Sridhar Vembu’s Blog

They say it’s hard being from the state of Tamil Nadu and not concerned about politics…atleast during the elections ! Well, it is true personally speaking. I can’t comment abt inflation, currency devaluation that you discussed (but very interested to know more) but one thing is certain: the goodies have been announced by all parties. I’m not quite sure if “most voters are pretty sharp and well informed about these matters these days ” and if “these promises tend to be kept”. I have my doubts. But you are right, pages have been devoted in all the dailies - even in the Bangalore based ones - on the various competing promises (and mostly as to how ridiculous the claims are). What pained me was someone like the federal Finance Minister backing it and when you see the guy making statements - given his credibility and standing on the general Indian economy - you tend to think “who is the one really speaking the truth ?”. Oh well. But then how many elections are based on sound logic…..?

Wolfram states in his Principle of Computational Equivalence

I believe Wolfram’s book A New Kind of Science contains profound insights, in spite of the fact that the author repeatedly hits you on the head with how profound his insights are.

The basic premise is that every real world system is equivalent to a computer program. One corollary to Wolfram’s principle, which he also talks about in his book, is that most problems arising in nature are formally undecidable. Think of real world questions like “Is this face is beautiful?” or “Will this product succeed in the marketplace?” I believe most such questions are formally undecidable, meaning there is no deterministic computer program that can provide correct answers all the time. The issue is not simply the imprecise nature of the definition of words like “beautiful” or “succeed”. Even if precise definitions can be made, the problems will remain undecidable. The key reason is self-reference: at some level, the definition of beauty can only be made against a definition of “normal” or “plain” or “ugly”, which basically means beauty will refer to itself.

Isn’t there a contradiction here? We say real world systems are equivalent to computer programs, but we also say that no computer programs can exist to give always correct answers to questions? There is no contradiction. The real world system is equivalent to a computer program, but no computer program exists to give an always correct answer about that real-world-equivalent-computer-program. To put it differently, the only way to get answers about the real world is to run the real-world-equivalent-computer-program i.e experience life itself! Wolfram points out the triump of experimental science (just run the experiment, don’t just sit and theorize) this corollary implies.

The always-correct part above is important. It is possible to develop programs that give mostly-correct answers. For example, a face recognition program, like the one developed by Riya could aim to provide mostly correct answer to “Is this face beautiful”, under commonly accepted notions of beauty. The program will make occasional errors, but as long as the errors it makes are similar to the errors humans make, it is OK. The program will pass the Turing Test in mimicking a human being in the judgements it makes.

I have found Wolfram’s Principle and the corollary of formal undecidability very useful. A lot of times, arguments and debates and flame wars we get into are really about formally undecidable questions. Questions like “Is Lisp better than Java” fall in that category - in this case, the self-reference is obvious because both are Turing complete, and so both Lisp and Java can be used to mimic the other. So we have to resort to statistical algorithms to decide it, which means, for example, we select a set of representative programs to code, select a set of representative programmers, select some metrics, and then find results. The only problem, of course, is that every step above will be contested (what is a set of representative programs to code? what is a set of representative programmers?)

The search engine ranking problem is formally undecidable too. The self-reference and feedback looks (ever heard of SEO?) are obvious.

Statistical algorithms, like the ones search engines use, seem like the right answer. But we must never confuse almost-correct with always-correct. This is OK for search engines, but in real life, a lot of insights could be hidden in that gap between almost-correct and always-correct. More on that in a future post.

PS: If it feels I am on a spree of documenting the uncomputable, the undecidable and the unmeasurable, I plead guilty as charged.

All the heat and noise over the delay of Windows Vista reinforces one thing I have come to believe: software project schedules are uncomputable (more on this technical terminology below). It is foolish to be going around announcing dates years in advance. I have no idea how project managers in Microsoft can ever precisely predict a delay of a few weeks several months in advance. We have taken our lumps when things got delayed [OK, we don’t have complex multi-year schedules, and frankly, I wouldn’t know how to create one!], and what we learnt was to be very careful promising release dates. I posted on this before at Scale Invariant Slips in Software Release Dates.

I believe the problem is fundamental and theoretical. I alluded to this earlier when I used the word “scale invariance”. Basically, the problem is that the job of creating a project schedule from a given set of requirements is an uncomputable function. The everyday meaning of the term “uncomputable” is that there is no step-by-step, determistic algorithmic process that can guide you in effort estimating the project. The human project manager is basically doing a whole lot of guess work. Let me elaborate.

Consider the situation in a well-understood project management activity like building a house. In building a house, it is possible to precisely spell out every step in the process. In effect, the project manager has created a “program” i.e schedule detailing the steps in the process. In a few years, it is possible that this schedule could be fed to an automated robotic construction system which builds the actual house. What is required is to estimate how long it takes to run the schedule, assuming human or robot workers carrying out the various steps of the schedule. We now have to make a crucial assumption: the schedule has no loops, no go tos and no recursions, so estimating how long each step would take should be possible, so the entire building schedule could be estimated. What is the meaning of not having no loops, no gotos and no recursions? Basically the building construction schedule is a linear workflow, where an activitiy sequentially succeeds another. Branches are possible, but no looping is allowed.

It is obvious why we need to no while loops, no recursion, no gotos condition: the effort estimation problem is impossible with these constructs, due to the famous halting problem. If any significant innovation is required in building a particular house, these conditions will likely be violated, and effort estimation again becomes uncomputable. This is where software project scheduling comes in. In order to effort estimate a project, the project manager is in effect constructing a “schedule program” detailing how the software will be built, based on the given requirements. But unlike in the building example, it is rarely reasonable to assume that the schedule program will have no while loops, no recursions or no go tos. In fact, a trivial example of a while loop is the bug-fixing loop: while (bug_exists) {//fix it} and that is unavoidable in any software project.

Any reasonably complex software schedule will have loops. If the high level schedule for a complex software project does not contain loops [the “pointy haired boss schedule”], then it is likely that some box in that schedule itself does. In effect, what the higher level schedule would have done is to specify “Build this module in 3 months”, without specifying the schedule for the module in question, and it may turn out that the module is itself not estimatable.

Given that the schedule program has loops, the time it takes to run the schedule program [i.e the time it takes to build the software according to the requirements] is not predictable in advance. The only way to effort estimate such a schedule is to actually run the project and see how long it takes. Doesn’t it feel like this is what Microsoft is effectively doing - just “run the project and see what happens?” The only part they seem to be doing wrong is to announce their estimates of when the project will end.

Now coming to scale invariance. Scale invariance is basically a recursive process. As soon as you have recursion (or equivalently while loops or gotos) in your program, you cannot algorithmically guarantee runtime behavior. My contention is that software project schedules have a lot of recursion in them.

Isn’t this a cop out? What are project managers paid for, if they cannot estimate and stick to schedules? That is a subversive question - I will take that up another day.

There were two articles in NY Times in the last 2 days that caught my attention. The first is an Op-Ed To All the Girls I’ve Rejected by Jennifer Delahunty Britz, the Dean of Admissions in Kenyon College. The basic thrust is that there is now a disproportionate number of female applicants who are more qualified. If colleges strictly admitted students based on objective criteria like grades, SAT scores and recommendations, they would end up admitting many more women than men. Here is quote (emphasis mine):

And here is a story on the psychotic effects of drugs like Ritalin, widely prescribed for treating attention deficit disorder & hyperactivity in young kids, disproportionately boys. The news story Panel Advises Disclosure of Drugs’ Psychotic Effects is not really about AD/HD or boys at all, but it tangentially reports that (emphasis mine):

So we need to medicate 10% of boys aged 10-12 so they can attend school normally. I am pretty sure there a connection would be dismissed, but could it be possible that the college admission numbers reported could have something to do with such extensive medication needed in boys at an earlier age?

It is worth drawing the connection to the fact that boys suffer from autism at rates 4 or 5 times greater than girls. I don’t know the ratio of boys to girls in AD-HD but I wouldn’t be surprised if it is at least 2 to 1. Now, here are my questions.

1. Clearly the number of children diagnosed with AD-HD, prescribed stimulants like Ritalin, has increased massively in the last few decades. Could all this increase be pegged to “better diagnosis”? Kids with AD-HD are very difficult to teach in school, because they are disruptive. Did teachers put up with disruptive kids before, and now they don’t? Particularly in an age when “discipline” in kids was more valued than it is now?

2. Could the rise in AD-HD have anything do with the decline in qualified male applicants to colleges?

3. What could cause the increase in AD-HD in boys?

Let me offer some speculation [no claim to originality made]: there are some things that have changed in the last 25 years. One is that the rising levels and the rising number of vaccinations given to kids. Yes, I know, quack theory, fed by frustrated parents. I just took a look at the vaccination schedules for my new-born niece in India, and there were no less than 14 separate shots in the chart, starting on day 1 of her life. 30 years ago, kids were probably given 3 or 4 shots in total, starting much later.

The hypothesis [to many of us parents of kids with autism, it is getting closer to being proved] that many of us are working under is that there is a connection between the dramatic increase in vaccinations, the associated overall rising levels of toxic exposure in children, and disorders like autism and AD-HD. It is by now clear that boys are the weaker sex - they are disproportionately affected by environmental assaults.

The medical establishment insists that vaccinations are perfectly safe. But are there any long term behavioral consequences? I am sure they will dismiss any such connections like the ones drawn above. I am not against vaccination per se. I was lucky to survive Diphtheria as a kid, which has killed millions and millions around the world, and still kills. I know many people in my age group in India afflicted with polio. But it is still worth asking Are we going too far? Are we being too aggressive with vaccinations?

I hope the medical establishment would at least start to investigate these connections. The numbers are staggering.

There is a op-ed article in the Chennai Edition of The Hindu today http://www.hindu.com/2006/03/22/stories/2006032202841000.htm

Reflecting the general leftish politics of The Hindu, the writer calls for a constitutional amendment in India to ban privatization of water. Latin American examples are liberally quoted.

I don’t know enough about the Latin American situation to comment, but I assume the writer knows about our own backyard in Chennai. Even in relatively good times, water is scarce in Chennai. Government water supply is so pathetic that almost all citizens, rich & poor alike, buy water from private sources, carried by tanker trucks. If Chennai citizens were left to the tender mercies of the government, the only solution would be a massive depopulation of Chennai. So water is effectively already privatized in Chennai, but in a kind of quasi-legal way, with no enforceable contracts or well defined property rights over water.

Most of the problems of water in Chennai actually arise from the informal legal status under which the private water tanker operators function. This drives away credible long term capital and investment. With their capital at risk almost daily (the government could crack the whip and capriciously ban them any time), today’s private operators have to aim for the very short term. Lacking any kind of incentive to build a brand or stay around to serve customers longer term, the operators predictably ignore quality and consistency of the product they deliver. They price the water for the short term (on the age old principle “make hay while the sun shines”!), because there is no such thing as long term customer satisfaction for them - they simply don’t expect to be around.

They also lack the incentives to invest in long term, stable sources of water. So the entire business operates in a “fly-by-night” mode (literally so in this case). It is further exacerbated by the lack of defined property rights over water. In a more market oriented system, farmers will have tradable property rights over irrigation or ground water, with allocations based on the extent and the nature of the land owned. In the present system, a tanker operator can suck the ground water of an entire village on a small plot of land leased from one owner.

Instead of identifying such real problems, the writer indulges in predictable left-wing prescriptions. Just passing a constitutional amendment saying water is essential to life and therefore only the government has to provide it is the kind of “solution” that the left usually comes up with (while we are at it, why not pass a constitutional amendment to abolish hunger, by nationalizing all food production - oh wait, hasn’t that been tried before?)

Water is scarce in Chennai, and most of urban India. The only viable solution to allocating this scarce resource is the free market. With a fully legal private sector water system, Chennai residents can get more and better quality water at lower prices than they currently pay. A constitutional amendment to ban private water will simply drive the private water business further underground, raise prices even more, and cause an even greater deal of misery than exists today.

My son turned 7 today. He went through a 2-week round of intravenous chelation, which was finished about a month ago. The chelating agents were DMPS and Calcium-EDTA, both of which have a fairly well-understood safety record in toxic metal removal. The medical establishment doesn’t yet accept this as a valid treatment for autism [because they don’t accept that heavy metals, particularly mercury, have any role to play in autism], but my own personal experience with the academic research establishment during my PhD taught me to question authority, and keep an open mind. I am glad I did.

Urine tests done after the rounds of chelation showed high levels of cadmium and lead; we didn’t see much mercury, but I believe mercury, if present, won’t come out until lead does. At the very least, the urine tests show there is something unusual going on in terms of toxicity, which suggests a biochemical basis for autism. Clearly genes also play a role, but even the statement “Smoking causes cancer” needs to be qualified with “in the set of people that are genetically predisposed to it”, because it is obvious that the majority of people who smoke don’t and won’t get cancer.

We went in to the treatment with our eyes open, not expecting anything dramatic. In the event, we have been pleasantly surprised. Our son has shown measurable step-change in a few critical areas, enough to suggest that the treatment had effect. Nothing miraculous or dramatic, but definitely a positive, noticeable change. His ABA therapists and school teachers also noticed the changes, and they are at best neutral towards these biomedical interventions.

Here is a quick list of changes we have observed in the last 3-4 weeks after chelation. These changes did not happen all at once, but there was a progression.

1. His response time when we ask him a question has improved fairly significantly. The answers come quickly now. Previously, we would often end up repeating the question.

2. For the first time, I spoke to my son on the phone. Before this round of treatments, someone would have to prompt him on the other side to say hello and respond to questions. This time he responded spontaneoulsy to simple questions like “What did you eat?” and “What do you want me to bring for you?” To that last question, his response was “Bring Dell” - meaning my laptop computer. At least he has his priorities right!

3. His speech seems more spontaneous and natural.

He still has a long way to go to attain normal speech. Sometimes we feel we are in a thousand mile journey and only on the hundredth mile, but before he was at the seventieth mile, so that is a welcome change. And he is still showing improvements.

Overall, I am pretty pleased with chelation. We met numerous other parents, many of whom doing a second or third round of chelation, and they all reported seeing very positive results for their kids.

As for the medical establishment, I hope they would pay attention to parents like us, who see clear results from these treatments. Somehow, I am not hopeful they will.

While it has become almost fashionable to pile on President Bush lately, I think the US-India deal is in the long term interests of both US and India. A quick summary: this deal provides American technology for India’s civilian nuclear power program, in exchange for safeguards against military use, and some inspections. Effectively, it recognizes that India is a nuclear power. There is considerable opposition in both countries to this deal, for entirely different reasons.

The Indian critics of the deal are mostly leftists and those instinctively distrustful of the US. Then there are nationalists in India who don’t want any kind of inspections on India’s military program - by the way America or China or UK or France or Russia don’t accept any inspections either. Thankfully the deal opponents constitute a minority in India, so the deal will most likely win approval.

The US critics of the deal are a combination of leftish non-proliferators, and conservative critics of India, and the main grievance is that the deal recognizes a nuclear India. I think both sets of people need to understand that the genie is out of the bottle. India is a nuclear power. The only way India will give up nuclear weapons is if there is a global deal to abolish all such weapons. Fat chance that Americans will approve of that!

Americans may not appreciate it, but we Indians live in a pretty dangerous neighborhood. As our President Kalam (who himself played a vital scientific role in India’s nuclear program) has said, India has been invaded, plundered and colonized again and again in our history, going over thousands of years, and nuclear weapons are our insurance policy that it will never happen again. India has never invaded or colonized anyone, so the weapons are safe with us, as safe as any of those other recognized nuclear powers. The vast majority of Indians will not grant America or the world a veto on this issue. That is why India never signed the Non-Profileration Treaty, with its discriminatory provisions of nuclear haves and have-nots. India never signed that treaty, so it never violated international law in producing its nuclear weapons.

The US Congress may well vote this deal down. It would be a short term set back to India’s civilian nuclear power program. In the long run, that won’t matter, because India is perfectly capable of doing its own nuclear research. Indian scientists are probably secretly hoping the US Congress will vote this deal down, so they can prove themselves by producing the technology. America needs friends like India, and by voting down this deal, it will lose the good-will of the vast majority of Indians who admire it.