Wednesday, 17 December 2014

On Beliefs, Assumptions and my World View

Beliefs are not equal.

I had a discussion with a gentleman recently who refused to accept that someone could have no beliefs. This because I told him I had no beliefs. I personally don't like to use the word belief. I would rather refer to any position I hold as a model or approximation of the truth. These models are in turn built on assumptions of properties about this world that could change. It is easy to dismiss the difference between approximations and beliefs as mere semantics, but do keep in mind that semantics is the first thing you learn in a Philosophy 101 course, and ensures that everyone begins a discussion on the same page, instead of ending up talking about different things while referring to the same term. 

Anyway, I see all views, opinions and theories as merely models that are built on assumptions. Nothing is based completely on evidence of course, as even the most basic evidence requires assumptions of the properties that the evidence is based on. For example, the colour red is not really seen by everyone in the exact same way. Our vision ensures that we all see the colour slightly differently, even if this difference is practically negligible. We still call it red though. This is an approximation. A generalisation. But there's more. We assume that the colour red, like other colours, exists as waves made up of photons. We don't know much about how light exists as energy, but we have created models that explain and exploit its properties to a degree that is useful to us. 

Of course, none of this may be real. We may all be plugged into the Matrix. This could all be a dream. The colours may not exist. This universe may not exist. The properties of this physical world that we think we know about may only be a function of a dream world we inhabit and not part of whatever is really out there. But we don't know for sure if this world is fake, and so we act under the assumption that this universe and all the properties in it are real. Because this is the only practical way to live if our goal is comfort and happiness. We don't know if we exist in the way we experience. But it is best to assume that we do.

So in this sense everything is an assumption. But that doesn't mean that all assumptions are equal. There is a hierarchy. If there weren't then any view or model we created, no matter how crazy, could all be equally plausible. So what we do to maintain order in our world is assume that certain things are probably real like our universe and our existence. We then build the rest of our models on top of these basic assumptions.

Now we need to be really careful about how we construct these models, because a lot of them are based on questions that involve incomplete definitions and subjectivity. For example, do we have free will? Luckily, a lot of our models are objective, and built on physical laws whose properties we can approximate quite well. We use mathematical operations to build bridges. Mathematical identities themselves like Pythagoras theorem are perfect and exist for themselves with no exception, at least under the assumptions of the mathematical laws of this universe. We don't know why these identities exist, but we know that they do and how to exploit them. This is not an excuse for a belief in the supernatural. That is simply uncalled for given the evidence. We simply do not know why identities exist. That is all. Any models explaining why will need additional evidence.

You could of course make up your own inductive proof for a supernatural entity that exists outside the laws of this universe and space and time and matter but you would eventually have to face the fact that the properties of this proof are made up by you. i.e the proof works by induction, just like mathematical proofs work, because you assume all the properties needed for it to work, like we do in math. You don't know if these properties are real, you just assume that they are. A logically valid argument will still lead to a false conclusion if its premise is false. If your assumptions are unfounded, then no matter how good your argument, the conclusion you reach will still be only as good as your assumptions. This is why models for God's existence are both perfect and probably wrong.

Moving on, when you build a model, you identify a pattern and make predictions based on evidence. Sometimes, you use other people's models. You act on expectations that another from another model that you know very little about. For example, when you get sick and pop a pill. You don't know anything about what you're consuming. But you take it anyway expecting to get better. Is this a belief?

You could call it a belief, yes. Like the belief that the sun will rise tomorrow morning, given a normal solar system, or the belief that you will be able to walk or talk tomorrow, given no major changes to your body. You could call these beliefs, and they are all based on assumptions. But are they the same as religious beliefs? No, of course not. Because unlike religious beliefs, all these beliefs are verifiable. You cannot know for sure if a pill will cure you, but you know that you can look up the details of the pill if you wanted to. You can examine the skies or your body for patterns if you want to confirm the expectations you have for your model. In other words, these models are verifiable. Not a 100% verifiable of course. Pills do not always work. Solar systems and human bodies do not always work the way we expect them do. Errors abound. Things unaccounted for. The model is updated with new data. This is how critical reasoning works. Religious models are different. They rely, as I have said, on assumptions that are unverifiable. They might lead to useful but false conclusions. Religious belief may be useful, but it is also unverifiable. 

So now we have not just beliefs, but levels of belief. There are verifiable and unverifiable beliefs. This allows for some degree of subjectivity, as what is verifiable depends on how good the evidence is, and all evidence comes down to further assumptions, which always comes down to our assumptions about this universe and our existence. But we can say for sure that some beliefs are more verifiable than others, because some evidence is better than others, assuming the basic laws governing this universe. Evidence that holds up to falsifiability and has predictive value will always be better than anecdotal evidence that relies false premise reasoning and confirmation bias. This is not to say that unverifiable assumptions are wrong. This is impossible to tell, but that's the problem. 

We are mostly concerned with truth or falsity of assumptions based on the evidence we have. But since we can only examine the evidence in light of what we know about the universe, and since this is itself a series of assumptions that do not take into account what we don't know, then of course anything we postulate about God or a supernatural being could be true. Not probable, but possible. I wouldn't say that  assumptions based on rules outside this universe are something we shouldn't bother to think about. But we definitely are limited in the ways we can verify them, given that all our means of verification exist only in this universe.

So yes, perhaps I do have beliefs. I suppose I do live my life along expectations of how the world should work even though I don't always understand why it works this way. These could be called beliefs. And they are certainly different from supernatural beliefs. My beliefs are based on assumptions that are verifiable, at least to a certain extent. I think is a more practical way to live for the moment, compared to holding beliefs that are unverifiable, because at least I can explain why I hold a belief. I can justify my beliefs with evidence. What about you?


Sunday, 14 December 2014

Stats Blogs I Follow

These are the stats blogs that make me better at what I do.

For stats literacy -

This is the main one for serious statisticians. Andrew Gelman, a statistics professor and Bayesian statistician and programmer, critiques poor statistical practices. Very informative - 

For some advanced talk and a lot of useful links -

Deborah Mayo writes about philosophy of statistics -

Great learning resource for advanced statistical concepts -

You might learn a few things from Daniel Lakens' blog -

A nice revision of important concepts with comics -

More on probability theory -


Saturday, 13 December 2014

Education without Innovation

Most complaints about education systems revolve around them being mostly theory without any practical application. This is a problem because practical application like research methodology & computer lessons are a large part of what you need to go from being a theorist to a practitioner. It is no use studying concepts if you can't use them. But there is another problem I have with the system and it is lack of innovation. 

Students are great at learning theoretical concepts. They are great at regurgitating what they are taught in the form of an essay. Sure this is a form of learning. But it is not innovative. When you gain knowledge being taught to you, you grow to the level of the person teaching you, but you don't necessarily exceed this level. This is why learning itself is useless for humanity without innovation. To truly make a change you need to go beyond what you are taught. You cannot simply learn concepts in a vacuum. You have to combine them to come up with new concepts. This is how new things are created. 

It is the same for practical lessons, which may suffer from the same problem. I can put students through computer classes, but it will not mean much if they just recreate what I can, unless you want no new development. The best way forward is to teach your students the basic concepts with practical application, but to connect these lessons with existing questions, theories or ideas that they already have. This makes their learning context dependent, and motivates them to go beyond their lessons, to use what they have learned to create something new. 


To see this in a broader historical context, countries that invested heavily in scientific innovation have always also been quick to reduce poverty and grow economically following innovation. Innovation makes you rich.

For example, Britain once had a lot of poverty. They were able to grow as a nation and coloniser and reduce their poverty because they innovated. This does go hand in hand with how much poverty you have of course. Britain had some labour, but not a lot i.e their labour was expensive because they were few, so they were forced to innovate, to find ways to mechanise processes that didn't require labour. They invented the steam engine, among other things, which meant that more resources could be processed, and which made the means of processing them even cheaper. This also meant that they could now do things quicker and cheaper than other more labour intensive countries could. 

Compare this with India, where everything was done by hand. It still is in many villages, because it is still economical to do so. The low cost of living and availability of cheap labour acts as a deterrent to invent in technology. This is of course fine if you aren't competing with other countries and if those other countries are peaceful. But this isn't the case. Exploiters gotta exploit. Britain had superior technology because they innovated because they were under pressure to do so. India had cheap labour so there was no pressure to innovate and so didn't have superior technology. It was the same with a lot of African and Asian countries where labour was cheap. No investment in technology. No incentive to innovate. And of course the countries with superior technology ended up colonising the countries without any. 

Bouncing back

You also see this with countries like Japan and Germany. Germany of course had a history of scientific development. But Japan didn't. It is interesting to see how these two countries managed to become economic powerhouses and developed countries despite losing world wars. Germany invested heavily in industries prior to both world war one and world war two. Even though they lost the wars, they still had the brains, the skilled technicians to build their economy, to continue creating, processing and selling products and services that other countries needed, which kept the money coming in, which meant they could continue to invest internally, in infrastructure, healthcare, education, and yes, in science and technology, to keep that loop going. 

It was the same with Japan. A country with immense poverty before world war two, they invested heavily in technology and innovation. They knew they were decades behind other countries in scientific development because of their isolationist policy. Political ambition and conquest drove their industrialists and businessmen to invest in technology, to send their best people abroad for training, to bring back, adopt, copy or recreate whatever they could, to bridge that gap between themselves and the west. Which they finally did. In a very short time frame to boot. Sure this was partly driven by war, but following their loss, which included recovering from two atomic bombs, they still had the scientific knowhow to become the number one economy in Asia. Because they had invested in technology like no other country had. So even though they lost, they were still number one in Asia in science and technology. 

History shows us that winning or losing wars doesn't matter as long as you own superior technology and a workforce that knows how to use it. You might occasionally grow overambitious, make dumb decisions like invading another country, and getting your ass kicked and pride hurt, but as long you still own superior technology, you will always bounce back quickly.

Owning the future 

China and India were happy being agrarian societies, while Japan correctly ascertained that if you wanted to be a world leader, you had to own the technologies that no on else had, because this gave you an advantage. You had to have products and services that made you more powerful, because you were able to do things better than any other country (like build better factories that built better cars, faster planes, etc.). This not only gives you a military advantage, but also something to sell to other countries for a very high value. 

Having better weapons not only gives you a military advantage, but it also creates a new market for exports. Having more money go into medical research means a better healthcare industry which means better trained doctors and hospitals with more advanced tools and techniques, which they can export. It also means better pharmaceuticals, which can be licensed or manufactured abroad. Again, the foreign countries that lack innovation only get to do outsourced blue collar work, not highly paying work. R&D stays at home. No country that owns technology is going to sell it. This has changed to some degree in recent years, with companies becoming more global, and R&D happening worldwide, which is an interesting change. It flattens the playing field somewhat.

But it's still shocking that people ruling countries today still act like they don't get the fact that for innovation to truly benefit you, you have to partake in it, so you end up owning the technology that results from it. When you look back at the recent history of India, it is shocking that there has been no efforts at home grown anything. If all you do is import foreign technology, you aren't owning the technology, you're simply renting it, or buying an end product of that technology, which is easily outdated. When India buys weaponry from Russia, Israel or France, it's buying old technology, perhaps even second hand products. Even if it is better than what its competitors have, it is still no comparison to having your own state of the art military industrial complex, like the US, Russia or France have. 

This doesn't only go for weaponry, but also for public infrastructure like trains. Why does India have to go to France, China or Japan to build a Metro or Monorail? Because it doesn't have the technology to do it internally. It has to contract the design work out to foreign firms, and then use local labour to build them. This despite the fact that monorail technology is over a hundred years old. This shows you how backward India is, how lazy it has been at innovating. It isn't like the incentives weren't there. They were, just as they were there for Japan. I don't mean war, but the incentive of not being left behind, of wanting the best for your people. 

Indian leaders simply do not have this vision. If they did, they would invest more in education and research. Without these, you're always going to be second best. You're always going to be left behind. And your country might always be exploited, particularly in terms of trade. Crops don't fetch the same prices that advanced technology does. A lack of innovation means that you're constantly dependent on other countries and their aggressive policies for products and services. To be the best you can't keep chasing the best, you have to outrun them. To chase is to lose. If in ten years you aim to be where the US is today, say n years ahead, then you're still going to be n years behind the US ten years from now. Your goal should be to grow at a faster rate than your competitors if you want to catch up with them. 

Reducing poverty

This is also how you get rid of poverty. Yes, low cost labour intensive production provides jobs to everyone, but it also sustains poverty because it doesn't really help the economy. In 50 years, when other countries have moved on to other technologies and you're still using a low cost labour intensive system to make things by hand, your economy will be in bad shape. Your workers might have jobs, but their pay will be low, because their work is simple and there's many of them. They might have just enough to cover food and basic living expenses, but no money to spend on more expensive goods and services, which means low purchasing power and a smaller market for expensive goods, which hurts the economy. Whereas other countries that abandoned labour intensive production ended up with a highly skilled workforce that are highly paid because labour is now expensive because their skills are valued, and they can now buy expensive stuff, which creates a market for more expensive items, which in turn drives the economy. This is pretty much a comparison of socialist India and capitalist USA in the 80s. 

This isn't a bad thing if Indians don't care about foreign products. But they do. They care about a better quality of life. However, because their economy is in bad shape, there's no money for the government to invest in infrastructure. This is partly due to subsidies, but those subsidies wouldn't exist or matter if your people were richer, which they would be if they had higher order skills that they could sell for more money. If you invest in innovation, say in factory production, there might be some job loss, but in the long run, you will need a highly skilled workforce to manage these new processes. You could have a million people harvest cotton by hand, or you could have machines do it, and have those million people do more specialised, highly paying work, like overseeing the machines, maintaining them, working towards business strategy, doing logistics, HR, marketing, PR, sales, client relationship management, IT. 

Through innovation, investing in scientific development and developing new and better ways to do things, you make life better for your workforce. Instead of earning a pittance doing low value work, they're now earning a lot doing high value work. This is how countries and economies grow. Innovation makes you rich. It's a costly investment, but the returns justify the costs. You see this as a historical pattern when you look at present day industrial nations that used to be agrarian - China, USA, Japan. All you need are leaders who can see and learn from history.