Wednesday, November 30, 2011

Syrian Parade

I've always hated fireworks. They are a boring waste of gunpowder and sky. But I love parades. Particularly crap ones with donkeys and trucks that break down and a collective shambolic happiness. The best parade I was ever at was in the Syrian town of Hama. In Hama stayed in a hostel with a parrot for a receptionist who could say "Welcome. Please wait in the tv room" in one of 6 languages depending on where he thought you were from.

Hama has a weird atmosphere. There is the contrast of really old norias on the river with the rest of the town. These ancient water wheels are in stark contrast to the rest of the city that is just 1980s concrete block apartments. Out in the countryside everything gets old again. They even have beehive huts like in Ireland

Its only when I got back did I read about the Hama Massacre
The Hama massacre (Arabic: مجزرة حماة‎) occurred in February 1982, when the Syrian army, under the orders of the President of Syria Hafez al-Assad, conducted a scorched earth policy against the town of Hama in order to quell a revolt by the Sunni Muslim community against the regime of al-Assad...
Initial diplomatic reports from western countries stated that 1,000 were killed. Subsequent estimates vary, with the lower estimates claiming that at least 10,000 Syrian citizens were killed,[4] the majority civilians, while others put the number at 20,000 (Robert Fisk), or 40,000 (Syrian Human Rights Committee).

It has the same weird cult of personality photos of Assad everywhere that the rest of Syria has. And the same bird like men that are the security agents everywhere in the world

The parade was in the town when I was there. It was something like a Saint Patricks Day Parade 30 years ago and with better weather.

What has always stuck in my mind for some reason was one group of kids form a karate club who happily marched through the town happily waving coloured flags.

I see the the news footage now from Syria of the protests and the Assad regime violent response. I worry about those kids I saw years ago. Those kids would be the perfect age now to be angry protesting. I'm scared to think what has happened to them. They are having a march in Syria right now and its not nearly as happy as the one I saw.

Tuesday, November 29, 2011

Predictions for 2030

I have no idea what will happen to the euro in the next month. And whatever happens will have big consequences. Yet I was willing to make predictions for 2030 about lbr, driverless cars, solar power and education in my last post.

The thing is I think I am cheating with technology predictions. Theres a wildly unpopular branch of Marxism that argues for technological determinism. Marx can be interpreted as saying that our technology is inevitable and will change us in unavoidable ways. "The windmill gives you society with the feudal lord: the steam-mill, society with the industrial capitalist"

After reading "What technology wants" by Kevin Kelly I am a convinced technological determinist. There is a podcast from Kelly on technology at econtalk

The book makes a compelling case that

1. You cannot shut yourself off form technological change. The Japanese tried it and failed. The Amish dont try it, they stay about 50 years behind on average but they do not avoid new technology forever.

2. No technology ever dies out.

3. Each new technology is an inevitable consequence of the last. No one person of country can cause or prevent a new technology, though they can shape its exact form.

Almost all patents of significant technologies have multiple very similar independant patents lodged at nearly the same time. Airplanes, radio, transisters, computers, television whatever you can think of about three guys thought of it and implemented it independantly within a few months of each other. This suggests that once the right prior technologies exist the next one is inevitable.

Kevin Kelly: What Technology Wants

Keynes wrote about 2030 in 1930 in the article “The Economic Possibilities for Our Grandchildren.” which has been surprisingly on accurate for the last eight decades. So I am betting it will continue to be for the next two. He made brought claims about future growth rates that have been accurate. I will make similar claims now. Given that I dont think non technological predictions can be made I am going to try anyway
1. Grinding poverty will be gone. This is a low bar of 365 dollars a year in 1990 dollars. Thats really bad. Much worse than medieval England but it is one I think we can reach.

2. Polio and guinea worm will be eradicated.

3. World population will be slightly lower than the medium UN estimate of 8321380. The high is 8776486 and the low is 7867332. So I will guess 8250000.

This is just a few predictions based on the world continuing to go the way it has for the last two hundred years. But like I said I cannot predict what will happen to the cash in my pocket over the next two months so two decades away is being ambitious. If you have any predictions for 2030 please put them in the comments. It might be fun for the floating brains in a jar to laugh at them at the time.

Monday, November 28, 2011

Technology of the Year

Loads of brilliant technologies went mainstream this year. Some new technologies were created but I don't know enough about any of them to make claims for their future. I'm going to describe what I think these old but newly well known technologies mean in reference to the world when my new daughter is 18 in 2030.

Learning by reading
Watson came along from IBM and answered quiz questions really well. Siri did something similar with peoples requests.

This might not seem like a big deal. Not many of us make a living answering quiz questions. But how many of us have jobs that involve finding the right page based on some search and parsing out the right bit of text? A surprising amount of medicine, lawyering, general office power point monkeying involves this. What happens when a cockroach knows everything? Watson is the answer to that as that is about the actual intelligence level it has. What happens when in 2030 the cockroach is a thousand times smarter? I have no idea.

Depressingly most of these systems seem to be owned by big companies like apple, Google and IBM. Most of the data used and the tools that analyse it are open source. It will be a really cool project when someone makes an opensource wikipedia based answering bot.

"if, in like manner, the shuttle would weave and the plectrum touch the lyre without a hand to guide them, chief workmen would not want servants, nor masters slaves." Aristotle wrote in Politics which says that smarter machines will free us from servitude. So far this has been the case but will that continue?.

Driverless cars

Will my daughter ever drive for non fun reasons? People will always drive the same way people still ride horses. Just now they do it for fun rather than transport. Two trends are making it less likely that my daughter will ever have a drivers licence. The increasing sophistication of driverless cars and our decreasing acceptance of the idiocy of youth.

If Ferris Bueller had a day off now, would he spend it on Facebook? Because cars are relatively more expensive and relatively less exciting teenagers are less inclined to drive now. Facebook and youtube provide entertainment for free so you are less inclined to work thousands of hours so you can afford a mildly entertaining car. Cars are now only mildly entertaining as safety nazis have turned them all into a homogenous look that cannot even be modded easily. If you cannot put a spoiler and underlights on your Clio your just not going to look cool as a 17 year old.

Combined with cars being less attractive to teenagers there is increasing control placed over them. New drink drive limits for learners have been imposed here. I have heard talk of curfews for young drivers. We are less inclined to accept the god given right to load a car with 7 of their mates and wrap it around a tree at 4 in the morning. In 2030 well have so many constraints on teenagers to prevent them killing themselves as to remove much of the attraction from driving.

The technology element of this is the driverless car

Hanson wrote
So a huge upcoming policy question is: when will what big cities manage to coordinate to change road law to achieve these huge auto-auto economic gains? Thirty years from now we may look back and lament that big city politics was so broken that no big cities could manage it. Or perhaps history will celebrate how the first big city to do it dramatically increased its importance on the world scene

Cowen said
The typical American spends an average of roughly 100 hours a year in traffic; imagine using that time in better ways — by working or just having fun. The irksome burden of commuting might be lessened considerably. Furthermore, computer-driven cars could allow for tighter packing of vehicles on the road, which would speed traffic times and allow a given road or city to handle more cars

These technologies will come in gradually drive train technology, lane assist, parking assist, crash avoidance are all present in next years s-class Mercedes. Probably legal hangups will delay driverless cars. If I was to guess it will be the old that get them pushed through. The old are increasing in numbers and will continue to vote. The baby boomers won't accept the loss of independence that goes with not being able to drive currently. Legal changes to allow driverless cars could be a vote winner. I'm willing to bet that by 2030 a combination of our scardy-cat nature about risk, the grey vote and improved technology will mean that my daughter may never have to drive.

Solar Power

The exponential improvement in solar power became news this year.
Averaged over 30 years, the trend is for an annual 7 percent reduction in the dollars per watt of solar photovoltaic cells...10 years later, in 2030, solar electricity is likely to cost half what coal electricity does today

This article started a great big argument about why solar is not replacing fracking and such. Including why environmentalists are still worried about co2. What will happen in 2030? I dont think we'll be using enough less energy to prevent climate change that way. I think well still be using fossil fuels to fly planes. New electricity generation methods take decades to become part of the grid, but I am still optimistic about solar.

Education has not changed much in a long time. This is partly due to Baumol's cost disease where a teacher now can only teach the same number of kids as one in 1900. Since that time manufacturing jobs can produce vastly more than they used to. But education is one of those human centric industries that has not changed much.

Since videos came out it has been possible to watch lectures at home. Just giving kids computers does not improve education. If you look at the list here of the best educational strategies Tutorial instruction, Reinforcement, Corrective feedback, Cues and explanation are now part of online educational programs.

Stanford this year ran three computer courses where anyone can sign up for free, get homework graded by computer and do exams at the end. This is a graduate level course in one of the worlds best universities for free. The Khan academy has been around for a while but this year it moved from just a list of videos to include problems. The site is now using machine learning and gamification techniques to improve these problems and thus student learning. They are also using statistical tests to improve the website layout and which version of a explanation gets used. These computer assisted techniques are even being used by humans to get better at Jeopardy as the video below describes.

Interestingly the same guys who got learning by reading and driverless cars to their current state are also important in these changes to education Sebastian Thrun, Andrew Ng and Peter Norvig.
This year technologies became mainstream that finally turn computers into the amazing educational tools they have always promised to be.

How did these areas change from when I was born till I was 18?

Cars became safer and drink driving became mildly uncool. The main change here was removing lead from petrol though. The recent drop in crime and increase in IQ may be largely due to removal of leaded petrol. I think driverless cars are important but not as important as this.

Power generation changed a bit, Chernobyl made nuclear uncool but it was probably never economical anyway.

Education did not change much. I did example questions from the year I was born to practice for the leaving cert. Corporal punishment still existed in my time but mainly for nostalgia rather than because they got true joy battering children. I was born after the education system demilitarized. I predict education will change more in my daughters childhood than mine.

Computers did not really visibly exist for people when I was born. By the time I was 18 it was obvious they were going to change the world even if they had not fully done that yet. I think the change from no computers to some was bigger than useful computers to really useful ones that I predict will happen.

Here is the prediction in 2030 for the technologies that became public this year. Many of the cars will be driverless. The world will get a most of its new generational electrical capacity from solar. Education has already changed massively this year just people have not realised it yet. We will each have an assistant that knows everything. I don't know how that will effect employment, my guess is she will end up working in a job that does not exist at the moment.

Wednesday, November 16, 2011

Memorising Books

"All of us have photographic memories, but spend a lifetime learning how to block off the things that are really in there." Granger in Fahrenheit 451. In Fahrenheit 451 there is a group of men who memorise books because they want to preserve them if all the physical copies are burned.

This was the natural state of books before the printing press. The Iliad, the Táin Bó Cúailnge and other epic poems were designed to be memorised and recited rather than read. But what books do people memorize now? I have included the text file size to give some idea of how long the book is.

The Bible 4.2 MB
I have never heard of anyone who memorized the bible but it does seem oddly popular. A Plan For Memorizing The New Testament In 5 Years. The New testament is 996 kB in size. I have never met anyone who has memorised the bible. How many people have?

Milton: Paradise Lost John Basinger did this in his 70's. 12 books, 10565 lines and 60,000 words in 8 years. The filesize is about 495 kB.

The Qu'ran 1.1 MB People who have memorised this have a special name of Hafidh. 'The total number of hafidh and hafidhas currently alive in the world has been estimated in the tens of millions.'

'But becoming a hafiz is also believed to bring rewards in the hereafter, guaranteeing the person entrance to heaven, along with 10 other people of his choosing, provided he does not forget the verses and continues to practice Islam.'

'The children, ages 7 to 14, are full-time students, in class 8 a.m. to 5 p.m., Monday through Friday, even in the summer. But they are not studying math, science or English. Instead, they are memorizing all 6,200 verses in the Koran, a task that usually takes two to three years.'

that is three years of full time work? Over 5000 hours or over half way to Gladwell's 10000 hours theory. Presumably once you spend 5000 hours memorising the Qur'an you would spend a considerable amount of time thinking and discussing it. Which means there are millions of near genius level experts in the Qur'an in the world.

I can think of things I would rather be an expert in but it is worth considering what it is about the Qu'ran that devotes this level of commitment to it.

Mao: The Little Red Book 273kb

One of the most read books ever. It still seems remarkably popular given it was written by the man who killed the most people ever.

'Acknowledging that she had never memorized Mao's quotations, Chen Di said the indoctrination of Mao, which was still inscribed on a few school walls today, was impressed in her mind as a primary school girl.' I cannot find a figure on how many people memorized all of this book but at 900 million copies 'sold' and many of them expected to memorise large passages it must be in the millions.

The Torah: Memorizing the Jewish holy book seems to have an important place in the religion but I cannot get any figures on how many people today have done this.

Can you think of any others? Dr Seuss' 'Cat in the Hat' has probably been memorised by many adults just from reading it to their kids a thousand times.

Books designed to be memorised, like the iliad, have tricks involving characters, meter, rhyme and rhythm. I have gotten two books recommended by Josh Foer, The Book of Memory and Memory in Oral Traditions that examine these methods. So hopefully I will blog about these tricks again.

The effort people must go to to memorize a book is incredible. We can now buy books for the cost of a few hours work. Or a kindle that we can get thousands of books on gutenberg for under sixty euro. Books are 0.2% of the average Americans spending. To think that we have such a low cost per hour worked on a physical or digital book now but people will still spend thousands of hours memorizing one is amazing.

Sunday, November 06, 2011

When could we see ourselves?

I've wondered for a while about how far away we could see the earth if we were using our current technology and we had our current light output. At what distance could we see the artificial light the earth produced?

This paper Detection Technique for Artificially-Illuminated Objects in the Outer Solar System and Beyond provides some sort of evidence on the question

Existing optical telescopes and surveys can detect artificially illuminated objects comparable in total brightness to a major terrestrial city at the outskirts of the Solar System
For this signature to be detectable, the night side needs to have an artificial brightness comparable to the natural illumination of the day side. Clearly, the corresponding extraterrestrial civilization would need to employ much brighter and more extensive artificial lighting than we do currently since the global contrast between the day and night sides is a factor ∼ 6 × 10^5 for the present-day Earth

Lets ignore that our telescopes will get better (about 2.5% a year?) so how quickly at current rates will the earth take to be this bright?

over the last three centuries, and even now, the world spends about 0.72% of its GDP on light. This was the case in the UK in 1700 (UK 1700), is the case in the undeveloped world not on grid electricity in modern times, and is the case for the developed world in modern times using the most advanced lighting technologies.”

But what is that in terms of extra light produced? "In 1700 a typical Briton consumed 580 lumen-hours in the course of a year, from candles, wood and oil. Today, burning electric lights, he uses about 46 megalumen-hours—almost 100,000 times as much."

So that means if we grew at the same rate we have for the last 300 in about 300 years we will be producing enough light that if it was all shone out into space the earth would be as bright at night to an outside observer as it is in the day. I dont think you can make projections this far and it still leaves the question of how long after that ambient external light would be that bright. Still I think its interesting that even with current detection technology as civilisation the 300 years advanced from us would be nearly visible to us. I would guess this prediction will reduce but 300 years is my first estimate to when we could see ourselves.

Tuesday, November 01, 2011

The Population Surge is Stopping

Population surge difficult to halt and almost impossible to reverse was published yesterday in the Irish times. It is also availible in blog form here. The article makes some interesting and arguable claims about human environmental damage to the planet. These are based however on claims about human population that do not match the evidence or the UN's demographic predictions.

Today, just like every day for the last 50 years, around half a million babies will be born.

This is not true the figures from the UN are here. In detailed indicators look in births and in select country look in world.
Between 1960-1965 302136 babies were born each day. Between 1985 and 1990 375909 babies and between 2005 and 2001 367320 babies. The 500000 figure is not just wrong but drastically wrong.

The geometric nature of population growth makes it extraordinarily difficult to arrest, and almost impossible to reverse. The last population doubling took only 40 years. Even if global population growth rate drops to just one per cent, today’s seven billion would swell to an unimaginable 14 billion in 70 years.

The growth rate is being arrested. As I have said before in "We have reached Peak Baby" the number of children each woman has has been falling for decades.

"1968 women got 1.87 adult daughters. Ehrlich called it population bomb. Now women get 1.07 daughters, 92% of way to population balance done." Says Hans Rosling. In 43 years we have gone from 1.87 adult daughters to 1.07 a decline of .8 and we only need a decline of .07 more to reach long term balance of the numbers of women. Rural agrarian Bangladesh has reached stasis in adult female population levels for example.Look at the gapminder video here for yourself the number of children per women is declining

Though surprisingly accurate population estimates get better over time. For example the 2050 estimate has been recently honed in

U.N. Raises “Low” Population Projection for 2050
The "low-variant" scenario of population growth now foresees 117 million more people on the planet in 2050 than it did two years ago.
While the "median-variant" scenario, often seen as "most likely," remains almost the same as before - predicting a world with 9.2 billion people by mid-century
The high projection, however, foresees some 10.5 billion people - a 295 million person decrease from the previous high projection. The medium projection is 9.2 billion people,

It says something that an decline in the high estimate an over 2.5 times bigger than the increase in the low estimate is not the headline.

There are still countries that have very high birth rates per woman. Senegal for example had 7.5 babies per woman in 1968 4.8 babies and in 2010. But these countries are developing at a speed that is likely to see these birthrates drop rapidly. "Senegal has lower borrowing costs than Ireland." and a GDP growth rate of 4.2%. As these countries that still have a high birthrate develop their birthrate will drop rapidly the way ours and other developed countries did.

The article is in the Irish Times gives the wrong figure for the number of births per day and "unimaginable 14 billion" scare figure. The demographic evidence and historical trends indicate the population will not go to this level. Current UN estimates are for the high population prediction do not match this figure for 2081. There were never that many babies born per day and the birthrate has been falling so fast that we will not reach the 14 billion in 2081 figure.