On blue sky work…

Useful perspective to keep in mind for the next time one receives unfairly critical comments about speculative work:

Successful research enables problems which once seemed hopelessly complicated to be expressed so simply that we soon forget that they ever were problems. Thus the more successful a research, the more difficult does it become for those who use the result to appreciate the labour which has been put into it. This perhaps is why the very people who live on the results of past researches are so often the most critical of the labour and effort which, in their time, is being expended to simplify the problems of the future.

– Sir Bennett Melvill Jones, British aerodynamicist.

On disciplinary boundaries

A poet once said, “The whole universe is in a glass of wine.” We will probably never know in what sense he meant that, for poets do not write to be understood… How vivid is the claret, pressing its existence into the consciousness that watches it! If our small minds, for some convenience, divide this glass of wine, this universe, into parts– physics, biology, geology, astronomy, psychology, and so on– remember that nature does not know it! So let us put it all back together, not forgetting ultimately what it is for. Let it give us one more final pleasure: drink it and forget it all!

-Richard Feynman, from Six Easy Pieces

How do innovations spread?

Atul Gawande raises a very interesting question in his recent piece in The New Yorker.

The question is, why do some innovations spread so swiftly and others so slowly? He compares two innovations that arose at similar times, surgical anaesthesia and antiseptics. They both were invented in the mid 1800s, in established centres of medical practice. Anaesthesia spread rapidly and quickly became the default practice, even though there were all manner of people who objected to it, ranging from clerics to old surgeons. On the other hand, antiseptics took much longer to be accepted. Although the basic ideas were already published in the Lancet in 1867, even after the 1880s, it wasn’t common practice – not even in the big hospitals! Why?

A key difference, as Gawande points out, is that both innovations helped the patient and saved lives, but one of them required a lot more from the surgeon (proper antiseptic treatment was hard and required some discomfort for the surgeon, who had to put up with carbolic acid and so on). Equally importantly, one of these – the antiseptics – was a hidden cause and effect phenomenon so that it was hard for everyone involved to appreciate the significance of each little step when the phenomenon isn’t so clearly visible.

Apparently, a lot of the same issues prevent effective solutions to some very common but pressing worldwide health issues. Gawande talks about neonatal care in places like India. Many many simple solutions, such as swaddling a baby along with the mother to let the mother’s body temperature regulate the baby’s, aren’t systematically implemented while too much faith is placed on technology-driven solutions such as baby warmers which may not get used due to lack of electricity or spare parts. How does one get these solutions implemented? One could use punishments, or maybe incentives. Neither seem to actually work on their own, and are too hard to ensure correctness of. What one wants is for these things to become norms, and norms get created by an expensive process of teaching cohorts of people who go out and train others, and so on. So, “diffusion is essentially a social process through which people talking to people spread an innovation,”

The rest of the article, which is a fascinating read, goes on to talk about the story of oral rehydration and cholera. The gem of the idea here is that, “coaxing villagers to make the solution with their own hands and explain the messages in their own words, while a trainer observed and guided them, achieved far more than any public-service ad or instructional video could have done.”

I think these lessons are much more broadly applicable beyond public health. Within my own research community, robotics, sometimes we seem to be cooking up more and more elaborate justifications for complex technologies that seem (with my private citizen’s hat on) only suited to showing that we researchers are smart enough to build really complicated systems. We are also very much susceptible to rewarding only the obvious solutions, like anaesthetics, and almost never rewarding the hidden drivers, like antiseptics. If we are serious about one day having some real impact on the world at a large, we also need to heed the rest of the message in Gawande’s story and think about exactly which bits are real bottlenecks to be targeted and how our technology is a solution to some meaningful question out there in the world.

Put your money where your mouth is

A while back, when I was in industry, I was asked by a senior manager if I’d put my retirement savings into a project I was arguing in favor of. His point was that there seems to be a disconnect between the way scientists often vigorously support the feasibility of something and the actual support they’d express if asked to quantify it in terms of a test like above. This is especially true now that I am an academic and I guiltily wonder about the hype in funding applications. One is constantly bombarded with technologies that are ‘just around the corner’, to be addressed by the next ‘breakthrough project’ (i.e., the present proposal) – would the PI actually put his own money into the project?!

In this context, I quite like this wager: http://spectrum.ieee.org/tech-talk/computing/hardware/why-im-wagering-100000-on-quantum-computing/.

On a related but different note, shouldn’t someone be proposing something similar for AI – ask people to provide the impossibility proof instead of deriding people who make attempts to provide positive answers?!

Pearls before breakfast

I am getting ready to leave on a much needed break from work. Strictly speaking, it is not a clean getaway and I’ll be carrying a few trailing ends with me, but this is not the place for discussing that.

As I approach the end of a busy year at work, it is perhaps worthwhile to reflect a bit. This seems particularly apt given that this year, more than the previous couple, was marked by key milestones – my first PhD students graduating, the start of national and international research projects on which I am a co-PI, which brought our group’s first postdoctoral research associate, etc. I finally feel firmly in control of a medium term research agenda that I feel quite excited about.

Amongst all this activity, and the immediate considerations about papers, proposals and laboratory admin, I notice how incredibly easy it is for an academic to get lost in the details – and to be successful without actually doing the things that motivated one to go into this profession. I mean that in a specific sense, which is perhaps best explained by looking outside academia and through an interesting story from the Washington Post, from a couple of years back. The renowned violinist, Joshua Bell, played some of the most exquisite pieces of music in a Washington DC metro station, and nobody really gave a second glance. Here is a snippet of what actually happened:

Now, there is a lot one could say about what this actually means about the people who walked past, why this is or isn’t surprising. I do not want to discuss the reaction of the people in this incident. I do want to make the observation that as I get ensconced in the mechanics of life as an academic scientist, I find myself increasingly at risk of being like those busy passers-by. Everyday, I see people around me exhibit exactly this kind of behaviour because they can’t be sure that stopping to listen is worth their time or consistent with the things they are rushing towards. This is not always accidental and the most disturbing instances are when the behaviour is rationalized away in terms of the way things are.

So, a note to myself is that I should remember to stop and listen when something beautiful calls out. At this early stage in my career, it is entirely uncertain if I can go on to ‘be someone’ or ‘just another …’. However, if that rational calculation gets in the way of doing the kind of free thinking that justified my becoming an academic scientist in the first place, then I will surely only be left embarrassed when I figuratively look back at such a video of my professional life, at the end of my career.

PS: Of course, all of these comments are perhaps even more relevant in the personal sphere. Happy Holidays!

Hard Problems in Social Science

I came across these short talks on hard problems in the social sciences, modeled after the famous Hilbert problems. I am not sure if the analogy holds under strict scrutiny (Is there a Hilbert today who can command the same influence in such a diverse area? Is there even a cohesive community at whom these challenges can be aimed?), but the talks were certainly very interesting and thought provoking.

Clearly, some talks/questions are closer to home than others but even the ones that seemed initially unrelated are thought provoking. The ones that came closest include a talk by Susan Carey on the problem of concept learning – something that is directly on the critical path of intelligent autonomous robotics. I had already read some of her papers before but this talk gave a concise summary and highlighted a connection to human learning in the classroom, something I rarely think about. Taleb’s question was the well known one about how to design robust strategies in the face of ‘black swans’. Despite his notoriety and (as I am told) penchant for personal rants, he managed to be quite constructive in the way he delineated the problem.  This is a problem that does show up quite centrally in our work – how do you learn policies for an a priori unknown open world. Some others were interesting too – Ann Swidler asks what makes a good theory of institutions (e.g., how and why exactly does al-Sistani come to have such enourmous influence in Afghanistan, even more so than the people with the guns or the money)? Her hypotheses were quite interesting and may turn out to be relevant as we wonder how to make resilient organizations of artificial agents!