug policy research, which has been supporting this research, told a tale from her personal history. When psychedelics were still legal, they had a massive impact on her life, and she began working on the connection between consciousness states and blood flow to the brain way before fMRI was everywhere.

Prof David Nut, used his exquisite British wit and love of Aldous Huxley to delved into the idiocy of policy makers when it comes to drug research.  “Orthodoxy is the diehard of the world of thought. It learns not, neither can it forget.” – he quoted Huxley hoping that after 50 years of spewing blatant lies policy makers will start looking at the science that shows the massive positive impact  psychedelics can have and indeed have had on most pre-Christian cultures.

The results in a nutshell
With great clarity and a tad of British shyness Dr Robin Carhart-Harris presented his experiment and preliminary results to the mostly no- scientific crowd. 20 healthy subject with previous experience with psychedelics (16 m 4 f) received 75 mcg of lsd intravenously (which gives a stronger affect that digesting it. This was chosen because of the extremely expensive price of 3000 pounds! for a dose of ‘legal’ LSD from a lab in Switzerland). They were scanned in an fMRI machine while resting and then also listened to music in an experiment designed by Mendel Kaelen.

The ‘normal’ brain on the left (few controlled high ways) and the brain on Psilocybin (many new roads open), a graphics based on fMRI functional connectivity data.

I’m going to start with a metaphor that wasn’t used and is a little wider than what can totally be proven but I’ll show some evidence to support it and suggest some more experiments that can be done to test it so bear with me. Imagine your brain is a city with many neighborhoods each of them with its distinctive characteristic and function in the city. There is the stressed financial district dealing with money, soho with its art and hipsters, the industrial area with all the factories and down town with its restaurants. In order for the city to function these different areas have to communicate with each other and pass information and goods between them. To do that, imagine these neighborhoods were connected by highways but not everyone could get on these high ways. There were tolls and  these highways were highly regulated by a central government that decided what has to go where. Then, imagine one day the toll company went on strike and these highways were closed down. What would happen?

At first each neighborhood would be less influenced and constrained by the others and could have more freedom in managing its activity. But for the city to still keep running some information would need to be shared and collaborated so the back roads would start getting used and new roads would start to be built, these roads wouldn’t be as regulated at the high ways so some soho artists could go to the financial district and start painting the buildings even if they weren’t asked to while some accountants and lawyers could go downtown to have a good time.

In this metaphor the people are the electrical brain activity, the neighbourhoods are the specialized networks that deal with specific types of information and the. The 5ht2a serotonin receptors in your brain sit at critical nodes constricting information flow, performing the function of a “Toll”. Lsd and psilocybin over stimulate these receptors and the ‘toll company’ goes on strike.

Carhart-Harris uses the term The Entropic Brain Hypothesis to outline his theory that claims that the brain in in a high entropic state with a higher disorder and more flexibility while on psychedelics.

How is this seen in brain imaging?

Passive eyes closed resting state fMRI scan
Psilocybin decreases BOLD signals in DFM and decouples them. Similar results found for LSD.
(R. L. Carhart-Harris et al., 2012; 2013)

It turns out that this ‘toll company’ is actually responsible for a lot of the brain activity, and the main results seen in fMRI for both LSD and psilocybin is less blood flow in these critical areas of the brain. Especially in the Default Mode Network that deals with ‘self-memories’ and according to Carhart fits well with Fraud’s notion of ego (and my idea of a self being a model that predicts the organism’s behavior)

MEG data showing decrease in brain wave activity (Muthukumaraswamy et al., 2013)

Now when you look at what’s left of the blood flow you see that the correlations between areas grow representing this higher degree of connectivity and when looking as close by areas the blood flow becomes less homogeneous.

MEG results for LSD weren’t presented yet, but a previous paper (Muthukumaraswamy et al., 2013) showed that decrease in activity of brain waves each, bandwidth decreasing in a different area.

One  surprising result in which LSD was different from psilocybin was that LSD increased blood flow to the visual areas while psilocybin didn’t. Carhart postulated that this might be because of the different time line of the drugs, the psilocybin when administered intravenously takes affect very fast while LSD doesn’t and the timing of the scan might show a specific phase of the drug. I would be very interested in getting the exact details of the experiment and comparing them, to see if there were any other changes in the experiment design that could explain this, some environmental influence that would activate the visual area in the LSD experiment more than it did in the psilocybin one.

Predictive Coding and my dream thesis

If you’ve been following my latest blog post you’ve already heard of predictive coding. The predictive coding framework claims that in essence the brain is a prediction machine, it tries to predict it’s sensory input to form a model of the environment (including itself). Predictive coding explains that the cognitive system is ordered hierarchically, in levels (Bastos et al., 2012). For any pair of levels, the higher-level will have hypotheses predicting the bottom–up signals from lower-levels. If the predictions are good, the bottom–up signals will be ‘explained away’. Only discrepancies between the winning prediction and the bottom–up signal remain as ‘prediction error’. This framework explains that sensory information is processed probabilistically, with prior predictions and posterior inferences made based on Bayesian optimized probabilities.

Google deep dream.
10-30 stacked layers of artificial neurons. Allowing a layer to become over active will produce “hallucinations”.

Since psychedelics seem to change the flow of information represented in the brain this framework can be a great model to help explain what is actually going on. Within the predictive coding framework I would postulate that Priors coming from the Default Mode Network aren’t passed to other networks and become less important in explaining away the bottom up sensory information. This will cause an increase in prediction error in lower levels of the brain. To deal with this, these lower levels will have to update their model more frequently without the constraint from higher layer priors.  This can explain the visual hallucinations and the subjective experience of heightened senses, more vivid colours, sounds, taste and touch. The tree is perceived as greener because higher layer abstract prediction saying ‘yeah, it’s green, we’ve seen this a million times’ can’t reach the lower layer and turn off the incoming data. Furthermore subjective reports claim that hallucinations start out as just variations of a noisy signal, like clouds shifting and morphing, but after a few second of maintaining the same visual input a very vivid totally unrelated image might appear. This can be explained by accumulation of “prediction error” that finally makes it’s way through some back road and connect with some higher level network which suddenly explains away the prediction error with wrong higher level priors. This can also explain “trippy repetitive thoughts” because higher layer abstract networks can also be triggered in certain cases and hyper activated.

In other words Prediction error that gets into a network will be ‘overly explained” within that network which is released from it’s normal constrains  This is exactly what is seen in Google deep dreams artificial neuron network. Neural networks trained to discriminate between different images can generate images too by inverting them (Mahendran & Vedaldi, 2014) and letting  a layer enhance what it has ‘found’ causes very trippy images to be produced. Indeed, Cartheit mentioned that he believes something very similar is happening to the brain on psychedelics.

Forward vs backwards connections –
Granger causality
(Zheng & Colgin, 2015)

Psychedelic Treatments

Psychedelic treatment for addictions and post trauma are getting amazing result (see ref at the end).  My speculative theory is that traumatic memories or addictive behavior are usually stored in the ‘worst neighborhoods’ isolated from the rest of the brain activity but still have managed to get enough access to the “tolls” and send out their predictions to the rest of the brain, effecting the general mood and behavior.  While on psychedelics other networks that are being activated have an opportunity to reach that bad neighborhood and renovate it, add some color, good food or financial maintenance. This does work both ways however which is why the right “set and setting are crucial” and why Psychedelics should be treated with the utter most respect and care. There are many reports out the of a bad trip having long term and even permanent results on people.

My dream thesis

Recent papers have found that slow brain waves (alpha and beta) correspond with top down activity and expectations the brain creates while gamma activity corresponds to bottom up activity. They did this by using MEG data and performing a Granger causality analysis that can assess what signal is more likely to cause the other signal. I believe that comparing this type of analysis on the brains of people that have ingested psychedelics will allow us to characterize the different flow of information and possibly prove this theory. Specifically I would expect to see a lower Ganger causality in the Alpha bands coming from higher levels of the hierarchy to lower levels showing the decreased effect of the higher level priors. While within lower close by areas I would expect to see higher Granger causality in the Alpha band (if MEG is accurate enough to source locate close enough areas). And perhaps it would even be possible to locate areas that have a higher Gamma band Granger causality than usual because prediction error has been rerouted through them.

Short Summary 

Couldn’t resist being a fan and asked Carhart-Harris for a photo

The science of psychedelics is finally starting to develop after years of being blocked by political reasons. Understanding how these substances affect us has the potential not only to teach us about ourselves but also to help treat some of the 21st centuries biggest problems. We really shouldn’t let closed mindedness stop the progress of this field, so how about sharing these thoughts with some friends or colleague?

___________

Psilocybin showing success in treatment of addictions and OCD:

A collection of information processing units designed through millennia of evolutionary trial and error to effectively model the environment and simulate interactions between the environment and these units.

Effectively?

Receptive fields of V1 neurons. Combining these basic structures is the most effective way to encode complex natural images.

Yes, one buzz word would be ‘sparse coding’.  Only a minority of units are needed to encode a typical stimulus. This is well documented in the low level V1 visual system in the brain. The receptive fields (the stimulaus that causes the neurons to fire) of V1 neurons that create a basis for all images are optimized for ‘natural images’.

It seems plausible that the rest of the units work in the same manner. Take emotions for instance. Wouldn’t they just be ‘sparse coding’ efficiently encoding interactions with the environment?

Fear, love, hate, disgust…they are all just a basis set, optimized to encode ‘naturalistic’ interaction.

To what end?

To perform actions on the environment that will ensure survival.  Any complex pattern requires a local decrease in entropy in order to different between inside and out, between environment and organism. In life forms the whole metabolic system couples exothermic reactions, that is, energy producing processes with endothermic processed that need that energy. Most life on earth arises from the energy delivered from the sun’s photons exciting electrons to a higher level, or as the biologist Albert Szent Györgyi put it “Life is interposed between two energy levels of the electron” (Although some life depends on deep sea thermal outlets or volcanos).  The survival of life depends on slowing the downwards cascade of this energy; maintain this local minimization of entropy. Any interaction with the environment is very costly although some interaction is needed to obtain more energy/food. This is why the the end goal of these units is to simulate possible action outcomes with the environment and choose which of them will minimize the loss of entropy/maximize survive chances. Another buzz word in this context would be, minimizing ‘free-energy’, or surprisal which is the information theory equivalent of entropy. The free energy principle tries to explain how (biological) systems maintain their order (non-equilibrium steady-state) by restricting themselves to a limited number of states. So if you are one of those who is constantly searching for a ‘goal’, minimization of entropy/free-energy seems to be as good as it gets. Striving for complex knowledge structures and creating just about anything and minimizing waste fit well with this idea but so does bringing more life into this world and cleaning my room.

Is that all that you are?

Me chronologically

me chronologically inverted

me, least information on top

‘I’ am a model within a model, projected out into the world allowing others with a similar algorithm to effectively model me. Or perhaps one could turn the tables. The effective modeling others conduct when encountering ‘me’ is mirrored back into these units as a simplified singular version of itself.

One could postulate that there are as many ‘I’ as the people who know me, although they too are probably composed of an effective sparse basis.

For more reading:

Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36(03), 181–204. http://doi.org/10.1017/S0140525X12000477

Cook, N. D., Carvalho, G. B., & Damasio, A. (2014). From membrane excitability to metazoan psychology. Trends in Neurosciences, 37(12), 698–705. http://doi.org/10.1016/j.tins.2014.07.011

Friston, K., & Kiebel, S. (2009). Predictive coding under the free-energy principle. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 364(1521), 1211–1221. http://doi.org/10.1098/rstb.2008.0300

Friston, K., Thornton, C., & Clark, A. (2012). Free-energy minimization and the dark-room problem. Frontiers in Psychology, 3(MAY), 1–7. http://doi.org/10.3389/fpsyg.2012.00130

Olshausen, B., & Field, D. (1996). Emergence of simple-cell receptive field properties by learning a sparse code for natural images Nature, 381(6583), 607-609 DOI: 10.1038/381607a0