Archive for January, 2015

Two Dimensions of Data: Newsletter #25

January 26, 2015

What was that old saw: in God we trust, everyone else bring data? Data and information are the bedrock of modern society. Money, numbers, bits; however you count the beads, it’s data everywhere. 

Yet, there’s no real understanding of data among scientists and scholars, let alone the general public. Even the experts view information from within their specialization – let’s say machine learning or information visualization – than an understanding of the science as a whole. Imagine a world in which people learned numerical simulations for space travel without learning classical mechanics. Physics is a great science because it’s basic concepts – not it’s foundations, but the concepts that all physicists need to know in order to apply their methods to problems in the world – are drilled into physicists from mechanics 101 onward. 

There are two sciences of information: computer science and statistics; both are backed by mathematical theory, but go well beyond mathematics in their real world applicability. Still, there’s a tendency to identify these subjects with their (current) mathematical foundations, i.e., the theory of computation and probability theory. A physicist would find that strange; physics is mathematical, but no physicist would confuse the foundations of physics with the foundations of mathematics. 

Until our understanding of information makes that transition, we won’t have a robust science of form. I believe that transition will require a deeper unification of computing and statistics than is on offer today and in order to do so, we will have to look at the two disciplines from a bird’s eye view first and then narrow down on important questions for unification. It’s a topic that’s beginning to concern me more and more, so I am going to use these newsletters to talk about my ideas every so often. Bear with me if you think I am going all technical. 

Let’s first note that computing and statistics bite different chunks of the information universe. Computing helps us engineer information systems – desktop, laptop and mobile computers and computer networks being the most important. Computing (and once again, let me emphasize that I care more about computer engineering than computer science) integrates information vertically, i.e., it’s about engineering information systems from logic gates all the way to iPhone apps. 

Statistics on the other hand helps us with experimentation, getting data from the world. The integration is horizontal; statisticians care about experimental designs and survey techniques; as the data is brought in for analysis, statisticians also care about techniques for crunching and visualizing the numbers.

Computing and statistics have stayed away from each other for most of their history, starting with training and ending with their typical applications. Statisticians learn continuous mathematics and most of the important applications of statistics have been in unsexy fields such as agricultural genetics and psychology. Computer scientists learn discrete mathematics and from the beginning the science and engineering has been very sexy – from it’s involvement in code breaking to the foundations of mathematics. 

The proliferation of data is the main reason the two fields are beginning to come together. In particular, we need the vertical engineering of computing systems to be driven by the horizontal flow of data. Incidentally, this is exactly what my PhD supervisor, Whitman Richards, was advocating several decades ago. He got the germ of that idea from David Marr’s work on Vision. The marriage of the vertical and the horizontal is not only interesting as engineering, it’s arguably the best way to understanding the relationship between the mind and the brain as well. Machine learning is at the forefront of the marriage of vertical information and horizontal information. I believe that merger will expand to more and more fields in the future. To be continued


Alien Minds: Newsletter #24

January 12, 2015

This week, I am going to talk about something I have puzzled about ever since I was a child but never really taken seriously: the search for extra terrestrial intelligence. SETI, like AI, is one of those elusive, almost dream like goals whose goalposts keep changing. What would count as a truly alien intelligence? When can we say we have discovered (or more likely, stumbled upon) an alien civilization?

I remember Carl Sagan talking about the Golden Record in the Voyager spacecraft, which was his view of the top ten hits of human existence. It has the usual suspects, starting with Mozart and going on to other peaks of civilization as conceived by white male nerds in 1977. OK, that was probably a little unfair, but in retrospect, Sagan’s idea of intelligence and civilization looks rather parochial to me. We are still saddled with a view of aliens as green eyed monsters who play the world of warcraft at a cosmic scale.

The search for intelligence remains the most anthropomorphic of quests; which means that asking whether robots will ever be intelligent is a little bit like asking whether planes fly or not. There’s no principled answer to that question: most of us intuitively think that planes fly, but that’s about it as far as science goes.

Certainly, planes don’t fly in the way birds and insects do and their capacity to fly isn’t based on a genetic endowment of the kind birds and insects have. On the other hand, both mechanical and biological flight are grounded in the principles of fluid dynamics. We can’t build aircraft without understanding how air flows around wings, though it goes without saying that a bird doesn’t understand the principles of aerodynamics in anything like the way an aerospace engineer does. These are different regimes of knowledge. 

In other words, flight is a believable abstraction; we are able to separate out the ability to be in the air for extended periods of time from its biological or mechanical implementation. It doesn’t depend on having feathers or landing gear. Flying doesn’t mean flying like a bird anymore. 

SETI is quite different. We are still focused on finding traces of advanced civilizations, i.e., beings who are like us, but better. I think that’s a major problem in AI as well. Take the Turing test for example: the goal is to create a machine whose answers to questions can’t be distinguished from a human’s answers to the same questions. How much more anthropomorphic can you get? 

SETI and AI pose a metaphysical quandary: on the one hand, we want to understand alien or robotic intelligence on it’s own terms (where the term “alien” encompasses terrestrial intelligence that’s very different from ours – gut bacteria, redwood trees etc) but the only tools and intuitions we have are our own minds and our cultural presuppositions about intelligence. 

Strangely, I think we should explore SETI for the same reason we sit down on a cushion and meditate, i.e., to explore ourselves but also to set aside and ultimately reject self-indulgent and parochial impressions of ourselves. It’s really a religious quest as much as a scientific one. Seen this way, it doesn’t surprise me that the techno-religious cults that have sprung up in the last fifty years (such as the “singularity”) and their manifestation in art (“The Matrix”) are all to do with AI and SETI. As religions go, these alien dreams are shallow spiritual systems, but they have unerringly identified a new direction for contemplation. 

The exploration of Mind and minds – our minds, the minds of other species, the minds of aliens, the minds of robots – and ultimately, the search for the origins of order and organization, is exactly the kind of exploration that brings science together with religion. It’s a search that would be as familiar to the Zen masters of China as the astronomer in her observatory. It is for that reason, not the preserve of scientists alone. Or in some crazy inversion of priorities, to be located in an imagined past of Vedic astronautics.

The adventure of the mind is a new adventure, pointing toward the future, not the past. It’s like Siva’s marriage procession, with room for gods and humans, beasts and demons. Inner space and outer space are deeply intertwined after all. 

In Doubt we Trust. Newsletter #23.

January 4, 2015

Fundamentalism is one of those modern predicaments that often come clothed in ancient garb. Religious fundamentalists like to tout their faithfulness to a pure version of their tradition. In practice, fundamentalism is more about exclusion rather than purity; co-religionists are often targeted for their impure faith – perhaps they sing and dance or celebrate a festival that they shouldn’t. As for those who are outside the cicle, they are fair game. There’s no room for doubt or accommodation; certainty is the hallmark of the fundamentalist. When seen this way, there’s no shortage of scientific fundamentalists either. People like Richard Dawkins are as vehement in their atheism as any Taliban preacher. 

It’s easy to see that certainty is incompatible with humility; without humility, there’s no going forward. Let me be clear, I am not talking about humility as an emotion – some of the most fundamentalist people I know are humble in their external attitude and fanatics in their faith. Humility is an orientation that recognizes one’s humanity and the incompleteness of one’s knowledge. That’s the attitude of the seeker, who is full of doubt, even if she comes to that doubt with great faith. If certainty is the standard of the fundamentalist, doubt is the engine of the seeker. 

I like doubt because certainty is boring. Humility is not just a negative attribute, i.e., the lack of arrogance or omniscience; it is also a positive energy that propels one forward to ask new questions. Let’s put it another way: there are two ways of being: the answer way and the question way. The answer way wants certainty, though it will settle for closure when it can’t get certainty. Consider science, both as it is taught and how it advances: it does so by stacking one answer on top of another. Papers get published because they settled a doubt or verified a hypothesis. There’s no journal of questions. Engineers are more modest. There are no final answers, but products have to ship and customers have to be served and until then there’s a temporary freeze on development. That’s what I mean by the termclosure, you close off all options until further notice. 

The question way has much less prestige. There are no patents for questions. There are no named professorships at Harvard for questions. In fact, it is often dangerous, as children learn quickly after asking awkward questions at home or school. On the other hand, a good question is like an arrow pointed at the uncovered belly of the dragon (I just saw the last episode of the Hobbit); it can bring the whole edifice down and usher a revolution in thought. To the questioner, an answer is just a question’s way of asking another question. A hypothesis might well be verified, but verification is important only to the extent to which it is the key to another door. 

The answer way makes a concession to the fundamentalist. It says, “I am ready to believe, but only when I see it.” Like the fundamentalist, the answerer wants certainty; he is just willing to test his faith a little more. Trust but verify. The question way makes no such concession. There’s always grass to be gathered and a fire to be lit.