Learning neuroscience, machine learning, programming, and social science without a degree

Fanli (Christian) Ramsey

Lead AI Engineer

发布日期: 2019年8月27日

We live in an age where most of our current knowledge is accessible and freely available. We no longer have to limit ourselves to waiting for a specific course to be available or for a professor to recommend a textbook or cutting edge research paper. We can learn just about anything we want and go deeper and further across disciplines than most degrees would ever allow all without incurring massive debt.

10 years ago I decided that there was enough content online and off to fulfil my own knowledge acquisition goals and it would allow me the freedom to study what I wanted as deeply as I want. 500+ nonfiction books, a seemingly endless amount of free courses and lectures later, I realised that this is available to anyone who has access to the internet. Today we call such a person an autodidact but in the future, we might just call it “modern learning” ( I talk about this in my deep learning and education paper).

autodidact — If you’re an autodidact you’ve done most of your learning on your own, outside of school. “Having learned Greek and Latin, as well as landscape painting and auto repair, without any formal training makes you quite the autodidact.” — Dictionary.com

In this article, I’ve put together a curated collection of my last 10 years of study to help others who might also be on this path or just interested in a specific subject. My deeper study subjects are neuroscience, machine learning and artificial intelligence, anthropology, psychology, design, biology, ecology, programming, mathematics, probability, language and philosophy.

Let us begin. For each subject, I’ll provide you with an overview, why study it, and a list of curated books, courses, and people with links that have inspired and informed me in that specific area of study. I’ve tried to pretend that every course, person, and the book didn’t cut across all the subjects to simplify but let’s keep in mind that everything is fuzzier than our language and concepts would have them appear.

Neuroscience

I studied neuroscience because I wanted to know the biological and neural substrates of why so many of us suffer from symptoms of mental disorders and the physical basis of stress, emotions, and decision making in everyday life. Grounding or looking underneath psychological theories which focus more on behaviour and mental constructs, I wanted to know how it worked at the neuronal and hormonal level. I was introduced to neuroscience by a book titled Embodied Cognition by Varela et al and then from there went on to read the highly acclaimed book by neuroscientist Antonio Damasio titled Descartes Error. Damasio was the first to help me understand the relationship between emotion and cognition. He described, in great detail, his own and others neuroscience experiments showing the limbic system’s complex relationship to the prefrontal cortex. Most notable were studies of patients with damaged or removed brain regions ( see frontal lobotomy ) and how they systematically affect behaviour, especially in making certain decisions. Most memorable was the somatic marker hypothesis where when one reasons, they look for an emotional marker to discriminate between choices. That marker is experienced as a feeling, not simply a purely reasoned calculation. This connected the prefrontal cortex, specifically the ventral medial prefrontal cortex (VMPFC), with the limbic system. After giving this book to a friend, they became slightly frightened by the implications that if one simply prunes the connection between the VMPFC and limbic system that the so-called personality of the person is altered and likely irrevocably. It was clear to me that the brain and the mind must be learned deeply in order to make sense of ‘me’ and ‘others’. I then focused on psychophysiology alongside neuroscience papers because I wanted to understand the neurophysiological substrates of stress and if it could be continuously measured. I studied electrodermal activity and psychophysiology via textbook and then decided I was ready to purchase the classic neuroscience textbook, Principles of Neural Science. It seems to be nearly infinitely useful. Whenever I want to know how something works in the brain, I turn to this classic. From there I rediscovered Robert Sapolsky. He deserves a Nobel prize for his teaching ability; he has impacted literally 100,000s of online learners (if not millions) and has shed a convincing neurological light on depression and anxiety disorders. I read and watched everything (literally) on the web I could find from Sapolsky. His recorded Human Behaviour Biology lectures on youtube are the gold standard for foundational for understanding human behaviour. He is both a primatologist and neuroscientist, which suited me well as I was transferring from the world of primate and human cultural anthropology into neuroscience. At the same time, I started directing my studies towards emotion and cognition, which is also called affective neuroscience. I studied the work of Ledoux (fear), Damasio (emotion and cognition), and Panksepp (mapping affective states in rats) and others, each of whom have contributed groundbreaking work to the science of emotions, affective neuroscience. Later, I became concerned with how neural action drives behaviour along with how the nervous system interacted with the endocrine system, specifically hormones. It seems the difference between a neurotransmitter and a hormone has a lot to do with the medium; whether it’s in the brain or the blood and this makes a vast difference in terms of how they unfold over time. When I was trying to put all of my knowledge together, I found Sapolsky to have one of the most effective frameworks to think about what produces any given behaviour. A framework is almost always an oversimplification but I’ve found it useful in my personal and work projects. Sapolsky provided us with a temporal framework to study and reason about any given action or better put by him, why did this behaviour occur?

what happened in the brain one second before the behaviour?
what sensory inputs entered just seconds before?
what hormones were released into the system hours to days before?
what in the environment changed weeks to years before?
what was childhood like for the individual?
what about their fetal environment and genetic makeup?
what cultural factors shaped the behaviour?
What were the ecological factors that shaped the behaviour?
What evolutionary factors shaped the behaviour x millennia before?

“when you explain a behaviour with one or any of the disciplines, you are implicilty invoking all the disciplines”

Robert Sapolsky

Sapolsky’s framework goes well beyond neuroscience and really provided a way for me to begin studying more topics in depth like evolutionary psychology, ecology, and genetics as well as a latticework to integrate my knowledge across subjects. I encourage you to consider Sapolsky’s framework when trying to understand animal behaviour. It is never just x gene or because these neurons happened to fire. It’s always multifactorial and interdependent and very unlikely to be as simple as some of us would like it to be.

Soon after this, one of my mentors Jayant, a practising neuroscientist, gave me a great book titled, Neuroanatomy through Clinical Cases which is just as fascinating as it is useful. The book has helped me understand how physicians diagnose and understand cases of the mind. This opened my eyes to the real world of neuroscience rather than just the seemingly precise concepts you read about in textbooks. In the real world, everything is exponentially messier. So inspired by this book, I then set a goal to work in the neuroscience field and am now a collaborator at a neuroscience lab where I get to see the data up close and apply my machine learning skills to real-world problems concerning the neuroscience of emotions and mental health disorders.

Why Study It:

Neuroscience grounds behavioural science in the brain and some would go further and say it grounds it in the mind. Neuroscience allows you to go beyond observing animal behaviour and towards the neural correlates that often happen before, during, and after a behaviour occurs. It should be noted, that I have mostly studied mostly animal behavioural neuroscience, neuroscience is vast and doesn’t just deal with the question of why did this behaviour occur; there are people who have been titled and title themselves neuroscientists that study the chemical and electric activity in the brain and those on the more computational side who try to simulate networks of neurons to better model neural activity without considering a given animal behaviour. There are numerous studies on almost any question you might have about life that neuroscience has attempted to shed light on. What makes up human or primate intelligence? Why are humans aggressive? Are emotions real? How do we recall the “past” from our memory? Why am I so anxious? How do we process morally bad acts? Many of these questions cannot be answered directly but neuroscience can shed light on how people, on average, process emotions or recall a memory or which brain regions are active during a morality test. But you should be warned, neuroscience will not give you “the” answer to any particular question other than, it’s complicated and these brain regions were more active than others across time. This is substantial, but like most behavioural research, it is correlative and the sample sizes are often quite small. A few neurotransmitters exciting or inhibiting a given neuron may be simple to model but when you scale that toy model to the ~ 86 billion neurons we have and the neuroevolutionary architecture we’ve inherited, this system goes from classical linear system to a non-linear and non-deterministic system fast. What you see one day in a group of study participants you may not see in the next set. Neuroscience does give you a view into what is happening at the neuronal level, and there are people who are “transcending” coarse brain region (i.e. limbic system, prefrontal cortex) for x behaviour or mental disorder research towards studying “populations of neurons” which may or not cross classical regional boundaries. There are also people, who at times, utilise neuroscience to prove their point using the language of determinism and causality which I view as wrong-headed. Some go further to tell you what is happening in your head when you do action x. But unfortunately, although there are similarities, our brains are incredibly plastic and flexible, which means we can use different networks in very different ways at the individual level. Many of us who feel disgusted at a person’s behaviour, might be activating our olfactory bulb and insula as if that person actually smelled or tasted bad, but we might also be processing this within our prefrontal regions.

One to one is an exception or fluke. It’s almost never the norm.

If you read biomedical engineer Alan Janasoff’s book, A Biological Mind, you’ll likely help yourself develop a healthy ‘neuroscepticism”. I am also convinced that we have put too much collective epistemological and even metaphysical weight on neuroscience ( or rather our projections unto neuroscience research) and this makes us suspectable to sounding intelligent but being wrong or convincing others of ‘facts’ that are wrong. Jasanoff gives a refreshing account of the neuroscience literature and its inherent limitations. He describes how people are easily manipulated into finding a proposition more credible and believable when there is an image of the brain next to it (fMRI). He describes the different techniques being used (i.e. transcranial magnetic ) and some of their limitations. The conclusion is that if you go into neuroscience believing that there is a brain region for supposedly special human construct x and a neurotransmitter for special human trait z, you will find yourself not only out of line with the literature but likely teetering on the edge of dogma. The brain has a basic structure and it doesn’t appear there is anything “extra” special about human brains that cannot be explained away by the sheer number of neurons it has at its disposal.

The deeper you dig into any subject the more propositional truths begin to dissipate into fuzzy regions of truth and non-truth that can’t be captured exactly by any single cause or, in this case, brain region.

Even though some of the popular media looks to neuroscience as the latest non-theistic religion, all is not lost. There is much you can learn if you have a discerning eye and curious mind towards true understanding. Learn neuroscience if you want to understand yourself and others at a deeper level.

Anthropology

Why study it?

Anthropology was one of the first subjects I dived deep into using books, open courseware, and to my luck, two talented and well-versed anthropologists who volunteered to mentor me along the way.

Anthropology is defined as the study of humankind. This is both narrow and broad. It’s narrow because it focuses on one “species” but broad because it encompasses any and every interaction (rather implicitly or explicitly social). This means you can study any phenomenon that exists at the level that we see our world. What I think is good about studying anthropology is that it gives you the ability to study human behaviour from a social perspective using a toolkit called ethnography that allows you to make sense of how people behave in terms of the structural forces like institutions or the neighbourhood we live shape the behaviour of groups as well as how individual agents act on the system and their social group. I was disenchanted with anthropology after doing a two-year-long ethnography that I’m yet to release, mainly my disenchantment came from the fact that I wanted anthropology to be behavioural science and it’s not until much later that I see the importance of explanatory knowledge and writing a thick description of people in action.

Study anthropology to de-centre yourself from your traditional view of the world as solid and whole. Anthropology can help you see the holes in your view. If you are interested in the post-modern trend occurring, study anthropology because it gives you a very interesting view and the ethnographic fieldwork will open new ways of considering human activity. I’ve read ethnographies on gift-giving, homelessness, finance, capitalism, sex, gender, and sports. Warning, there is a group of people within the social sciences that view the idea of sociological reflexivity as knowing where you are coming from. I fundamentally think this is wrong-headed and leads to researchers collapsing themselves into the 3 or 4 most trendy dimensions ( race, gender, sex, age) rather than taking the view that those dimensions are not ever-present, solid, and not always a part of the stage of phenomenon. Gibson, Heidegger, Mearlau-Ponty do a great job countering this by showing that all of our dimensions are not points, they themselves also rise and fall depending on the “situ”. This is much better than the very cheap version of reflexivity that I see being employed, but not everyone sees it as an upgrade.

Books

Analysing Social Settings by John Lofland
Presentation of Self by Erving Goffman
Imagined Communities by Benedict Anderson
Suicide by Durkheim
It’s Complicated by danah boyd
Computational Social Science by Michael Alvarez

Courses

Anthropology by Oxford University
Biological Anthropology: An Evolutionary Perspective by Barbara J. King

Biology

Why Study It

I didn’t study biology well in school but after immersing myself in cognitive science and pushing into neuroscience, it became clear that I needed a deeper understanding of how organic “life” works. Learn biology because you’ll give yourself a lens to see the world that reveals it’s underlying structure. It’s interdependence and the sheer amount of complexity. Cell biology gives you the basic building blocks of how life “works” with some of the most basic units we’re all made up of. The fact that it looks more and more like a more or less prokaryotic cell made a leap to the eukaryotic through an aggressive symbiosis (endosymbiosis) with what we now call mitochondria and that this transition may be the foundation for all of the plant, fungi, and animal life today is not only fascinating but it is practically useful in unhinging our obsession with human exceptionalism and pretending we aren’t made up of entirely non-human things. Our ancestors are weird, not because they actually are weird but because our level of collective self-deception doesn’t match the underlying reality. So instead we pretend that we’ve always been human.

Evolutionary biology, which Darwin is viewed as a pioneer of the field, helped many people upgrade their cognitive toolkit and has since evolved into a robust modern evolutionary theory providing us with a way to reason about the past. It cuts away at the idea that genetics and natural selection resulted in a nicely pruned tree of life that we can traverse with ease and intuitively. The reality is that it’s difficult to draw any “species” line or any interspecies line, there don’t seem to be any pure “essences” and more about a multifactorial combination between natural selection, genetic drift, ecology, culture, and random mutations where the variance within a group of a so-called population is typically the same or greater than the variance between such populations. So fairly new terms that we hang on to, especially in identity politics and some social sciences, are quite arbitrary and it is delusional to deeply identify with these categories. This includes race, gender, and so on. If after deep study and contemplation in the biological sciences, we should begin to see that the lines we draw harshly are more conceptual than they are “real” and yet we suffer because of our deep identification in them. Study biology because whatever you learned in high school was not enough to get you beyond reproduction and basic natural selection. If we take biology seriously enough (by reading broadly and deeply), we will have a new way to reason about ours and others best and worst as well as that the world is fundamentally interconnected. Biology tells us that we are the same and that we are different. It is helpful in deconstructing our view of others as having intrinsic negative qualities and ourselves as having advantageous qualities that are made up of something beyond basic cell biology. It isn’t, this is likely your ego at work.

Books:

Life Ascending by Nick Lane
Origin of Species by Charles Darwin
The Tangled Tree by David Quammen
Power Sex and Suicide by Nick Lane
Symbiotic Planet by Lynn Margolis
Human Biology: An Evolutionary and Biocultural Perspective by Douglas E. Crews

Courses:

Introduction to Biology
Biological Anthropology: Evolutionary Perspective
What Darwin Didn’t Know — The Modern Science of Evolution

People

Nick Lane
Lynn Margolis
Brett Weinstein
Louis Pasteur
Frederick Griffith

Computer Science and Programming

I started programming towards the end of high school. Long story short, I needed a way to make money to fund my self-education and building websites was my way of solving the issue. So after using WYSIWIG editors for some time I became frustrated at the impossibility of creating or replicating websites I found to be good using these editors. Then shortly after, overly confident, I tried to build a social network, for a whole city. After using many off the shelf components, I found myself learning PHP, SQL, HTML/CSS and Javascript to customise the website to fit user requests. There was an overwhelming pressure to learn it because the site had amassed more than 3,000 people on it and the site was constantly crashing during peak load times. So I called my server host and found out they couldn’t be of help on my almost free plan and suggested I use larger servers but since I couldn’t afford it I tried to use my limited programming skills to improve the performance. After that, I decided to really dedicate myself to programming. I had a friend who taught me Java after work and gave me toy problems to work on which was very helpful. I eventually landed on python as I shifted from front end development to back-end development.

When I landed my first “real” job, they told me that I needed to learn python if I wanted to try and improve their software modules so I learned just enough to figure out website programming or scripting and software programming can be quite different. I wouldn’t really learn python until after the job. I truly fell in love with programming. I loved the blank script file and how anyone could get started instructing a computer with a few lines of code. Coding became addictive for me after a while and I used it to automate parts of my life such as social media posting, summary generation, and updating websites. I didn’t know it when I first learned it, but programming is now seen as essential for future work. I’m not so easily convinced but I do believe it has been worth all of the long nights of reading and debugging for countless hours.

But what is more exciting than coding? Computer science while coding. I avoided computer science for years and finally gave in when I discovered MIT’s CS101 and their Discrete Mathematics course. Many of the issues I was running into in my roles as a software developer now made sense after learning data structures and algorithms. Why some of my code ran slowly, how timing my code seemed like a fool’s errand. How knowing the levels of complexity (i.e. On ) of a given piece of code can provide the necessary abstractions needed to reason about how efficient a piece of code is. So instead of programming to just get all of your spaghetti code eventually working, you can begin to program by using the many computer science abstractions and paradigms that have been tested and validated much more thoroughly than the clever spaghetti code you are planning to write.

I also learned about computability. I learned about Turing machines and information theory. I will talk more about this in the machine learning and mathematics section.

Why Learn It

Because computing empowers individuals to create at a level that is unprecedented in the analogue world ( in terms of scale and speed from thought to manifestation ). Computer science provides you with the level of abstractions needed to apply working architectures, paradigms, data structures, and tests to evaluate your code and improve it to make it more useful, reusable, and efficient. If you already know how to program but don’t know computer science, I assure you, computer science is instantly applicable to the code you are writing right now. So learn it deeply, I’m in fact, still on a mission to learn it more thoroughly in order to add discipline and intelligence to my 8+ years of programming. If you don’t learn any computer science, programming is still worth learning. Learning to program with a basic understanding of logic will help you improve your ability to move from idea to implementation in the software and app world. It will help you think more analytically and improve your mental capacity for thinking logically and improve your long and short term memory if practised well.

Books

Python Programming: An Introduction to Computer Science by John Zelle
Head First Design Patterns by Kathy Sierra
Head First Python by Paul Barry
The Pragmatic Programmer by Andrew Hunt and David Thomas

Courses

Introduction to Computer Science and Programming
Mathematics for Computer Science
Design and Analysis of Algorithms
Introduction to Computational Thinking and Data Science

People

Tom Leighton
Eric Grimson
John Guttag

Mathematics and Probability

I discovered quickly that my mathematics skills were underdeveloped as soon as I started to dig deeper into data and computational social science. I could not speak the language in the slightest. So after trying to read E.T. Jaynes Logic of Probability, I decided I needed to go all the way back to Algebra. Then from there, as Jaynes recommended, I found myself studying boolean algebra. Which lead me back to logic. I read George Boole, some Leibniz, and Spinik as I went from logic to calculus. This is where MIT OCW seemed to really shine. After learning enough boolean algebra I began studying calculus along with the history of the use of mathematics in artificial intelligence. This gave me a patchwork path to chart through mathematics. I took Tom Leighton’s wonderful course in Discrete Mathematics. I learned so much about how to think and prove things using mathematics. I did many proofs, mostly proof by induction, and learned how computers can help prove this much faster than you can. Mathematics had never been this exciting for me. I learned about beautiful mathematics like Turing machines, Cantor’s infinities, Godel’s incompleteness theorem, and many other mathematical treasures. But more than mathematics, probability appealed to my ‘fundamentally ambiguous’ theory of the world and so this is where I ended up spending most of my time and still do. Learning about Komologrov’s axioms, Jaynes extension of logic to plausibility, Bayesian credibility, Shannon entropy, the laws of probability, random variables, likelihood functions, distributions, fractality, limits and bounds, optimisation and the like, pushed me towards a new understanding of the word. Towards the idea that we live in a probabilistic universe, where nothing is certain, yet we have a well-developed toolkit to reason under such uncertainty. All of this content excited me and I can still feel a bit of the rush I felt when I was just started to learn it. Eventually, I found myself learning the basics of complexity science and following the TensorFlow probability thread.

Why Learn It?

Learn it because this way of knowing about the world is fundamental to our understanding of reality. A phenomenon occurs, and it recurs. These patterns can be modelled mathematically. Sometimes deterministically. Sometimes things go chaotic. These events are not just edge cases we can leave out of our data or model; in many cases, classical mechanics works only when conditions are stable enough to support its axioms but you can be sure that every system will eventually violate these assumptions and the only question, which is not as simple as it seems (i.e. 2008 financial crisis), is when. If one understands the language of mathematics and how probabilistic systems tend to work, one can gain intuition about the world through simulation and modelling. But in order to simulate and model well, you need the underlying mathematics, probability, and statistics to understand and tinker with your assumptions in code or equations. It can be a means to an end, or the end itself if it becomes your fancy. I recommend studying it deeply even if you see it as intrinsically evil. Start from scratch if you have to, all the way back to addition.

Books

Probability Theory The Logic of Science by E.T. Jaynes
Doing Bayesian Data Analysis by John Kruschke
Introduction to Probability by John Tsitsiklis
Boolean Algebra by Paul Halmas
Introduction to Logic by Henry J. Gensler

Courses

Probabilistic Systems Analysis and Applied Probability
Coding the Matrix
Fundamentals of Probability
Khan Academy Probability and Statistics

People

John Tsitsiklis
Sal Khan
Komologrov
E.T. Jaynes
Claude Shannon
Kurt Godel
Alan Turing
Gottfried Wilhelm Leibniz
Alfred North Whitehead
Georg Cantor
George Boole

Machine Learning and Deep Learning

I dove into machine learning after learning statistics for social science and reading up on complexity science (especially social simulations). For Machine Learning I have professors who were then at MIT and Stanford to thank for open sourcing their courses as well as Bengio et al’s Deep Learning book. I started with CSS (computational social science), quickly finding out I needed a deep understanding of mathematics and probability to grasp the basics of machine learning. I then went on to read books in parallel like Elements of Statistical Learning and papers from the likes of Vapnik, Bengio, Minksky and so on. I became quite obsessed with Bengio’s idea of distributed representation and how deep learning intersected with neuroscience so I jumped in. To solidify my learning, my partner and I created an 8+ hour course titled Applied Deep Learning with TensorFlow which dives deep into deep learning and neural networks using the most popular library for machine intelligence.

Machine learning is an incredible field that is very approachable if you have the basics of mathematics. You’ll give yourself access to a world of open access machine learning papers and an extremely helpful community of practitioners hungry to learn more. I started more from the statistical learning side than the software engineering side of machine learning so I was quite attracted to statistical learning so I started reading the classic Introduction to Statistical Learning by T and A [fix]. From there, while learning and doing featuring engineering and parameter tweaking for support vector machines I discovered neural networks. I applied them without really knowing what they were doing fundamentally to outperform SVMs. That’s when I discovered the very fragile and early version of TensorFlow. I was quickly able to apply TensorFlow and machine learning techniques to problems I was interested in. My interests were to discover how one could take psychological constructs or better “heaps” like love and teach it to a machine in interesting ways. So I used TensorFlow and therefore deep neural networks to create word and sentence embeddings using the latest recurrent neural networks and integrated this approach into my research process. From there, I started getting into data annotation and labelling migrating psychological constructs and measures into my data annotation steps. I eventually created a new method called Social Signals which combines computer vision (CV), natural language processing (NLP), and psychometric research to gain a deeper understanding of human behaviour in the world based on online traces.

I also went on to combine my learning in machine learning and deep learning into my first paper (here) a couple of months after the release of my 8 + hour course on applied deep learning and TensorFlow (here).

Why Learn It

Learn machine learning because you’ll understand your world better. Your world is full of recommendations from companies who have an enormous amount of data about you and others which is fed into a machine learning algorithm, often a deep neural network, in order to “predict” what you would be willing to purchase or consume next. So learning how machine learning works will shed some light on why you are likely getting the recommendations you are getting.

Learn machine learning to investigate the human condition. When you get deeper into machine learning, you’ll be faced with how to deal with human bias and that is also an opportunity to understand humans through machine learning. Encoding their heuristics and moment to moment opinions in order to understand decision making. Machine learning can be a powerful tool to inquire into the collective psyche of people around the world.

Learn machine learning to design better products. I say this with a large disclaimer. Better is very difficult to define. In fact, it’s highly likely you’ll build something that never makes it into “society”, and even if you do, it’s more likely you’ll provide society with yet another distraction that you‘ve convinced a growing group of people that it’s for the greater good. Machine learning, like any algorithm, allows for amplification. But every now and then, people make something that makes the world a bit better. Machine learning is typically used to enhance a products ability to keep you using the product, so I would encourage you to design intelligent products, that are intelligent not because they know how to manipulate users in ways that are inconsistent with their goals (i.e. tempt you by way of recommendation to binge-watch that show that you told yourself to stay away from in order to get something more meaningful done ), but instead let it be because you helped people reach a worthwhile goal that would leave them better as well. I could go on.

Books

Deep Learning by Ian Goodfellow, Yoshua Bengio, and Aaron Courville
Elements of Statistical Learning by Trevor Hastie, Robert Tibshirani, Jerome H. Friedman
Information Theory, Inference, and Learning by David J. C. MacKay
Probability Theory: The Logic of Science by Edwin Thompson Jaynes

Courses

Probabilistic Systems Analysis and Applied Probability
Statistical Learning Theory
Artificial Intelligence ( by the late Patric Winston)
Applied Deep Learning with TensorFlow (by me)
Deep Learning.ai (Andrew Ng)

People

Andrew Ng
Trevor Hastie
Robert Tibshirani
Vladimir Vapnik
Patrick Winston
Alan Turing
Ian Goodfellow
Yoshua Bengio
Aaron Courville
Fran?ois Chollet
Shakir Mohamed
Warren McCulloch
Walter Pitts
David Rumelhart

Psychology

My self-led knowledge acquisition journey started with 2 books I found when I was younger (maybe 7–12 years old, foster/orphan kid memory issues). One book was about electronics which lead me to break apart a dozen cd players and toy cars to take the motors out to look at the coils and magnets and get it running by myself. The other book was called Abnormal Psychology. I haven’t a clue who owned it but it was a college-level textbook that I read either because I felt deeply bored or it seemed interesting. My interest in psychology reemerged again when I was studying celebrity culture and considered different ways of seeing masculinity and so I eventually arrived at cognitive psychology and the popular neuroscience of embodied cognition against the background of Cartesian Dualism. Eventually, I would find the venerable William James who is considered one of the fathers of modern psychology, starting with a grand book, The Principles of Psychology. This book and my reading of it, has simply been life-altering. It may have been written 100+ years ago but it still holds fundamental principles for studying the mind and behaviour. You also begin to see James’ philosophy seeping out throughout the text which is a bonus. James introduced me to a rigorous and more pluralistic view of animal behaviour and atop that, he is also an astounding role model. James cared about psychological theories as much as he cared about the involvement of the nervous system in such processes. From there I branched into mental disorders, where knowledge of biology is useful since many disorders have to do with neurotransmitter and hormonal changes in an individual (serotonin levels), to understand myself and how and why they exist. The more I studied the technical or analytical side of psychology, the more I knew I needed a deeper understanding of the biology of the brain and the endocrine system, so be prepared to learn some biology in parallel. Once I started studying mental disorders like OCD, I became interested in the therapy and medicine that can help regulate or even cure these disorders, so I studied everything from Jungian archetypes, behaviorism, cognitive behavioural therapy, to Buddhist psychology. One of the most surprising areas of study for me, since I am not religious as such, outside of my core study areas (cognition, therapy, disorders, and perception), was Buddhist psychology. In recent years, there has been a significant number of researchers focused on Buddhist approaches to psychology that have been subsequently tested and provide some early credibility for the approach. The main area being stress reduction techniques. Psychology has helped me in the design world as well as in my personal relationships by giving the ability to converse and identify patterns of communication, attachment styles, cognitive bias, social interactions, and human perception.

Why Study It:

Learning psychology will help you explain the world in terms of the individual with concepts and descriptions that are, on average, better than our day to day chatter and gossip about people in person or on social media. If you want to upgrade how you think about people you like and dislike, psychology can give you a nuanced view of the world that provides a convincing and sometimes empirically backed view of why people behave in certain ways. If you are studying to better understand and treat yourself then the therapy side would be useful. I recommend exploring everything from Freudian and Jungian psychotherapy, cognitive behavioural therapy (CBT and DBT), growth mindset training, and mindfulness stress reduction therapy as evidence-backed therapy from some of the most successful methods being used in the field. Psychology is vast and seemingly endless, you can get lost in all sorts of psychology, but the main point is to bring the subject a question and learn to refine it. Psychology is currently the best way to understand individuals in your social network. It provides us with better theories than the ones we usually generate about ourselves and others. But this can also be dangerous if applied frivolously. So it’s best to go deep so you don’t find yourself using psychological theories and bias to satisfy your ego that there is something wrong with person x (i.e. she definitely has a narcissistic personality disorder).

**Warning: knowing concepts in psychology doesn’t automatically prove you know how to classify people into them.

Books

The Embodied Mind by By Francisco J. Varela, Evan Thompson, Eleanor Rosch
The Principles of Psychology by William James
Computational Social Psychology by Robin R. Vallacher, Stephen J. Read, and Andrzej Nowak
Designing with the Mind in Mind by Jeff Johnson
The Ecological Approach to Visual Perception by James J. Gibson
Handbook of Psychophysiology by John T. Cacioppo and Gary Berntson
Mindfulness and Buddhist-Derived Approaches in Mental Health and Addiction by Shonin, Edo, Gordon, William Van, Griffiths, Mark
Phenomenology of Perception by Maurice Merleau-Ponty
How Emotions are Made by Lisa Feldman-Barret

People

William James
Maurice Merleau-Pointy
Lisa Feldman-Barret
James J. Gibson
Robin Vallacher

Design Thinking

When I found design, I was searching naively for a way to change the world. I had been designing apps and websites but none of this was fundamentally tied to solving problems in the world. Design thinking allowed me to take a completely new approach to generate solutions for hairy issues (i.e. poverty). Design thinking is intrinsically optimistic and places people and their stories at the centre of its production. I was invited to become a Stanford d. School design thinking trainee and after my training there, I began creating startups using the methodology. My LinkedIn doesn’t contain much in terms of formal institutions but my experience at the d. School was so life-changing that I simply had to give those at Stanford credit. I now work at a design thinking firm, where I work as a data scientist focused on creating human and ecologically centred artificial intelligence.

Why Study It:

Because the world is utterly complex and therefore we need complex methods to make sense of the world in order to design products with the goal of alleviating human and non-human suffering. The goal of design as I see it is to improve human and non-human flourishing. Design thinking concepts such as convergence and divergence or abductive and inductive allow designers to tackle hairy issues where the end isn’t already insight and the solutions aren’t obvious. Design thinking is a toolkit and philosophy that you’ll be able to use in every aspect of your life. I used design thinking to design my knowledge training programmes, my language learning method, my mindful productivity system. Design enables you to take the creativity you already have and take a disciplined approach to solving a problem by first understanding people. By spending deep time with them, then synthesising the stories and themes learned in the field into specific needs, themes, and insights, then given those insights we then begin (kai shi) to prototype solutions that we believe may alleviate their issue. It’s an embodied process. Also, you can bring all of your tools and crafts into design thinking. I have been developing machine learning methods that can be applied throughout the design process.

Books

About Face by Alan Cooper, Robert Reimann, David Cronin, Christopher Noessel
10 Faces of Innovation by Tom Kelley
Design of Business by Roger Martin

Courses

Human-centred Design
IDEO U ( design thinking )
d. school boot camp
Design Thinking and Innovation

People

Tom and David Kelley
Roger Martin
Tim Brown
All of my co-employees

Ecology

Ecology is interdependence. It is the result of discovering or rediscovering how everything is in fact connected at each level of organisation and an attempt to describe the total relations and interactions within an abstracted defined ecological unit. Our psychologically tilted society has lost the ability to see our deep entanglement, ecology helps loosen our delusion of the world as centred around ourselves or even our species. As we now know, human activity has had some disastrous effects. It is obvious that our mantra of innovation and technological advancement are not aligned with human and non-human flourishing as much as we imagined it would be. People are hooked on processed foods and stuck scrolling through infinite timelines. Ecology helps us see this from a different angle.

I started studying ecology when I was very young as I was influenced having spent much of my time on a farm where I took care of livestock and the back and front garden. I somehow became interested in soil and started learning about how soil works to help plants grow and that is when I discovered that topsoil was being ruined by human activity. Topsoil, as I had discovered, is extremely important, the first three feet of soil we grow our food from is vital to the plant surviving because of the complex array of microorganisms in topsoil and it’s abundant nutrients that plants use to grow. In other words, topsoil is alive. This fascinated me and scared me, what were we doing as “humans” to create such damage? Unfortunately, I wouldn’t formally come back to ecology until late last year because I found myself more and more interested and concerned about how human activity has and still is playing a huge role in destroying the earth. From my own studies of our activity, there is a decreasing tolerance for anything that slips over the “nature” vs “us” line. Our buildings are often nearly bugless, odourless, and our modern view of what is deemed sanitary is a subtle but still apparent form of genocide (bugs, odour molecules, bacteria, “pests”) to everything that makes us feel insecure and therefore fearful. We want to transcend “nature” so badly but it isn’t possible since we are in fact nature. It is only our perception that convinces us that we are somehow different. What Timothy Morton calls “The Severing”. Our unrelenting progression down this path is now showing it’s distinct consequences. The current human project has mostly succeeded in trying to bring the myth that we are separate from nature to life and it’s in vibrant form. Just look around. At the same time, this means that we have to attack ourselves at least the parts that most resemble nature at this moment in time, just look at capitalism’s effect on agriculture as an algorithm tuned for human experience, to study these complex relationships I went back to ecology. It was already clear to me that we have helped create and are also victims of a marketplace which thrives on human ignorance. We use very basic neurobiological bugs (there is much neurobiological evidence to support this) that can be defined by a handful of neural pathways that any algorithm wanting people to buy more widgets would target for optimisation (inadvertently or advertently). It should be clear that “markets” have found it. Unfortunately, it seems to have nothing to do with what people actually need to reach a state of individual collective wisdom. Ecology provides a way to understand this relationship from the perspective of an organism’s relationship to another organism, in this case, symbiosis. Ecology also gives us different levels of organisation to view such relationships; communities, populations, and then there is the all-encompassing biosphere. So we can observe certain relationships and classify them loosely into ecological units, which include mutualism, parasitism, commensalism, and symbiosis. This gives us a lens to understand our relationship with technology and algorithms ( I speak more about that here ) and we can use the ecology literature to discover how certain organisms shifted between these relationships in order to shed light on how we might shift our relationship to non-humans and humans alike. Ecology textbooks are dated, so many of them do not touch on modern ecological units like the internet, phones, apps, and the algorithms that run them, although I’m sure it is being worked on. If you simply observe your relationship with your phone or a given social media app, you’ll discover a very strange relationship indeed. It isn’t clear who is eating who or who is nourishing whom. The algorithms behind apps and social media sites, do not make it immediately clear that they won’t always be optimising for some dimension that bleeds into parasitism rather than mutualism since constructing an algorithmic goal of human flourishing that works for everyone everywhere is a fool’s errand and the value alignment of engagement is not necessarily healthy for the targeted organism. I talk about this in my article what would life be like without a cellphone.

Ecology has given me a macro view which considers the sheer complexity that can and does happen in even a simple ecosystem. Ecology felt right because it showed how fragile things are; in the first chapter of Economy of Nature, the author points out that the delineation between forest and meadow isn’t a sharp line or captured by any set of rules. In fact, it’s a fool’s errand that is due to the underlying ambiguity of reality and the fact that it escapes our concepts. The book states that former ecologists use to try to define typologies my experience told me that we were always in a state of precarity and many of us were trying to collapse it to provide ground underneath our feet.

Perhaps even the goal of civilisation is to pretend that things are not inherently precarious, not rising and falling from moment to moment, so the underlying angst doesn’t overpower us, providing citizens with some version of solid ground. But perhaps we need citizens who can deal with reality an inch deeper, where uncanny valleys begin to dissipate.

When you study ecology deep enough, you find that everything is fleeting, falling, arising, changing, shifting right under our sensory perception of it. It’s not clear if I and my partner’s relationship is bordering on parasitic or mutualistic today. If we were in fact parasitic in some impossibly permanent way, it still wouldn’t be clear if there was anything inherently wrong, in fact, it might be necessary. Ecology gives you the holistic view that psychology has to do away with to create the deepest concept of an individual. Ecology is the economy of nature in all it’s minute details.

Why Study It?

Study it because you need to broaden your scope of what is.

3 Books

The Economy of Nature by Robert Ricklefs
Dark Ecology by Timothy Morton
Hyperobjects by Timothy Morton
Humankind by Timothy Morton

5 People

Timothy Morton
Robert Ricklefs
Charles Darwin
Rachel Carson
Ernst Haeckel

Philosophy & The Path

I’ve decided to try and split spirituality which I prefer ‘The Path” and philosophy mainly because my study of philosophy happened in very different periods. But let’s be honest, you can not cut a reasonable line through any of these subjects especially not philosophy and spirituality.

Philosophy

I didn’t seek out philosophy as an area of study. I also wouldn’t recommend it when one is in search of meaning or truth for that matter. I discovered philosophy during a mild to severe depression and while I had crippling anxiety. I dove into several flavours of philosophy which might be termed existential, phenomenological, continental, and philosophy of science and mathematics. I prefer the term deconstructive philosophy. I swam in the black sea of Heidegger’s Being and Time which reoriented my relationship with what it means to “be”. First inspired by Nietzsche, while in parallel discovering Wittgenstein, Kierkegaard, Thomas Kuhn, Ernest Becker, and other iconic philosophers, I found my way to Heidegger’s dasein and even named my artist name after one of his terms(being towards death). As I made my way to the shallow depths of my mental life, I moved more towards phenomenology and even dove into Freud, Jung, and Otto Rank, who had centred themselves around the individual and how discriminates between good and bad suffering. The book that broke me into a thousand pieces was The Denial of Death by Ernest Becker. It is an empty masterpiece that if given the right light, will punish you existentially for your ignorance and then show you a glimmer of hope by taking on valid heroism to make your life worthy.

Why Study It?

I’ve found my studies of philosophy and spirituality a rabbit hole worth floundering in. I’ve struggled to understand some of the most basic questions that science presents as “research findings”, but they lack the material to create a lattice-work of how to live one’s life. Studying philosophy and spirituality broadly may allow you to begin to knit a toolkit together. But that being said, philosophy and the study of philosophical groupings and their concepts, can be incredibly misleading if you rely on reasoning alone. Philosophy, like anything, works extremely well when you are willing to put it to the test in your everyday experience, it is at it’s worst when you think you understand the world after you finished a particular book or course. There are different ways of So use it to guide your path towards mental well-being and learning how to live a life. But do not mistake conceptual knowledge with wisdom. Knowing what Heidegger and Kierkegaard said doesn’t make you a wise sage. In fact, it might make you the opposite; it is worth considering that their knowledge could have come from contemplation or direct experience. I prefer the latter. So take the concepts and bring them into action in your daily experience else they are just fascinating symbols that make fake conversations slightly more sophisticated but still lacking in wisdom. Study philosophy and live it, it should be obvious you already are.

Books

The Denial of Death by Ernest Becker
Cutting Through Spiritual Materialism by Chogyam Trungpa
Fear and Trembling by Soren Kierkegaard
Stepping Out of Self-Deception by Rodney Smith
Training the Mind by Chogyam Trungpa
Being and Time by Martin Heidegger

People

Ernest Becker
Thubten Chodron
Soren Kierkegaard
Pema Chodron
Robert Sapolsky
Chogyam Trungpa

The Path

More recently, I’ve been quite attracted to many threads of what is labelled buddhism. I do not claim to be a buddhist and I prefer to call anyone who is in pursuit of deeper truths, who want to give up their delusions and hallucinations and are seeking the cessation to suffering as beings on The Path. Some people might call you a buddhist or themselves because of this but some would say you aren’t until you’ve read the Buddha’s teachings. I’m not convinced that is the only way to awakening. There are many zen priests who have said, “kill the buddha” if you are clinging on to it. The Buddha was not a Buddhist. He was on The Path and he was able to cut through deception, delusion, hallucinations, to arrive at the way things actually are. His grand synthesis shows how much clarity he was able to contain. His teachings even now, hold up to many of our developments (i.e. neuroscience). We are thrown into a world full of suffering and we wish to alleviate that suffering so in ignorance we build up a sense of self that grasps on to external objects and sensual experiences to temporarily alleviate or distract us from our underlying suffering. Each of us can look to our own experiences and see that this is most obviously true; if you do not notice it quickly, then your level of self-deception is likely quite deep. Buddha presents what feels like axioms such as the four noble truths and the eight-fold path. But he also cautions us to question tradition and anything that has not been directly experienced by us. He asks us to not take his axioms as truth (even though he posits them as clearly true) in and of themselves but use your experience of the past through meditation to determine if this is in fact true for you. I have tested many of the “axioms in some pretty extreme ways including different forms of asceticism and hedonism. It is clear to me that the reasons I suffer is due to many of the reasons that the Buddha states concisely. The Buddha was both concise and also exhaustive. Concise because the teachings provide a parsimonious explanation and a strange yet convincing solution to the human condition. Exhaustive because he was extremely redundant using different vehicles (i.e. stories, poems, logic) to explain the same underlying principle or reality.

Why Study It

I have had the chance to read many great writers of The Path including Bikkhu Bohdi, Dali Lama, Pema Chodron, Thubten Chodron, Shantideva, Timothy Morton, Thich Nhat Hanh, Dogen, Paramahansa Yogananda, Urggyen Sanggharaskkhita, Robert Wright, Sam Harris, Khenpo Tsultrim Gyamtso Rinpoche, Chogyam Rinpoche, Shunryu Suzuki, Bohidarma, Malipera, Rodney Brooks and many others (at least one book from each). Their teachings have had a profound effect on the way I choose to live my life. They all have advocated that you must use your time to meditate and contemplate. You must spend plenty of time with yourself doing absolutely nothing. Well maybe “nothing” is too clean, you also need to observe your thoughts as you focus on your breathing in order to become aware of the nature of your mind. Cutting through your egos seemingly infinite ability to twist any sensual or emotional input into something to inflate you to be proud or arrogant or deflate you to drive you towards “productive” action, ultimately protecting it’s own existence. Study The Path in order to become aware of yourself, your suffering, and to gain the clarity, compassion, and wisdom to cut through it all.

5 Books

The Denial of Death by Ernest Becker
Cutting Through Spiritual Materialism by Chogyam Trungpa
Stepping Out of Self-Deception by Rodney Smith
Training the Mind by Chogyam Trungpa

People

Ernest Becker
Thubten Chodron
Soren Kierkegaard
Pema Chodron
Robert Sapolsky
Chogyam Trungpa
Bhikkhu Bodh
Dali Lama
Pema Chodron
Shantideva
Timothy Morton
Thich Nhat Hanh
Dogen
Paramahansa Yogananda

Foreign Languages — Mandarin Chinese

Me, my partner, and my partner's parents

[Me, my partner and her parents]

I have been learning Mandarin Chinese for the last few years applying my own methodology. Mostly taking from what I’ve learned in neuroscience and cognitive science. My partner speaks fluent Chinese so she is my supreme teacher. I have learned most of my Mandarin Chinese without a language reference (without using English to translate it) using immersion techniques that I talk about here and here and also using Rosetta Stone.

Why Learn It?

Learning languages is difficult but worth it even if you never talk to anyone who knows the language. The transformation that can come with learning a language; realising that your language is not the centre of the world and that words don’t have a true ground to stand on since reality is inherently slippery. A moving target. So language is a way to collapse the complexity of the world into things you can point to and what things can do. It’s an impossible feat which is why language doesn’t converge over time. Instead, language is patchwork. Learn it because it might shift your perspective and keep your brain performance at the edge. It opens new worlds, but not in the ways people tell you; these aren’t totally new wholesome worlds, these are perforated leaky worlds that you’ve likely been apart of already and you now can access them in different ways.

Learn a new language because it’s good for you. It’s good for you to learn different ways to map the world to meaning linguistically. It’s good for you to keep your brain in shape. Learning a language immersively engages a large proportion of your brain learning and memory (hippocampus), decision making (should I say this in English? No, say it in Mandarin and the prefrontal cortex steps in), hormones (anxiety and nervousness because you want to communicate clearly but you cannot), reward circuitry (learning pathways), and so on.

Courses

Rosetta Stone Mandarin Chinese
So your partner speaks Mandarin (me)

Conclusion

More than 70% of people over the age of 25 do not have a college degree and they usually go on to occupy low paying jobs. It is predicted, that only 5% of those who get degrees will pay off their education loans in full. Yet the machine is still running, eating itself. It clearly doesn’t have to be this way. We need to upgrade societies cultural norms around education. It should not be the case that someone who has the knowledge of someone who has a BA in Computer Science should be selected over someone who has near expert-level technical knowledge in computer science who didn’t go to school or even if they have the same level. This is a problem of society overloading institutional symbols and the problem of knowledge representation (I talk about how we might solve this in my machine learning and education paper titled DeepEdu here and here.). We are and always will be students of life. With the advent of OCW and MOOCs and other technologies, we should begin to evolve as a society. The fact that it hasn’t might suggest that school is now more about the symbol or the badge than it is about the actual knowledge gained. If that were true, then more people would still be getting college degrees. This appears to be the case.

I’m not against degrees or any of these institutions. I received my education from professors who represent them. I am against delusion and deception. If most people believe that the fantasy of having a specific degree from x institution is better than the fantasy of having the specific knowledge regardless of if you get a badge from a known institution, then it’s likely proof that our society is tipping towards the corruption of education and the pursuit of knowledge.

This reminds me of a story from Buddhism.

“the teachings are merely a vehicle to describe the truth. Don’t mistake it for the truth itself. A finger pointing at the moon is not the moon. The finger is needed to know where to look for the moon, but if you mistake the finger for the moon itself, you will never know the real moon. - Thich Nhat Tanh

So this is our situation; the degree is the finger and the moon is the wisdom to be gained from knowledge acquisition. Do not mistake the finger for the moon itself. And do not mistake the finger to be the only one to point you towards the way. There are more ways now then there ever has been.

I hope this helps you find out more about the world and yourself. I am open to suggestions as well. In my next article on this subject, I’ll take you through the story of why I choose the path of self-learning and the twists and turns that enabled me and sometimes forced me down a strange loop bringing me to where I am now; at the intersection of all of these heap like subjects. On the one hand, I have great knowledge of the world from the perspective of the sciences and literature, but on the other hand, there is still great open space. I found some common threads that might still lead more to some wisdom that I’d like to try and share. Follow your questions with discipline and remember that knowledge is only a vehicle, where you go with it is just as important. Wisdom is the ultimate goal and it’s not clear from the outset how you get from here to there. Until next time.

“The teaching is like a raft that carries you to the other shore. The raft is needed, but the raft is not the other shore. An intelligent person would not carry the raft around on his head after making it across to the other shore. Bhikkhus, my teaching is the raft which can help you cross to the other shore beyond birth and death. Use the raft to cross to the other shore, but don’t hang onto it as your property. Do not become caught in the teaching. You must be able to let it go.”