Have you ever been unable to find a bixi? Or unable to find a parking spot?

I’ve downloaded a map of the Bixi system every 5 minutes for the last couple weeks, and got to working on an animation. In this daily segment for Tuesday June 30th, you clearly see downtown filling up after noon:

(red indicate a station is full; transparent ones are empty. Sizes are roughly proportional)

If you have better ideas for a color scheme, let me know. And if you want the code, it’s up on github.

Yesterday BellKor’s Pragmatic Chaos became eligible to win the $1 million prize by improving Netflix’s Cinematch performance by 10%.

As per contest rules, there are now 30 days for other contestants to try to beat them. The most likely contender is the Grand Prize Team, which allows anyone to join and will split earnings based on how much each member’s contribution adds to the final solution.

With a 30 day deadline, there is a stronger incentive than ever before for other contestants to join the Grand Prize Team.

One of the main lessons of this competition has been that ensemble methods vastly outperform single algorithms. Top teams have been blending hundreds of different results to eke out ever smaller improvements. BellKor’s Pragmatic Chaos consists of the old BellKor, BigChaos and PragmaticTheory, which were at a respective 9.40%, 9.47% and 9.80% improvements over Cinematch. Joining forces, their blended results beat the baseline by 10.05%.

The Grand Prize Team now stands at 9.68%. Blending works best when very different predictors are used, so they still have a chance. But what happens if they get close or equal to BellKor’s Pragmatic Chaos? Normally in game theory teams remain as small as possible to maximize earnings. However these conditions are anything but normal, as loose coalitions are easier to form online and results are visible and updated daily.

If we end up with a super-coalition winning the prize it will only reflect the collegial and cooperative nature of the top contestants. With so much money and fame at stake one of the most surprising aspects of this competition has been how generous most contestants are.

Other coverage:

•  Daniel Lemire: Netflix competition is over?
• Greg Linden: The $1M Netflix Prize has been won
• Wired: Winning Teams Join to Qualify for $1 Million Netflix Prize

If you care about cities, sustainable infrastructure, solar energy, life-logging and product-service systems, the Bixi is worth learning about. It is billed as a fourth generation bike sharing system: portable, solar-powered and wireless.

While it has many problems that have been discussed, I want to cover 4 elements that qualify this system as bright green.

Social Engineering wrought large

There’s already been one accident involving a Bixi rider, and the media has been focusing on the dangers of riding without a helmet. No doubt someone will die riding a Bixi, and the media will declare it some kind of bixipocapypse.

Melora Koepke writes that after the introduction of the Vélib’ in Paris, “Parisian motorists seem to have learned to share the road with cyclists and even respect them.” The same will happen in Montreal. Motorists will change their habits, making all cyclists safer - whether or not they’re riding Bixis.

Despite the media, the perception of cycling as a normal mode of transportation will keep on increasing.

Solar with immediate payback

While our parking kiosks also have solar panels, the Bixi stations are a lot more conspicuous:

Wireless communication and solar panels allow for faster and cheaper installation of stations. So fast that they plan to install temporary stations during special events, something which would be otherwise cost-prohibitive.

Even regular stations get moved. Due to higher than expected demand, a the Bixi station for the Mont-Royal Metro was expanded and moved behind the entrance. Examples like this change the way the ROI for solar technology gets calculated: the cost of saved kilowatt-hours is a small fraction of the cost of installation.

Conspicuous solar infrastructure, off-grid in the heart of the city. Planners from around the world will be looking at the Bixi and integrating this lesson. The meme will spread.

Product Service Systems

Renting rather than owning is not a new idea; We’re used to video rentals, laundromats, libraries, gyms, and taxis. Bixi stations every 300 meters are a high-profile and novel example, creating a teachable moment to talk about car sharing and co-housing.

For those who would compare the bixi to a cheap bike, it’s worth repeating that sharing is often more convenient. Take a bus to get to work in the morning because it’s raining, and you can still bike back at the end of the day. This will change the way you move around the city in surprising ways, just like a cell phone changes how you coordinate social activities with friends. We routinely defer making plans with friends until we’ve arrived near a meeting spot.

Now it’s transportation that’s getting untethered.

Life logging: making the invisible visible

I’ve used a lot of buzzwords already, but this hardly gets a mention in any of the reviews. When you buy a monthly or yearly membership, Bixi tracks all your rentals and that information is available to you on their website:

Bixi is new, and it shows here: there’s no easy way to get all the information. To make matters worse, Bixi’s administrators have sued one developer for unauthorized use of their data, so creating mashups is risky. Whether they’re logging for fitness, weight loss, pure curiosity or an obsession with pretty graphs, a lot of people would like to have their personal data.

To the extent that such applications can encourage more people to use and continue using the Bixi, its administrators should encourage it.

Conclusion

Bixi is a great idea that incorporates many elements of the future we want. Energy efficient, green infrastructure that encourages use rather than ownership. I hope bright green folk will help this meme spread.

Opaque software, politically engaged disability and queer activists, bumbling PR, religious zealots, hacking and a massive internet troll? Welcome to the future of the internet.

Easter week-end will be remembered as the great amazonfail. Here’s a quick recap of some of the most salient facts and theories.

In case you do not know, #amazonfail was Twitter’s top trending tag on Easter week-end, after people realized Amazon was de-listing queer books. The shitstorm starts with Mark Probst’s livejournal entry, “Amazon Follies”.

Amazon claimed it was a glitch, but few people are buying it. Only the de-listing started 2 weeks ago.
Besides, a software bug that classifies “Heather Has Two Mommies” as objectionable while leaving “Recovering from Homosexuality” untouched was swiftly called out as Bisonshit.

Adding to the accusations of homophobia, a search for homosexuality turns up “A Parent’s Guide to Preventing Homosexuality” as its first result:

There’s a problem with the Homophobia angle though: the de-ranking also affects titles targeting disabled people.
Amazon recently got called out by disability activists when they allowed publishers to turn off the text-to-speech function on their Kindle e-books, pissing off “the blind, people with dyslexia, people with learning or processing issues, seniors losing vision, people with spinal cord injuries, people recovering from strokes, and many others for whom the addition of text-to-speech on the Kindle 2 promised for the first time easy, mainstream access to over 245,000 books.” (“We Want to Read” petition)

Ironically, the Kindle edition of “The Ultimate Guide to Sex and Disability” wasn’t de-listed, but the dead-tree version was.

All these facts are consistent with the theory of a large internet troll, the “bantown” theory, advanced by “a former non-member of a non-IRC channel called Bantown which no longer exists”.
The bantown modus operandi? Exploit vulnerabilities in internet trust systems. Recruit a bunch of useful religious idiots to overwhelm Amazon’s automated classification system, and you make Amazon look bad. Time it for Easter week-end when Amazon’s PR staff is operating on a skeleton crew and you can start your very own internet shit storm.

Even if the bantown theory isn’t true, this whole fiasco demonstrates the inherent vulnerability of trusting customers having a political or moral/religious agenda. It is also a textbook example of how NOT to handle a PR crisis. The basics are simple: Google’s first hit for “crisis handling PR” gets you an excellent primer.

The lack of communication and half-baked non-explanation only enraged customers, many of whom already deleted their accounts.

One troll is already claiming responsibility.

Some people like to make beer or wine at home. Instead, I decided I was going to infuse cocoa.

Summary: In a 750 ml bottle, put about 300 g cocoa nibs (~ 1/3 of the bottle). Fill with 188 proof alcohol, and let sit for 2-4 weeks. Strain and mix with 2 parts light vanilla syrup.

You can serve this in shot glasses: start with 35% cream in the bottom, then top with liquor. Or you can spike hot chocolate, spiced with hot pepper, cinnamon and ginger and topped with whipped cream.

About Cocoa Nibs

Cocoa nibs are just pieces of cocoa beans, after the outer skin has been removed. My last batch was made with beans Marie-Eve brought back from the Dominican Republic. The first batch was with Madagascar beans, which resulted in a fruitier chocolate, with a citrus-like acidity. I prefer the stronger chocolate taste in liquor.

In Montreal you can buy beans raw or roasted from Cafe Rico. Online, check out chocolatealchemy.com (buy nibs if you can). They also cover how to roast and winnow beans, so check them out.

The second time I tried this infusion I ground the nibs, and straining was a pain. Not recommended, but it did infuse much faster, even though the ground cocoa solidified at the bottom.

Vanilla Syrup

• 4 cups water
• 5 cups sugar
• Vanilla bean, split length-wise

Bring ingredients to a boil, simmer for a few minutes. Remove bean after cooling. That recipe was too much, so I had a lot of leftovers to serve over pancakes.

Taste, smell, adjust

I’ve only done this a couple times, and haven’t kept the best notes. If you can’t get the same alcohol, you’ll need to adjust infusion time. There was a definite point at which the liquor smelled chocolaty, and the liquor had a rounded taste profile (Warning: dilute the liquor in 2-3 parts water. You’d have to be *stupid* to try tasting 188 proof alcohol. It will hurt ;).

I couldn’t find a recipe online, probably because most people don’t have ready access to cocoa beans. Hopefully that changes, and more people experiment. In the future I’ll try more varietals, and eventually plant to try adding nuts and spices.

Happy experimenting!

We’re starting this week with a new plugin released every day.

The warmongers manage to drag decent people into war with lies and by dehumanizing the “enemy”.

Amidst all the insanity of the Israeli-Palestinian war, there are people that want peace, and refuse to see another person or group as an enemy because of ethnic and religious divisions.

One of those peace makers is Izz el-Deen Aboul Aish, a Palestinian gynecologist who used to work in one of Israel’s biggest hospitals. He also regularly apeared on Israeli TV. One of the show’s hosts called him recently, just as his family had been shelled. The host decided to broadcast the phone call, live. It’s now available with subtitles on youtube; go watch it now.

Should we have more bridge-builders like the doctor and the show’s hosts, the war-mongers couldn’t operate.

Sadly our foreign policy is rather one-sided in its support for the most extremist Zionists. Even burying people alive in their houses does not rouse Canada’s indignation. Friends shouldn’t let friends seed so much hatred.

What’s a foreign peace maker to do? Some temeritous souls put their bodies in front of Israeli bulldozers to prevent houses from being flattened. I am no such hero and wouldn’t ask anyone to do such a thing.

But the conflict isn’t just “over there”. Here are ways we could help:

-Demand our elected officials take a balanced position in their foreign policy.
-Insist our media get facts straight and stop calling people names.
-Stop censorship in our schools

Protecting our ability to see the conflict clearly and refusing to dehumanize one side are objectives we should have for every foreign conflict to which we have ties. What else could we do?

Suppose you collaborate with a designer on a tiny project. 40 hours of work, 5 of which by the designer.

6 months later, you sell the project for $50k. What started as a fun, community-minded site is suddenly a commercial success.

Is the designer owed anything beyond the recognition they were originally offered?

Colektivo was just announced (StartupNorth, Hacker News), a new P2P financing site for small and medium sized enterprises.

I’d like to be excited about it but I smell a rat. Maybe it’s the arthritis making me cantankerous, or vague similarities with Capazoo.

Here’s the deal: if you’re going to make outrageous claims you better be able to back them up. You can have a few days grace while your web site gets spiffed up. If you want months it’s probably so you can rope in rich chumps.

Colektivo promises “more stable and better returns than mutual funds” to lenders and better interest rates for borrowers. I’d like to know what typical interest and default rates are in their financial universe. And what regulators will approve that kind of language on an investment vehicle brochure?

I love peer-to-peer finance, having lent through kiva and consulted for CommunityLend. For this model to succeed we need transparency and accountability.

I hope my gut feeling is wrong on this one, and I invite anyone from Colektivo to provide information that would reassure skeptics that these grandiose claims are legit.

Our single most urgent task today is to imagine a future after oil. A future without cars.

zelaurent is rightfully outraged that Canada is spending $4 billion to bail out car manufacturers. He accurately points out that the financial crisis is being used as a political expedient.

So let’s point out the obvious which politicians will never do: we’ve built a deadly machine.

There were 2,778 deaths due to motor vehicle traffic collisions in the year 2001 - a rate of 8.9 deaths per 100,000 population.1,2  In 2000-2001 there were 24,403 hospital admissions for traffic-related injuries, corresponding to a rate of 79 hospitalizations per 100,000 population. (tc)

We deal with the issue by campaigns against drunk driving, forcing drivers to use a seatbelt and extolling the virtues of air bags.

How about planning cities so we could walk every where? Fast and convenient transit options between cities?

Besides the deadly toll excised by collisions many people suffer premature death or lifelong respiratory diseases. Consider simply the cost to our health care system:

A recent study examined the economic value of reducing the health effects of air pollution by introducing cleaner vehicles and fuels in Canada.  This study found that the economic value of avoiding these health effects was $24 billion over a period of 24 years,  compared to a cost of $6 billion to implement the program. (hc)

This bailout is just the most recent contribution everyone is making to keeping these old industries alive. We’re paying in lives lost and shortness of breath.

Even those costs will seem small compared to the effects of climate change. The scale of the issue is vast and frightening: runaway climate change could kill hundreds of millions of our fellow humans.

It may be possible to eliminate 90% of our emissions by 2050 while still keeping cars around. Some prototypes already get 300 miles per gallon. If we continue bailing out car manufacturers, if we continue enabling their vision, will they ever build comparable machines?

For now most of the moral debate rests upon what will happen to workers.

Many people will be out of jobs. We as a society need to help people reconvert. Expect a fight: unions want to protect their nearly $34 an hour.

My grandfather was a barber, one of the last to shave customers. My father was in the last generation that received an SOS by Morse code. Not too long ago 98% of workers were agricultural. I don’t know anyone who is a tinker or a blacksmith. Very few people fish whales for oil.

Occupations change and economies evolve. We can ease the transition for affected workers if our politicians can see that the car era must come to a close. Throwing more money at the problem or extracting more empty promises of fuel efficiency from manufacturers is merely a delaying tactic.

Our financial crisis was built on the assumption that we could grow our economy safely and indefinitely, ignoring natural limits. Companies padded their balance sheets by externalizing monstrous and ballooning costs.

Let’s not quibble about a few more billion dollars. Let’s ignore the inflated job figures for people we can help reconvert.

It’s time our politicians showed real vision. We need a future where we breathe easier and climate changed is checked. A future where ecological destruction is not counted as a positive because GDP goes up. If they can’t provide that, they have no business calling themselves ‘leaders’.