New world: Outsourcing Your Crowdsourcing

I couldn't improve on the title of this O'Reilly essay. I think it's a fascinating story.

dSLR camera sensors are ridiculously prone to dust contamination. Most of us ignore the dust, but it's a problem for stock photos. The specks can be digitally removed, but the work is tedious and slow. On the other hand, it's not hard to do an adequate job -- just slow. A perfect outsourcing solution, but most photographers prefer to do their own dust cleansing. If the image is worth a lot of money it's worth the labor.

But, what if the image isn't worth much money? What if you're dumping images quickly to an online photo service that depends on "crowdsourcing" (millions of monkeys typing) to produce cheap, good-enough, stock photos?

Outsourcing Your Crowdsourcing - O'Reilly Digital Media Blog

... I recently read an interesting post over at AUPN about a company based in New Delhi, India that is doing post-production work for photographers via FTP. The company is called Differential Technologies and you can check them out yourself at their website — http://www.worldofdt.com

... I went and checked out Diferential’s website anyway, because I was interested to see if they might be able to help me with submitting photos to a micro-stock agency. A while back I wrote about my experiences in trying out the Aperture plugin for the micro-payment stock agency iStockPhoto.com. In the end I found that I really liked the plugin, but just couldn’t justify the amount of time that was necessary to prepare my images for upload.

The trick to iStockPhoto seems to be that one needs to submit as many photos as possible, the photos need to be fairly unique and appealing to a variety of markets, and they need to be well key-worded, and adjusted. They don’t, on the other hand, need to be from big budget photo shoots, shot with ultra-expensive cameras, or have hours and hours of post-production work put into them.

... The team at Differential set me up with a personal FTP account and I sent them an exported JPEG of the original Master image. I used the new Ubermind FTP plugin for Aperture to transmit the image, and selected a full resolution JPEG with the color space set to sRGB as my Export Preset.

About half an hour later I received an email from Differential explaining to me that the job was done and I could download the image from the same FTP account. The result is below. Differential’s email also explained to me that the charge for such an image would normally be $2.00 (USD) due to the excessive amount of dust on the sensor. The price, they say, ranges from $0.50 to about $2.00, so I guess I hit the max on this one....

... After I looked over the image, I wrote back to Differential inquiring about key-wording services. They said they would be happy to work something out with me. I think this could be the start of really great relationship.

On the Aperture side of things, the whole experience got me thinking about how I could optimize the process so that I would have to do the least amount of work, and keep things nicely organized. I really like the option of just having an “iStock” keyword on hand, perhaps in one of my Keyword Control bars, and a Smart Album set up to search on this keyword. I could continue using the Ubermind FTP plugin to send the images to Differential, and when they were finished, I could just import them into Aperture, and send them to iStock using the plugin.

The only work I would have to do would be to set the iStock categories for each image. Later, I could even go as far as to connect them with their corresponding Masters using the Stack tool...

Of course now that we're seeing high-quality online photo editing solutions, we will see Amazon 'Amazing Turk' services for similar post-production services. Home video editing is another obvious example. I'd be glad to send copies of my home videos out for video editing, though that's an example of only the outsourcing part of the equation. The wonder of this story is the clever combination of outsourcing and crowdsourcing. It's a fascinating parable for our times.

Morford and the virtues of small things

Morford, a columnist for SF Gate, specializes in 'over the top' rhetoric. Sometimes it's tiresome, but often it's oddly agreeable. I liked this one...

You Cannot Save The Earth / Does buying that cute recycled organic lip balm really do any good? Your government snickers

... maybe all this good eco-vibration spurs you on even further, and you decide to green up the whole house, get into gray water and solar and reclaimed wood and non-VOC paints and all the rest, and fill the joint with organic cotton sheets and chem-free cleansers and passive heating systems...

...Just look around. It feels as though your heart is being eaten by angry capitalist cockroaches. Like your id is being munched by deranged zombie architects. And your eyes, oh God your eyes, they can't help but be burned like charcoal as they take in mile after mile, town after town, dreary suburban dystopia after dreary suburban dystopia of massive gluttonous eco-mauling overdevelopment, more Wal-Marts and SUV dealers and scabby strip malls and so many generic prison compounds that are apparently actually tract-home complexes it makes you want to rip out your soul with a pickax and feed it to the few remaining wild coyotes in Joshua Tree before someone shoots them all to make way for a new Home Depot...

He's a sensitive soul, but he does have a good, if obvious, comment about individual versus global action. Organic cotton sheets are an aesthetic choice, not a game changing action. The same thing applies to global poverty -- education and tariff reduction do much more to reduce poverty than coins in a box. Still, if a butterflies wings can trigger a storm, then perhaps organic lip balm can lead to a carbon tax..

Why now? Blacksburg and Montreal

Within the past 20 years, in my hometown of Montreal (Quebec), there have been three sets of shootings of college students. One took place at the Ecole Polytechnique in 1989 (affiliated with the University of Montreal), another (smaller) at Concordia University, and one recently at Dawson College. In the first attack 14 women died, in the last the toll was limited only by courage, good fortune and rapid medical care.

The Polytechnique massacre led to stricter gun control laws in Canada. I believe the weapon used in the Dawson shooting was illegal in Canada, but it is widely sold in the US.

In the US there is Columbine and now Blacksburg.

Has anything changed? If we look back at the last fifty years in North America, and we adjust for population growth, will we see intermittent episodes of these events?

I suspect modern hand held weaponry is lighter and smaller, easier to acquire (even where it is not legal), easier to operate, more affordable and more lethal than the weapons of twenty years ago. Is technology change alone responsible for any increased lethality of school shootings? Or perhaps there's no clear pattern at all.

I can't imagine any easy fixes. NPR had an excellent program on US gun control recently. The universal judgment was that the NRA has been utterly victorious. They have cleared the board and crushed the opposition. The NRA made Bush president in 2000, and America learned its lesson. No American politician will dare challenge the NRA for at least a generation. I would not encourage challengers, I know when a cause is lost. For now.

Update 4/18/07: I've been thinking about this, of course. Given Mr. Seung-hui's age, the prevalence of disease, and the history we're given, there's a reasonable chance he was schizophrenic. If so, then the most pragmatic preventive actions, given the impossibility of weapon's control in America, is to focus on the disorder of schizophrenia. Should every university professor and staff-person be required to complete a program of study in schizophrenia and major depression? Should universities focus on improved early recognition and treatment procedures, and techniques to manage the very difficult intersection of culture and psychiatric disease? Above all, we need far more knowledge of how to prevent and treat schizophrenia, and we need to know methods to divert the victims of the disease from violent paths.

Saletan and the organ trade

The growing organ trade first came to media attention around 2002 or so. Saletan has the 2006 update. Growth has been exponential ...

The worldwide market in human organs. - By William Saletan - Slate Magazine.

...The numbers on the maps add up to thousands. According to the World Health Organization, the annual tally of international kidney transactions alone is about 6,000. The evidence includes reports from brokers and physicians, accounts from Indian villages, surveys of hospitals in Japan, government records in Singapore, and scars in Egyptian slums. In Pakistan, 40 percent of people in some villages are turning up with only one kidney. Charts presented at the meetings show the number of "donations" from unrelated Pakistanis skyrocketing. Two-thirds of the people getting these organs are foreigners. Data from the Philippines show the same thing...

This is about taking life from the weak and giving life to the strong. Here's one more reason why this is not a good idea. Someday the friends and family of those who give up their organs will go looking for revenge ...

The botched security of banks: Schneier ignites a comment storm

Schneier has been too kind to the banks and their increasingly inane security procedures; he's mostly left them alone. Today he finally picks on a misguided credit union, though he should be chewing on Vanguard:

Schneier on Security: Bank Botches Two-Factor Authentication:

... Um, hello? Having a username and a password -- even if they're both secret -- does not count as two factors, two layers, or two of anything. You need to have two different authentication systems: a password and a biometric, a password and a token...

The interesting stuff, however, is in the comments. Schneier hit a nerve, and his audience responds. Some comments claim banks believe their regulators want "two factor authentication" and, in the interests of doing nothing of value, they interpret this as multiple passwords, intermittent security questions, anonymous user IDs, etc. They probably figure they can avert expensive mandates for physical tokens with a load of smoke and mirrors. Their probably right, but the increased complexity and illusory security measures will almost certainly increase consumer losses.

I hope Schneier starts piling on to the banks ...

Corzine didn't wear a seatbelt - does your dog have one?

Corzine didn't wear a seatbelt. It's a characteristic of humans, they confuse power over people with power over physics. Kudos to McDonald for an excellent post, and for the valuable reference to Princess Di the seatbelt-less. In Minnesota Corzine would have gotten a ticket, but New Jersey's mandatory seatbelt law isn't due yet ($20 fine).

I wonder if Corzine will consider appearing in public service messages supporting the new law?

Despite my commie credentials, I have sympathy for the libertarian perspective. I think adults should be allowed to ride a motorcycle without a helmet, or a car without a seatbelt -- as long they then forfeit ownership of their organs in the event of brain death. The benefits to organ recipients will outweigh the costs of care for the seatbelt-less who survive.

BTW, your dog needs a seatbelt too. They need it because they're not sentient decision makers, and because a flying 80 lb dog can break your child's neck. Our dog wears a sled dog harness and appears fond of it; I think she likes the feeling of security it gives her. The harness was used for skijoring back when we had snow, and it works for high speed inline skate action too.

Accusation and injustice: Duke and Lacrosse

The accusation has been shown to be unfounded, and the DA will pay a price for what seems now to have been a politically inspired prosecution. Once again we are reminded of what happens when the law is used to further a political agenda, in this case the reelection campaign of a North Carolina district attorney. Of course, on a vastly larger scale, we are have recently seen how the GOP subverted the law to attack their political opponents, including putting one Wisconsin official in jail.

Will the media also pay a price? There's a long list of people who need to reexamine their readiness to mine this story and sell papers. I checked, and it appears that I'm not among the guilty. I recall thinking that the emotion had far outstripped the available data. The NYT had to pay out for prematurely assigning guilt to a Chinese-American physicist a few years back, I expect there will be some well-deserved payments this time as well.

Apple's OS delay: iLife, iWork, NetInfo, remote control -> it's bad

Mac Geeks are mostly fairly complacent about Apple's five month schedule slip to OS X desktop (vs. OS X TV, OS X iPhone, and OS X video iPod). The most knowledgeable, however, are quietly cautious. Gruber isn't saying this is bad, but my read is that he's starting to worry about Apple's priorities.

I'm also unhappy. It's not just OS X. I expected an Apple replacement for AppleWorks over a year ago. Microsoft still owns the only mass market spreadsheet solution on the Mac [1], and it's tied to buying into their entire suite of file formats. iPhoto hasn't had a major update in over a year, and it needs some serious help (such as importing Libraries). We don't have any adequate remote control solution for the Mac -- that's several years behind Windows. Apple's NetInfo based home and corporate network solutions are a mix of state-of-the-art and vintage 1978 unix. Aperture 1.5 is far slower than it should be, and the best explanations point to weaknesses in OS X data services. We all know the OS X Finder is very weak, and that OS X hasn't not yet equaled MacOS Classic file management. Simple Finder is a farce and OS usability leaves much to be desired. Not to mention the Dock ...

Oh, yes, and Safari. The browser that's stuck in 2004. The range of web sites that Safari supports shrinks every month. Google barely supports Safari; anyone who uses Blogger or Gmail with Firefox or IE can't tolerate Safari on Google. Hmm. Safari barely works on Google? Might as well say it doesn't work at all. There is a far better version of Safari in Apple's labs, but it's waiting on 10.5.

Speaking of core OS X applications, don't users deserve a version of Mail.app that loses massive email repositories less often?

Apple clearly has issues with their computing solutions that go beyond OS X. The saving grace for Apple is that XP is ailing and Vista is about as appealing as a root canal. Even so, there's no room for complacency. They're failing in more than one domain and OS X still has a vast amount of unrealized potential.

[1] Time for me to again test OpenOffice. The last time I tried it wasn't ready for my wife. Happily Nisus Writer Express is quite good.

Update: I almost forgot about Safari and Mail. app.

Warm-blooded Fish?

Ben can't sleep.

Dad, are corals warm or cold blooded?
Wikipedia has a lovely diagram. Coral don't have blood, they have a jelly like diffusion substrate.
Go to sleep Ben.

Dad, are there warm blooded fish?
Slashdot (of all places) says yes - sort of. The North Pacific Salmon Shark is particularly impressive, but many fish have various ways to use heat to improve performance. They don't, however, have the same obligatory temperature constraints we have.
Go to sleep Ben.

This is fun.

Humanzees and the Macaque genome

Carl (Loom) Zimmer reviews the implications of sequencing the Macaque genome: Meet the Monkey Cousins. Wow.

One of the most interesting examples was a few hundred gene sequences associated with disease in Humanzees (Hominid-Chimp hybrids, lately known as Homo sapiens). Genes sequences associated with diseases like PKU are normal in Macaque, the disease in us may represent a reversion to an old gene form in an otherwise updated genomic environment.

Astonishing stuff, and much more to come. I remember the day when it was novel to understand disease in evolutionary terms, back when I asked questions about the evolution of the immune system and professors seemed puzzled by my interest ...

The pernicious affects of taxes and accounting rules: crummy software and cubicle farms

Accounting rules for software capitalization are one of the reasons there's so much crummy software around. The rules favor "waterfall development", an expensive way to produce poor quality software. Incidentally, since waterfall development is very compatible with outsourcing, US tax law and accounting standards facilitate outsourcing software development.

It turns out tax laws also facilitate cubicle farms instead of pleasant work evironments:

Moveable walls - Joel on Software

... We're going to need a much bigger space now: on the order of 15,000 square feet. To build that much office space could cost a couple of million dollars. With the lack of deductability, your bank account goes down by three million dollars. The landlord will pay a fraction of that, but not enough to make it affordable.

There's a loophole. Office furniture can be depreciated much faster than leasehold improvements, over 7 years. So for $20 of office furniture you can deduct about $3 a year: better than nothing. Even better, office furniture is a real asset, so you can lease it. Now you're not out any cash, just a convenient monthly payment, which is 100% deductable.

This is why companies build cubicle farms instead of walls, even though the dollar cost is comparable...

Tax laws subsidize trailer parks in some states, and make them unaffordable in other states. Tax laws are why Tokyo once had rice paddies. The more convoluted the tax code becomes, the more unwanted side-effects occur ...

Globalization: poisoned pet food in Namibia

So it wasn't just the US market:

allAfrica.com: Namibia: Pet Food Shock - Manufacturers Recall Stocks (Page 1 of 1)

PET owners in South Africa and Namibia are reeling after two major pet-food manufacturers announced that they were recalling products that had caused kidney failure in dogs and cats...

...In South Africa, 19 dogs in Cape Town and Johannesburg that had been fed Vets Choice have been diagnosed with acute kidney failure, according to the news24 website.

... Earlier, the South African subsidiary of Hill's Pet Nutrition recalled a batch of its Prescription Diet m/d Feline dry food after a similar recall in the United States, where hundreds of cats reportedly died from kidney failure after eating contaminated food...

... The US Food and Drug Administration said tests indicated the food was contaminated with the industrial chemical melamine.

Recently, Woolworths in South Africa had to recall all of its dry dog and cat food due to contamination of certain products...

One wonders about human exposure outside of the US. Human kidneys, however, are far tougher than cat or dog kidneys -- we evolved in a hot, dry environment.

Failure to label on-demand reprints: shame on Peachpit

[Update: Peachpit's publisher responded in comments:

I'm the publisher at Peachpit, and I want to thank you for your comments. Your words sting, as they should, but they also inspire us to work harder to find a better solution for Print on Demand reprints. We don't use this reprint solution very often, and most certainly not for any of our image-intensive graphics and design titles. What exacerbates this is that although we try to monitor these reprints carefully, our supplier keeps changing their equipment. We're in the process of trying to find a new solution.

Which means, of course, I feel better about them.]

I'd naively assumed "on-demand reprints" would be an unmitigated good. Just-in-time printing should reduce waste and improve services. Unfortunately, at least one publisher's solution isn't up to the task. Instead of using a high quality PDF to generate reprints that closely resemble the original, Peachpit Press (Visual Quickstart, etc) uses something like a high quality black and white raster (scan) image for prints. That's unfortunate, but the real problem is a failure in product description (from my Amazon review):

Amazon.com: Python (Visual QuickStart Guide): Books: Chris Fehily (Amazon review by me)

Amazon doesn't allow us to rate the quality of an author's work separately from the publisher's presentation. That's unfortunate, because as an introductory guide to Python this is quite a good work.

Unfortunately, as of 3/07, it's also being published by Peachpit Press as an "on-demand reprint". There is nothing in the Amazon product description to tell you about this change, and indeed I'm not sure Amazon has any way of knowing about it. If you bought this book in a bookstore you'd see the "on-demand reprint" icon on the front cover, but Peachpit Press should have changed the description on Amazon. This reflects poorly on Peachpit, a company I've previously had respect for.

Peachpit's "on-demand reprint" technology is crude. The book resembled the sort of high-quality bound photocopies I used to see sold for $1-$2 in "third world" bookstores twenty years ago. It is entirely gray scale (black and white?) with blurry screen shots and irregular contrast.

The effect is quite annoying. It doesn't make the book worthless by any means, but it hurts. The cover price is $22, $14 is probably a fair sale price IF you know that you're getting an "on-demand reprint". If you can get a used copy you might do better, but of course you might end up with a used "on-demand reprint".

Of course, if you read this you now know what you're getting, and you can make an informed decision without any surprises. Which is as it should be.

Peachpit has a warning icon on the cover of the book, but it's insufficiently descriptive. They have not updated the Amazon description at all, the image picture on Amazon doesn't show the 'on-demand' icon. I'd have kept my respect for Peachpit if the Amazon product description included this warning text:

This product is now an on-demand reprint. It is black and white only with variable contrast. Some images and text will be difficult to read.

If they did that then I'd be fine no matter what they charged, because I'd know what I was getting.

If you get one of these on-demand reprints from Amazon, be sure to write a review describing what you're received. That's the best way to provide feedback to the publishers and to Amazon. In the meantime, I'll use Amazon to buy used rather than new books from Peachpit; that will give me slightly better odds of a much higher quality product.

BBC 4 Reith Lectures 2007: Sachs and the modern world - by mp3, podcast and rss

The BBC Radio 4 - Reith Lectures 2007, featuring Jonathon Sachs on the challenges of the modern world, have begun. Yes, it's no longer sufficient to avidly follow Melvyn Bragg's weekly Radio 4 program In Our Time, the intellectual-geek must also listen to the current and past Reith Lectures. [1]

This year, the BBC has made the lecture available by download (MP3 - only for one week after each lecture!), Podcast (iTunes) and RSS feed [2] . There's a one-click subscription option for iTunes users that's hidden away [2]. As an experiment I'm now subscribed to the feed via Bloglines and iTunes. Note the feed is for Radio 4 Choice, not for this specific lecture. When I subscribed I received both Lecture 1 (yesterday) and an option to get an program on the Falklands war.

Here are the lectures:

Lecture 1: Bursting at the Seams
Lecture 2: Science for Survival
Lecture 3: The Dethronement of the North Atlantic
Lecture 4: The Extremely Poor and the Extremely Worried
Lecture 5: A New Politics for a New Age

The BBC 4 is a cruel example of globalization at its best and most cruel. Best because there's no comparison between, for example, In Our Time, and anything available on NPR. Cruel, because I used to be a regular NPR listener, and I pretty much ignore them now. We still contribute, but how long will we keep doing that?

---

[1] A modern car radio helps.
[2] Some odd things happen in IE and FF when one clicks on the Podcast link. It's an XML document for the Radio 4 RSS feed, and depending on the browser and browser settings it may display a web page, ask you to add a feed to your feed reader, or tell you you've already subscribed to the feed! (This led me to change my FF settings so FF always displays a feed page rather than auto-subscribes). In addition IE 6 displays a quite different page than FF, only IE shows the explicit one-click subscription to iTunes option:

itpc://downloads.bbc.co.uk/rmhttp/
downloadtrial/radio4/radio4choice/rss.xml
vs.
http://downloads.bbc.co.uk/rmhttp/
downloadtrial/radio4/radio4choice/rss.xml

Such are the joys of the bleeding edge. You may be able to find the feed in iTunes, this worked for me:

1. open itunes
2. under the Advanced menu, select 'Subscribe to Podcast'
3. copy and paste the above itpc url.

Update 6/9/07: The mp3s are no longer available from the BBC, but you may be able to find black market versions online. More honestly, the BBC transcripts are now associated with the lecture links above.

Also, Variety has a profile of Sachs. It's a persuasive picture of an obsessive workaholic, humorless, driven, brilliant, relentless and probably often cruel and ruthless. He doesn't sound like someone you'd want to share a beer, or even a building, with. Perhaps this makes him the right man for the ultimate challenge -- the eradication of extreme poverty.

The strage saga of the scrambled letter not-urban myth

An investigation of a 'friendly spam' email leads to an urban myth that's not a myth, and to a comment on reading disability ...

Be the Best You can Be: Scrambled letters and reading disability, not to mention emergent memes and urban myths

"I cdnuolt blveiee taht I cluod aulaclty uesdnatnrd waht I was rdanieg.
The phaonmneal pweor of the hmuan mnid Aoccdrnig to rscheearch at Cmabrigde Uinervtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are, the olny iprmoatnt tihng is taht the frist and lsatltteer be in the rghit pclae.
The rset can be a taotl mses and you can sitll raed it wouthit a porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe. Amzanig huh?;"

...This text circulated on the internet in September 2003. I first became aware of it when a journalist contacted a my colleague Sian Miller on 16th September, trying to track down the original source...

... I've found a ... page that tracked down the original demonstration of the effect of letter randomisation to Graham Rawlinson. Graham wrote a letter to New Scientist in 1999 (in response to a paper by Saberi & Perrot (Nature, 1999) on the effect of reversing short chunks of speech). You can read the letter here, or in a link to New Scientist, here. In it Graham says:

"This reminds me of my PhD at Nottingham University (1976), which showed that randomising letters in the middle of words had little or no effect on the ability of skilled readers to understand the text. Indeed one rapid reader noticed only four or five errors in an A4 page of muddled text."

It's possible that with the publicity offered by the internet, that Dr. Rawlinson's research might be more widely read in future. For those wanting to cite this in their own research the full reference is:

Rawlinson, G. E. (1976) The significance of letter position in word recognition. Unpublished PhD Thesis, Psychology Department, University of Nottingham, Nottingham UK. (summary here)

So Dr. Rawlinson's unpublished 1976 thesis (31 years ago) has come to worldwide attention as the result of a propagating urban myth that's not a myth, and this story is best illustrated by a chaotic and scrambled web page that further extends the original work across multiple languages and references newer software for generating readable scrambles...

Self-referential on many levels.