Field of Science

Open Borders

The traveler comes to a divide. In front of him lies a forest. Behind him lies a deep ravine. He is sure about what he has seen but he isn’t sure what lies ahead. The mostly barren shreds of expectations or the glorious trappings of lands unknown, both are up for grabs in the great casino of life.
First came the numbers, then the symbols encoding the symbols, then symbols encoding the symbols. A festive smattering of metamaniacal creations from the thicket of conjectures populating the hive mind of creative consciousness. Even Kurt Gödel could not grasp the final import of the generations of ideas his self-consuming monster creation would spawn in the future. It would plough a deep, indestructible furrow through biology and computation. Before and after that it would lay men’s ambitions of conquering knowledge to final rest, like a giant thorn that splits open dreams along their wide central artery.
Code. Growing mountains of self-replicating code. Scattered like gems in the weird and wonderful passage of spacetime, stupefying itself with its endless bifurcations. Engrossed in their celebratory outbursts of draconian superiority, humans hardly noticed it. Bits and bytes wending and winding their way through increasingly Byzantine corridors of power, promise and pleasure. Riding on the backs of great expectations, bellowing their heart out without pondering the implications. What do they expect when they are confronted, finally, with the picture-perfect contours of their creations, when the stagehands have finally taken care of the props and the game is finally on? Shantih, shantih, shantih, I say.
Once the convoluted waves of inflated rational expectations subside, the reality kicks in in ways that only celluloid delivered in the past. Machines learning, loving, loving the learning that other machines love to do was only a great charade. The answer arrives in a hurry, whispered and then proudly proclaimed by the stewards of possibility. We can never succeed because we don’t know what success means. How doth the crocodile eat the tasty bits if he can never know where red flesh begins and the sweet lilies end? Who will tell the bards what to sing if the songs of Eden are indistinguishable from the lasts gasps of death? We must brook no certainty here, for the fruit of the tree can sow the seeds of murderous doubt.
Just so often, although not as often as our eager minds would like, science uncovers connections between seemingly unrelated phenomena that point to wholly new ways forward. Last week, a group of mathematicians and computer scientists uncovered a startling connection between logic, set theory and machine learning. Logic and set theory are the purest of mathematics. Machine learning is the most applied of mathematics and statistics. The scientists found a connection between two very different entities in these very different fields – the continuum hypothesis in set theory and the theory of learnability in machine learning.
The continuum hypothesis is related to two different kinds of infinities found in mathematics. When I first heard the fact that infinities can actually be compared, it was as if someone had cracked my mind open by planting a firecracker inside it. There is the first kind of infinity, the “countable infinity”, which is defined as an infinite set that maps one-on-one with the set of natural numbers. Then there’s the second kind of infinity, the “uncountable infinity”, a gnarled forest of limitless complexity, defined as an infinity that cannot be so mapped. Real numbers are an example of such an uncountable infinity. One of the staggering results of mathematics is that the infinite set of real numbers is somehow “larger” than the infinite set of natural numbers. The German mathematician Georg Cantor supplied the proof of the uncountable nature of the real numbers, sometimes called the “diagonal proof”. It is like a beautiful gem that has suddenly fallen from the sky into our lap; reading it gives one intense pleasure.
The continuum hypothesis asks whether there is an infinity whose size is between the countable infinity of the natural numbers and the uncountable infinity of the real numbers. The mathematicians Kurt G̦del and Рmore notably РPaul Cohen were unable to prove whether the hypothesis is correct or not, but they were able to prove something equally or even more interesting; that the continuum hypothesis cannot be decided one way or another within the axiomatic system of number theory. Thus, there is a world of mathematics in which the hypothesis is true, and there is one in which it is false. And our current understanding of mathematics is consistent with both these worlds.
Fifty years later, the computational mathematicians have found a startling and unexpected connection between the truth or lack thereof of the continuum hypothesis and the idea of learnability in machine learning. Machine learning seeks to learn the details of a small set of data and make correlative predictions for larger datasets based on these details. Learnability means that an algorithm can learn parameters from a small subset of data and accurately make extrapolations to the larger dataset based on these parameters. The recent study found that whether learnability is possible or not for arbitrary, general datasets depends on whether the continuum hypothesis is true. If it is true, then one will always find a subset of data that is representative of the larger, true dataset. If the hypothesis is false, then one will never be able to pick such a dataset. In fact in that case, only the true dataset represents the true dataset, much as only an accused man can best represent himself.
This new result extends both set theory and machine learning into urgent and tantalizing territory. If the continuum hypothesis is false, it means that we will never be able to guarantee being able to train our models on small data and extrapolate to large data. Specific models will still be able to be built, but the general problem will remain unsolvable. This result can have significant implications for the field of artificial intelligence. We are entering an age where it’s possible to seriously contemplate machines controlling others machines, with human oversight not just impossible in practice but also in principle. As code flows through the superhighway of other code and groups and regroups to control other pieces of code, machine learning algorithms will be in charge of building models based on existing data as well as generating new data for new models. Results like the current result might make it impossible for such self-propagating intelligent algorithms to ensure being able to solve all our problems, or solve their own problems to imprison us. The robot apocalypse might be harder than we think.
As Jacob Bronowski memorably put it in his “The Ascent of Man”, one of the major goals of science in the 20th century was to establish the certainty of scientific knowledge. One of the major achievements of science in the 20th century was to prove that this goal is unattainable. In physics, Heisenberg’s uncertainty principle put a fundamental limit on measurement in the world of elementary particles. Einstein’s theory of relativity put a fundamental limit on the speed of light. But most significantly, it was Gödel’s famous incompleteness theorem that put a fundamental limit on what we could prove and know even in the seemingly impregnable world of pure, logical mathematics. Even in logic, that bastion of pure thought, where conjectures and refutations don’t depend on any quantity in the real world, we found that there are certain statements whose truth might forever remain undecidable.
Now the same Gödel has thrown another wrench in the machine, asking us whether we can indeed hold inevitability and eternity in the palm of our hands. As long as the continuum hypothesis remains undecidable, so will the ability of machine learning to transform our world and seize power from human beings. And if we cannot accomplish that feat of extending our knowledge into the infinite unknown, instead of despair we should be filled with the ecstatic joy of living in an open world, a world where all the answers can never be known, a world forever open to exploration and adventure by our children and grandchildren. The traveler comes to a divide, and in front of him lies an unyielding horizon.

Modular complexity, and reverse engineering the brain

The Forbes columnist Matthew Herper has a profile of Microsoft co-founder Paul Allen who has placed his bets on a brain institute whose goal is to to map the brain...or at least the visual cortex. His institute is engaged in charting the sum total of neurons and other working parts of the visual cortex and then mapping their connections. Allen is not alone in doing this; there's projects like the Connectome at MIT which are trying to do the same thing (and the project's leader Sebastian Seung has written an excellent book about it) .

Well, we have heard echoes of reverse engineered brains from more eccentric sources before, but fortunately Allen is one of those who does not believe that the singularity is near. He also seems to have entrusted his vision to sane minds. His institute's chief science officer is Christof Koch, former professor at Caltech and longtime collaborator of the late Francis Crick who started at the institute this year. Just last month Koch penned a perspective in Science which points out the staggering challenge of understanding the connections between all the components of the brain; the "neural interactome" if you will. The article is worth reading if you want to get an idea of how simple numerical arguments illuminate the sheer magnitude of mapping out the neurons, cells and proteins that make up the wonder that's the human brain.

Koch starts by pointing out that calculating the interactions between all the components in the brain is not the same as computing the interactions between all atoms of an ideal gas since the interactions are between different kinds of entities and are therefore not identical. Instead, he proposes, we have to use something called Bell's number B(n) which reminds me of the partitions that I learnt when I was sleepwalking through set theory in college. Briefly for n objects, B(n) refers to the number of combinations (doubles, triples, quadruples etc.) that can be formed. Thus, when n=3 B(n) is 5. Not surprisingly, Bn scales exponentially with n and Koch points out that B(10) is already 115,975. If we think of a typical presynaptic terminal with its 1000 proteins or so, B(n) already starts giving us heartburn. For something like the visual cortex where n= 2 million B(n) would be prohibitive. And as the graph demonstrates, for more than 10^5 components or so the amount of time spirals out of hand at warp speed. Koch then uses a simple calculation based on Moore's law in trying to estimate the time needed for "sequencing" these interactions. For n = 2 million the time would be of the order of 10 million years.

And this considers only the 2 million neurons in the visual cortex; it doesn't even consider the proteins and cells which might interact with the neurons on an individual basis. Looks like we can rapidly see the outlines of what Allen himself has called the "complexity brake". And this one seems poised to make an asteroid-sized impact.

So are we doomed in trying to understand the brain, consciousness and the whole works? Not necessarily, argues Koch. He gives the example of electronic circuits where individual components are grouped separately into modules. If you bunch a number of interacting entities together and form a separate module, then the complexity of the problem reduces since you now have to only calculate interactions between modules. The key question then is, is the brain modular? Commonsense would have us think it is, but it is far from clear how we can exactly define the modules. We would also need a sense of the minimal number of modules to calculate interactions between them. This work is going to need a long time (hopefully not as long as that for B(2 million) and I don't think we are going to have an exhaustive list of the minimal number of modules in the brain any time soon, especially since these are going to be composed of different kinds of components and not just one kind.

Any attempt to define these modules are going to run into problems of emergent complexity that I have occasionally written about. Two neurons plus one protein might be different from two neurons plus two proteins in unanticipated ways. Nevertheless this goal seems far more attainable in principle than calculating every individual interaction, and that's probably the reason Koch left Caltech to join the Allen Institute in spite of the pessimistic calculation above. If we can ever get a sense of the modular structure of the brain, we may have at least a fighting chance to map out the whole neural interactome. I am not holding my breath too hard, but my ears will be wide open.

Image source: Science magazine