Wednesday, December 28, 2011

Most Popular Posts of the Year

Thanks to everyone who read and commented this year. The most popular posts this year were

Search Engine Deoptimization 16,000 Pageviews. Many people seem to dislike Ryanair and weird legal rules.

Eurovision Voting Fraud 5,500 Pageviews. A rather glib post from 2010 about human rights abuse that got popular when Azerbaijan won the Eurovision.

My favorite posts were

The Dead Zoo Dodo Get Anto the Dodo back to the public. Anto has set up a twitter account

The Kilbrittain Whale. This tourist attraction is brilliantly unhinged.

Wednesday, December 21, 2011

My favorite books read in 2011

Books really are kind of magic aren't they? Here is the ones I read this year that changed the way I think and turned me into a different person at the end. All for ten euro and a few hours of my time.

My favorite book was Moonwalking with Einstein. Without memory what would we be? This is a great diverse book on a really interesting topic.

Don't sleep, there are snakes. There is so much in this book from Language to God to child rearing to self sufficiency. It is just fascinating to read about people who see the world completely different to us.

The rational optimist nearly everything is getting better. This book left profoundly optimistic about the future. This is the book I have tried most to persuade my friends to read most this year.

Selfish reasons to have more kids. This book goes through the evidence that once your not an awful parent (one who would be refused an adoption for example) you really don't make much difference to your kids. It is actually a profoundly optimistic argument as this means you don't need to spend your time doing things you dislike to shape your childs path as by the time they are 18 they will end up where they would have anyway. It uses evidence from identical twin adoption studies to show how little parents need to worry about religion, education and all the other things they break their hearts worrying about.

The Great Stagnation By Tyler Cowen. We should have dinner in a pill, Optimus Prime, hover boards and lasers that turn dolls into women by now. Why don't we and what can we do about it?

Next year I plan to read mainly fiction. This year In cold blood by Capote was the one piece of fiction I read that I am sure will stay with me. Everyone says it is great and everyone ain't wrong.

Close but not in my very favorite books were fever by Shah, Adapt by Hardford, What technology wants by Kevin Kelly, Red Plenty by Francis Spufford, Gutenberg, race against the machine, good omens, the unbearable lightness of being, How the irish save civilisation and Launching The Innovation Renaissance by Alex Tabarrok, popular crime by Bill James

Sunday, December 18, 2011

How much would a driverless taxi cost?

This is a post where I do a back of the envelope estimate of when we'll see driverless cars, what will they do to taxi costs and what will that do to unemployment.

There are several estimates to when driverless cars will arrive. The New York time estimates 2020.
Self-Driving Cars
By 2030, Sebastian Thrun predicts, more people will use self-driving cars in their daily commute than manually driven cars.
Submitted by Sebastian Thrun, developer of Google’s self-driving car.
Our readers predict this will occur around 2020, having moved this date 1381 times.

Many similar bets of Driverless cars being regular enough in 2020 and ubiquitous in 2030 exist for example here,
By 2020 - Driverless cars are commercially-available and street-legal somewhere in the United States.
By 2027 - New driverless cars outnumber new cars requiring at least some human control, in the US market.

and here
By 2019, it has begun spreading to public roads, with significant numbers of driverless trucks appearing.

and here
In my post last week, my commentors took me to task on my prediction that cars will drive us in ten years. Some thought Americans would wise up and learn to love mass transit. They don't know Americans.

Others thought the hardware cost would even in ten years remain out of reach. Google did not build an autonomous car by creating the hardware but by harnessing and training good machine learning algorithms. No amount of hardware would have given you a car able to navigate the streets of San Francisco five years ago.

What effect will these cars have? There are all sorts of ideas about how they will alter parking and car ownership. I'm going to try do a back of the envelope here on how much Taxi fares will cost if you don't have to pay the driver.

Taxis cost about 120 cent for a kilometer

For each additional 1/6th of a kilometre or time 28 seconds)

(a) Day time 8am to 10pm €0.15
(b) Night time 10pm to 8am €0.20
(c) Sundays, Public Holidays, Christmas Eve and New Year’s Eve €0.20

Ignoring the pick up costs of about 3.40.

The AA says it costs around 25 cent per kilometer to drive a car in Ireland. About 25 pence per mile in the UK. Taxis charge 1.20 so the majority of the cost looks like the driver. You would have extra costs on top of a normal car with a commercial vehicle. But given the pick up costs a driverless taxi could be about a quarter the cost of a taxi with a driver.
The price elasticity of demand should allow an estimate of how this will alter taxi usage. This paper "Estimation of Price Elasticity for Taxi Services in Hassel" gives a PED of -2.644. Though others such as Schaller at -.22 and here of -.6 shorter term. Taxis that cost a quarter the current price with a PED of -2.6 would mean about ten times the number of taxi journeys. The long term viability of public transport should take this possibility into account. If by 2030 people will be taking ten times the number of taxi journeys would enough people be using Metro North to make it cost effective?

Transport employs nearly one hundred thousand people in Ireland. Which is about 1 in 20 people who have a job here. Or about a third of the number of unemployed. I doubt everyone who works in transport will lose their jobs overnight. But taxis provide an example of how economic effects could provide a huge incentive to move to driverless cars. So far technological progress has always eventually resulted in new jobs to replace old lost ones. The money people save getting into town for a night now could end up being spent in town and require more employment in restaurants and bars for example.

But I think it is worth considering the possibility that fairly soon we could have nearly a hundred thousand people who earn a decent wage at the moment becoming unemployed in a short period of time. Construction lost 160 thousand people in three years. Transport jobs do not pay as well as construction did. But if the construction change caused most of our current economic issues it would be unwise to ignore a large sudden future change in transport employment.

To put some skin in the game, I am predicting that in 2025 in a period of three years we will see structural unemployment of about 5% of the workforce, half of those that work in transport.

Tuesday, December 13, 2011

Eradication Game Theory

Can you crowdsource disease eradictaion?

I mentioned in this post on 2030 that I expect Polio and Guinea worm to be eradicated by then. It becomes a tricky issue when a disease gets really uncommon how you find and treat the last few cases? So far only smallpox and rinderpest have been eradicated. Eradication is great because once its done its done. You would not have to immunise every child for polio any more. Every polio vaccine has a small cost and risk and once thats gone you can go spend the money on something better

India has tried an interesting tack 'Cash awards for info on diseases'
Alert the government on the occurrence of new cases of certain ailments and you may get a cash award! The government has targeted vaccine preventable diseases such as diphtheria, pertussis, measles, tetanus, and leprosy for elimination by 2016. India is also on the verge of being declared polio-free.
District medical and health officer Dr G. Srinivasulu explains that even after being declared polio-free, there should not be a single new case for 14 consecutive months in the country. This is where the reward comes in. “If anybody succeeds in detecting a new polio case meanwhile, the government will give a cash award. Even in case of detection of new leprosy cases, ASHA health workers are given Rs 150-Rs 200,” he said.

There was a competition last year from DARPA to encourage new ways to pool information held my many people.
'MIT wins $40,000 prize in nationwide balloon-hunt contest'

A team from the Massachusetts Institute of Technology won $40,000 in a high-tech scavenger hunt on Saturday by discovering the location of 10 red weather balloons.

"We're giving $2,000 per balloon to the first person to send us the correct coordinates, but that's not all -- we're also giving $1,000 to the person who invited them. Then we're giving $500 whoever invited the inviter, and $250 to whoever invited them, and so on..." it said.

Some similar system that set up a chain of reward could be really useful in disease eradication. Many security protocols rely on a sort of iterative proof of trustworthiness. Something similar could be used to allow steps toward eradication without fear some other country is going to stop efforts.

Assurance contracts are another approach. It is possible that some regions are scared that once a disease is eradicated to their area they will lose funds. Some way to guarantee that funding will not reduce or to reward successful eradication might help here. Something like Dominant assurance contracts might help to change incentives to encourage eradication.

How about a guarantee fund for each of the remaining countries with polio and guinea worm that when who declares them free they get some cash bonus. You could imagine a kickstarter project that gave the minister of health in Nigeria money when polio was declared eradicated from the country.

Another good plan is not to have fake immunisation drives against polio

The juice of the carrot, the smile of the parrot
A little drop of claret - anything that rocks
Elvis and Scotty, days when I ain't spotty,
Sitting on the potty - curing smallpox

We want to discourage deliberately being lax about a disease so you have to be clever about incentives. Anyone have any good idea for how you could bribe people and governments to further incentivise disease eradication?

Friday, December 09, 2011

A Rose Named 20175 Does Not Smell as Sweet

Would we get more or less innovation without patents? The rose that holds patent no 20175 tells us how little we need patents. The plants have been patentable in America since 1930. 16% of Rose varieties are now patented but this increase in protection to rose growers has not increased the variety of rose varieties.

Launching The Innovation Renaissance by Alex Tabarrok is a short €2 book I read on the bus into work. It details as well as the perils of rose patents other hindrences that block wider innovation

The argument against patents is that they give too much protection to innovators. That in covering too much that is too easily thought of they actually hinder people coming up with new products. I have not seen a coherent argument that (non pharmaceutical) patents have too little protection. Tabarrok does a moderate anti patent argument forward in a clear convincing way though.

Next Tabarrok argues we should have more prizes for innovation. I have a major brain erection for kaggle and science innovation prizes in general. Tabarrok makes a clear case that prizes should be used when new ideas are needed.

Many of the faults in education are also well laid out. "the value of a permanent 25-point increase in scores ...would be 80 trillion" ..."A 25 point increase would bring the united states from about the level of mathematics education in Ireland and Spain to the level in Germany and Australia" How are we in Ireland worse at maths than Australia? Australia is mainly populated with Irish people, and the ones with an unhealthily active interest in sports at that. We have no excuse for the Aussies beating us in Maths. The education section gives several good pieces of advise for Ireland. Pay teachers based on their subject. Pay them on results not just years served. Sponsor college courses that result in more innovation that increases the standard of living more than some of the liberal arts that don't. While on the subject there is a very good blog on Irish education from the teachers side at anseo.

Immigration of high skilled immigrants is discussed next. Again like patents a reasonable achievable change is suggested rather than fully free immigration. I think Ireland should go moneyball on this. Find undervalued players in the market and buy them for our team (this would mean letting persecuted people live here, but it is an argument for another time).

The main thrust of the book is that innovation matters. Instead of a welfare state dividing a pie it might be worth thinking of an innovation state that constantly tried to make things better and increase the size of the pie. This would involve fighting entrenched interests "Few people lobby for innovation because almost by definition, innovation creates present losers and future winners and the present winners are by far the more politically powerful".

The optimistic view that we have in our power to create new and amazing things infuses this book. It is about fixes rather than showing how bad the innovation problem is as Cowen's The Great Stagnation does. None of the suggestions are on the extreme side and all are clearly explained. Because of this I think this book is likely to provide a blueprint for many improvements in the near future.

The Irish times has an interview with Tabarrok here

Thursday, December 08, 2011

Should Science Bloggers Have a Session?

I think it would be really cool if loads of science bloggers talked about the same thing on the same day. Beer bloggers have a monthly "session"

"The Session, a.k.a. Beer Blogging Friday, is an opportunity once a month for beer bloggers from around the world to get together and write from their own unique perspective on a single topic". The beer blogging community really get around the session and seem to delight in sharing stories around one topic. Something similar for science bloggers would be cool, it might already exist but I cannot find it.

If some neglected interesting topic was blogged about once a month that could be really fun. To see the many angles bloggers can take on a subject.

Here is a suggestion World Leprosy day is January 30th.

How about we all write a blog about leprosy for that day?

Off the top of my head here are some possible angles

1. Epidemiology
How is leprosy spread
Where is it common
Can it be eradicated? An irish leprosy charity The Leprosy Mission is here

2. Microbiology
Why can't leprosy be diagnosed until you have symptoms
Why is the Leprosy bacteria so hard to culture
Why do some people take decades to develop symptoms
Why are most people genetically immune from the disease
The reviled drug Thalidamide has a role in modern leprosy treatment why

3. History
Why is leprosy one of the oldest known infectious diseases. How can you tell a mummy has leprosy
How has treatment changed through the ages
How has leprosy effected history? From the king in Braveheart to the Gospels leprosy has been important in history.
Leopardstown I have heard was a 'leper colony'. Was it and what did it look like?

4. Sociology
Why is there such a taboo against leprosy? What can we do to ease this?

5. Medicine
What are the symptoms of Leprosy and why do they occur?

6. Zoology
Armadillos can spread leprosy to humans. Other animals have also been implicated. What does it mean for the animal to carry this disease? How can infection from these animals be minimised?

Thats just a few ideas. None of which I know enough about to be able to write up. But you or someone you know might. It is a neglected disease. I have heard it small pox described as the disease that is gone but not forgotten and leprosy as the disease that is not gone but is forgotten. How about a session for science bloggers on leprosy on January 30th? If you have any ideas or comments or will commit to writing a post please comment below.

Saturday, December 03, 2011

Lithium in the water supply

There is an interesting proposal in the Irish Times
Psychiatrist calls for lithium to be added to water

A consultant psychiatrist last night called on Government to add lithium salts to the public water supply in a bid to lower the suicide rate and depression among the general population.

At a mental health forum on “Depression in Rural Ireland” in Ennistymon, Co Clare, Dr Moosajee Bhamjee said that “there is growing scientific evidence that adding trace amounts of the drug lithium to a water supply can lower rates of suicide and depression”.

I would like to see what this scientific evidence is. The studies I have seen are from Japan where the lithium is naturally occuring. And one from Texas that seems a bit dodgy
Drinking water which contains the element lithium may reduce the risk of suicide, a Japanese study suggests.
Researchers examined levels of lithium in drinking water and suicide rates in the prefecture of Oita, which has a population of more than one million. The suicide rate was significantly lower in those areas with the highest levels of the element, they wrote in the British Journal of Psychiatry.

So taking about 1/1000 of what someone who need treatment takes could reduce suicide? There is an interesting correlation/causation problem with this study. Is it that the lithium in the water is reducing the suicide rate or could the same thing that reduces suicide increases the lithium amounts? One explanation I have heard is that places where it does not rain much water hangs around longer and rubs off rocks picking up lithium. In places where it rains a lot the lithium levels might be naturally lower as the water does not get a chance to pick up much lithium. The kicker here is that places where it rains all the time might be depressing and that could explain the increased suicide rate.

A quick check of this theory throws up some problems. Oita is in the south of Japan. Oita has 94 days of rain per year . It has Average rainfall of 1677mm

Tokyo where people top themselves at the rate of depressed lemmings at a Leonard Cohen concert also has 94 days of rain per year. 'The city of Hiroshima, in western Honshu, averages a sizeable 1,603 mm (63.1 in) of rain each year. Tokyo, further east near the Pacific, receives an annual average rainfall of 1,460 mm (57.5 in). The city of Sapporo, on Hokkaido, averages 1,158 mm (45.6 in) of precipitation per year. The southern end of the Kii Peninsula is known for a heavy annual rainfall exceeding 4,000 mm (157.5 in).'

A quick back of the envelope is not enough to discout the "rain means less lithium but also more depression" explanation. But I still don't think it is time to put lithium in the drinking water.

Takeshi Terao, a coauthor of the paper and a professor at Oita University.
"I do not think [cities should start adding lithium to the water supply], because our study is a preliminary one and further studies are required to establish evidence."
"Lithium does have its negative side effects as well. Some are mild: people often feel thirsty when taking lithium. Other side effects can be more severe, like weight-gain and diabetes and kidney problems. Lithium in the water supply could increase these side effects as well, although Terao's study didn't examine this possibility. Since the dosages are so much smaller, presumably the side effects would be as well, although more research is needed to prove that."

While on the topic of Lithium 7up used to contain lithium
The product, originally named "Bib-Label Lithiated Lemon-Lime Soda", was launched two weeks before the Wall Street Crash of 1929.[49] It contained the mood stabiliser lithium citrate and was one of a number of patent medicine products popular in the late-19th and early-20th centuries.[50] Its name was soon changed to 7 Up; all American beverage makers were forced to remove lithium in 1948.