Alexis C. Madrigal

After a weird IPO, no one knows exactly what's going to happen with the social media's shares on its first full day of trading.

Just when you thought absolutely nothing interesting could happen on Friday, when Facebook debuted on NASDAQ, something did: the company's bankers really did set the price of the offering as high as the market would bear. That is to say, when the shares made it out to the open market, they didn't immediately spike as has happened with so many other tech IPOs like LinkedIn's last year. Some have argued that's a good thing, others that it wasn't.

Regardless, as Friday's trading drew to a close, Facebook's underwriters (namely, Morgan Stanley and the consortium of banks the company put together) propped up Facebook's share price at $38. They just would not let the company's share price fall below where they priced it. If you want to see a detailed blow-by-blow of how it worked, check out this rather stunning video recap. Meanwhile, other social media companies' shares tanked for reasons that are not quite clear.

In any case, no one expects those underwriters to keep up that kind of buying. So, Facebook will have to stand on its own. The big surprise, it seems, was that Regular Joe investors seemed unimpressed with the company's offering. Most industry watchers expected the stock to pop because they figured retail investors would herd into the shares. For whatever reason, those regular people didn't on Friday. Now everyone is waiting for the opening bell to find out if they will be more interested now that the hype of the IPO is over. I know better than to offer a prediction.

Many tech twitterers thought the company's first day on the market would go a lot better than it did.

With the market closing up shop and Facebook sitting at $38 thanks to the deep pockets of the IPO's underwriters, so it's worth revisiting what people were saying about this IPO yesterday. Luckily, developer James Proud created a little app -- facebookipodayclosingprice.com -- at venture capitalist Chris Sacca's request to track predictions about the company's first day of trading. This got tweeted out to the Twitter tech elite and about 2,261 people entered their predictions. Mashable wrote it up like this: "Facebook IPO: Did Twitter just give us closing price?"

In short: NO.

What did they think Facebook would close at? $54.
What did Facebook close at? Exactly $38.

And I note that $38 should have an asterisk.

Only 26 of the 2,261 predictors offered $38 as the company's shares closing price. And as you can see in the chart above, the distribution of predictions were concentrated around $50 with a substantial number of people predicting a very high closing price. Put it this way: almost as many people predicted a close of $80 (21) as predicted a close of $38.

The big news of the day (the week? the month?) is that everyday investors can now buy shares of Facebook. Quiet. Please sit down, everyone. Ladies and gentlemen, quiet please. Please! I am holding the conch shell. I am holding the conch shell.



Everyone wants to have their say about Facebook today, even if there isn't too much more to say than there was yesterday. We know this is a company with an astounding number of very engaged users. We know that some people love it and other people find it annoying. We know their ads have a lower than average clickthrough rate, but that it seems like they have the potential to build the world's most formidable advertising business. We know they make some money but not nearly as much as companies that have much smaller market values.

That's the state of play. We do have a little new data: the stock didn't go wild. It priced at $38, debuted at $42, and is hanging around $41 as I write. That's technically what is supposed to happen, but practically never what happens, so people are a bit confused as to what that all means. I'm trying not to draw any huge conclusions from a few hours of trading.

More broadly, though, there are some interesting cultural, corporate, and catty things to speculate on, now that Facebook is public.

• Will people keep "sharing" as much as they do now?
  
The most basic assumption of the social media game is that people will keep sharing ever more and ever more. If only because it is so deeply ingrained, I think it's worth examining that idea. What if it doesn't turn out to be true? What if sharing peaks in 2013 or 2014 simply because people run out of time or because of some newfound sense of privacy or because some subtle cultural shift occurs? None of these things are impossible, though they would be running against the current trends.

One thing to note: people tend to think about privacy as a function of the amount of information that people share. So, if people are sharing a lot on Facebook, the idea is that they are OK with the privacy tradeoff or that they are fine with whatever information is available about them. But NYU philosopher Helen Nissenbaum sees things differently: she sees privacy as that information ending up in places you did not expect. So, as time goes on, all the information you've put on Facebook could still end up in a place you did not expect it to. That means that the aggregate amount of stuff that could lead to a Nissenbaum privacy violation continues to grow, particularly as people change contexts from college to work, say, or single to married.


• What are the Facebook natives going to think about online life?
  
There are hundreds of millions of people who are coming to Facebook at about the same time that they are coming to the Internet. They've always had a social web. They don't remember AOL. Most don't live in the United States. I'm not sure that those of us who have been on the Internet since 1993 are going to be able to accurately predict their online behavior or what they'll think about online life. Every day of their lives is going to be cataloged partially on Facebook, so what happens when they get older? Do they reject the service they grew up on or does it become a permanent social layer in their brains? I don't know.


• What are the mobile natives going to do with their time?
For just about everyone reading this post, your experience of the Internet began on a computer. For hundreds of millions in the developing world, this is not the case. Their primary or sole means of access to the Interweb is through a mobile device. I have a long running debate with Paul Kedrosky about what these people are going to do as the incomes of their countries rise fall and computer prices continue to decline. I think they'll switch to the old mouse-and-keyboard style as soon as they can afford them. Paul and many others think that they'll remain mobile device lovers. Why's that matter for Facebook? They're much stronger on the web than on mobile, most (but not all) people think

• If and when will Facebook's unusually centralized corporate structure be tested?
  Not all corporate structures are alike and as Matt Yglesias explained a while back, Facebook's power is concentrated in one person: Mark Zuckerberg. Is this going to become an issue for investors or are they all too happy having Zuck running the show?

• What are the thousands of new rich people going to spend their money on?
  If you live around the Bay Area, this is a big question. It won't play out immediately, but everything from the businesses they fund to the restaurants they frequent to the apartments they buy will be transformed by today's big event. New companies, new neighborhoods, new ideas about how cities should be run: we'll get all that and more as these people make their way further from Palo Alto's HQ, the money rippling reality before them.
  
  
• What will happen to other social media stocks?
  As Facebook debuted, shares of GroupOn and Zynga took a nosedive. Perhaps that's because investors who wanted exposure to social media now can get in on the main course instead of messing around with the side dishes. It'll be interesting to watch what happens as time goes on.
  
  
• How will being a public company change Facebook's culture?
  Some people think: not much. In fact, one of the first articles out of the gates today was about how after Mark Zuckerberg rang the NASDAQ opening bell from Palo Alto, everyone whooped it up for 10 minutes and then went back to work. But this is a long game here, and you never know how the company's new orientation might impact who is interested in working for them or how they have to shape their products to meet investor expectations. (Then again, see the point above, re: company control.)
  
  
• How will the Google and Facebook competition evolve?
  Last year, Google was the company stepping all over Facebook's turf with the launch of Google Plus and the social layer it represented. Will Facebook start to push into Google's information organization and discovery territory in 2012 or 2013? And for what it's worth, Google's actually up 7 percent since the Facebook's IPO.
  

I do not think that I can add anything to this video, but I would like to thank its creators, who have brought so much joy to me on this day. "Most of the people in this video have never met face to face," the video maintains. "We are a global family of people committed to inspiring and empowering each other via Facebook."

Don't miss the song's Facebook page, either.

OK, I do have one thing to say about this video. This is a celebrity singalong from our universe in which everyone sort of behaves like a celebrity. It's "We Are the World" multiplied by The Cult of the Amateur and raised to the power of Facebook's opening-day share price.

Via @normative

Your Google search experience is about to change.

No, don't worry, it's not another social integration. The latest update has nothing to do with Facebook and everything to do with Google's core strengths of organizing information so that you can find it faster.

Now, when you search certain things, say, Tom Cruise, a box will pop up in the right column of your search with structured data about the topic. Google can identify 500 million people, places, and things and can serve up a custom selection of data based on the nature of the noun.

Google knows that you are very likely to want to know certain things about Tom Cruise (e.g. his height) and other things about Bill Gates (his net worth) and other things about astronaut Don Pettit (which Shuttle missions he flew).

How good is Google at both guessing what you want to know and having that information in its databases? In some cases, the company is really good. "Based on the other things that people are looking for when they are looking for Tom Cruise, our knowledge graph is going to show you 39 percent of the answers to the next thing you might be looking for," said Johanna Wright, director of product management for The Google Knowledge Graph, which is what the company is calling this new feature.

In the online world, there is this relentless obsession with relevance. Everything has to be relevant, relevant, relevant. The way the word is normally translated in common speech is that something is relevant when you're interested in it at a particular time and place. Facebook's news feed is all about showing you things that are relevant to you from your friends. Google wants to show you the most relevant pages for a given search. And, to my mind, they actually do a pretty good job doing these things.

As businesses, though, what they really want to do is serve up the most relevant advertisements. That is to say, they want to serve up the ads that you're most likely to click on. So they refine and refine and refine the algorithms they've got with more and more and more data in the quest to find ads that people want to click on.

It's fair to ask, I think, how are they doing? One indication comes courtesy of this infographic that these marketers created showing the differences between Facebook and Google's ad networks. It contains three remarkable stats about clickthrough rate (CTR), which is the percentage of the time a user clicks on an online advertisement. The average, these marketers say, is about 0.1 percent. Facebook's CTR is below average at 0.051 percent and Google's is above average 0.4 percent.

While these differences are meaningful and say something powerful about Google and Facebook, let's do the math on those percentages to see how relevant the ads you're seeing really are. For Google, people are clicking on about 1 of every 250 ads they see while searching. For the average, it's 1 out of every 1,000 ads. And for Facebook, people are only clicking once every 1,961 ads they see.

That's the reality of relevance in the online world -- and, not so incidentally, while companies are so eager to scoop up any data they can to increase the likelihood that you'll click.

A preview of what the next wave of anti-corporate activism might look like. Call it Big Dada: speaking noise to power.

Through the 1990s, a practice called "culture jamming" grew in popularity and sophistication. It aimed to disrupt consumer culture by transforming corporate advertising with subversive messages. So, as in the example above, a Coca Cola sign has been defaced to note the company's other imperative aside from love. Another canonical example was current BuzzFeed chief Jonah Peretti's 2001 attempt to order a pair of Nike's through the company's website emblazoned with the word, "sweatshop." Culture jammers would use the power of brands against themselves. Their most famous organ remains the magazine AdBusters, which is widely credited with helping jumpstart Occupy Wall Street last year.

Culture jammers capitalized on the general feeling of many on the American (and global) left that corporations had (and have) too much power and that one very powerful expression of that power was advertising. Advertisements seemed to have mythic influence that could get people to do all kinds of things from buying Hummers and McMansions to starving themselves to attain fashion-model thinness.

Active culture jamming was always a niche activity, but failing active engagement with brand transformations, ignorance was considered the next best policy. Better to skip past commercials with Tivo or stick to NPR than watch or listen to the ads on these broadcast media. Being ignorant of advertising has been considered a moral good; it meant that one was not in sway to the corporate paradigm, etc, etc. The underlying idea is that the activist position is to transform or ignore corporate assets and advertising.

Fast forward to our world in which an increasing amount of advertising runs online. The old logic of culture jamming would say that anticorporate activists should run ad blockers or perhaps something like the (now outdated) Firefox extension, Add-Art, which replaced corporate callouts with curated art.

But the system of advertising has changed in the online world. First, because of the private nature of the browsing experience, there is no way to transform ads for others' political consumption.

Second, Google and Facebook ads are measured on what's called a cost-per-click basis. Advertisers are charged not by how many people see their ads, but by how many people click on them. That means that the old method of passive resistance to corporate power -- ignoring ads -- costs the advertisers nothing. In fact, it makes the delivery of those ads more efficient. Advertisers' dollars get spent on those who find their ads "relevant" and are open to their marketing methods. And because of the private nature of the browsing experience, there's no real way to deface or transform an ad as a political statement to others. Whatever personal pleasure one might find in Add-Art, it's not doing anything in the societal realm. (There are anti-corporate memes, sure, but those would not be a direct response to the ads that Bank of America runs when you search for mortgages.)

I foresee that activists might find the best way to disrupt corporate power on the Internet is to be begin interacting with the ads they're being shown and muddying the data that's being collected.

The counterintuitive logic of online advertising is that any time someone clicks on an ad, it costs the advertiser money. So, clicking on any, say, mortgage-related Google ad, would cost the company that placed it more than $1, according to current pricing. Other banking-related keywords are more expensive, too. "Jumbo mortgage" has an average cost-per-click of $2.42 (and you'll find Citi, Union, and Fremont banks advertising on the search). "Mortgage calculator" goes for $5 (presumably because those searches are more serious). One person's clicks, of course, don't mean much. But a million people's clicks would. Tens of millions of clicks would. And this is a kind of online activism that's closer in nature to Anonymous' famed distributed denial-of-service attacks than to protesting in the streets. It's something people could participate in without leaving their computers and it would not be hard to write tools that would help activists coordinate their actions.

In the commercial world, already something like 10 percent of all clicks on Google ads are perpetrated by various bad actors, usually bots or spyware that are trying to make a tiny bit of money for delivering clicks to Google. The company has to actually refund that money to advertisers, which means it's probably a vast sum when you consider Google's $37 billion of 2011 revenue. (Consider that the Move Your Money project claims they got about $50 million moved out of big banks to credit unions.)

This form of activism, however, wouldn't be click fraud if it weren't perpetrated by machines. It's hard to see how it could be against the law or even against Google's terms of service. It's "window browsing" as activism, sucking up corporate resources as a political act.

But beyond the immediate financial impact this kind of action could have on marketing budgets, if the collective action became large enough, it could begin to impact the quality of the data that Google and other data intermediaries are collecting about each and every Internet user. If enough people started to seem interested in home mortgages who were not actually interested in home mortgages, it might start to disrupt their ability to efficiently target users with behavioral advertising. This would be statistical noisemaking as a form of protest. 

After the attention that Occupy Wall Street garnered last year, the attempts to revive the movement in May have not caught fire. As they always have, activists will keep trying new things. How long before they realize that many businesses are valuable more for their data than their storefronts? Banks are troves of data being mined for profitable strategies. How long before activists see that making their data harder to analyze could be a political tool?

Call it Big Dada: speaking noise to power.  

The unlikely story of a company that built a business selling the recent torrent of digital photos.

The stock photo company Shutterstock has grown tremendously as companies of all sorts realize that they're in the media business. And if you're in the media business, you need visuals. The incumbents in the stock photo space, like Corbis and Getty Images, get expensive if you need to illustrate hundreds of pages on a website.

So, Shutterstock came along with an all-you-can-eat subscription model and said, "Here, use tons of photos from this library of 19 million images." Some of the images are cheesy, but they serve the role that clipart used to: filling a space that you know needs an image with something vaguely topical (see above).

Now, after doubling revenue growth over the last two years, the company is preparing for an IPO.

It's an interesting game that Shutterstock is playing. Individual customers pay an average of about $3 per image. That's dirt cheap, but they make up for it on volume, bringing in $120 million of revenue in 2011. On the producer side, my read of their SEC filing is that they paid out $39.3 million in royalties to 35,000 contributors. So the mean contributor is making something like $1,100 a year by posting their work on the site. (I don't know exactly what the distribution looks like; we only know that no entity received more than 10 percent of the royalties paid out.)

This is clearly a buyer's market. In fact, it's amazing that some entity, i.e. Shutterstock, has been able to build a nine-figure business on the flood of digital imagery emanating from DSLRs across the world given that buyers aren't paying much and sellers aren't making much. They've come up with a cost structure that's low enough to enable them to turn a decent profit. The company's had net income of around $20 million a year since 2008.

From an investment perspective, the most obvious red flag, though, is that their revenues have more than doubled in the last several years, but their net income has been stagnant. They're having to spend a lot more money on sales and marketing than they did back in 2008. If that trend continues, something will have to change on the revenue side.

Even the most cherished skills for manipulating our machines eventually lose their utility.

At some point in the not-too-distant past, the key technological moment in a teenager's life might have been when she learned how to depress the clutch with her left foot, change her car's gear with her right hand, while giving the engine gas with her right foot. As the driver improved, the action became automatic and if you were a particular kind of dumb, rural teenager (like myself), you may have tried to see how fast you could get your car going in a given direction. The keys to this (I may have discovered) are when and how you shift the gears. I felt much mildly unsafe joy in getting from 0 to 60 as quickly as possible in my little Ford Escort.

Nowadays, though, more than 90 percent of American cars come with automatic transmissions. And the deskilling of teen drivers, I'm sure, has begun. One more skill, like efficient rotary phone dialing, will go missing and more more system will become a little easier to use and more opaque.

So it was with great nostalgia that I watched this incredible video of the hundreds of gear shifting clips from all five Fast and Furious movies. I've never actually seen these films and luckily, now that we have this YouTube video, I'll never have to. Imagine what this will look like to the kids of the future, almost all of whom will not know how to "drive stick."

While climate change may be complex and difficult terrain, rediscovering our industrial infrastructure is compelling.

A coal mine in Utah (Reuters).

The Rio+20 UN summit is just around the corner, the latest in a decades-long string of international meetings that attempt to address one of the world's greatest and most global environmental problems.

What's that? Your eyes have already glazed over? Well, you're not alone. I just spent the last couple of days in Seoul for the Global Green Growth Institute Summit, where I spoke during a session on green journalism. A common refrain from both the speakers and the audience was that that people were tired of hearing the same jeremiads about greenhouse gas concentrations, sea level rise, and government panels. Even people who care deeply about the environment are fatigued. This is a particularly acute problem on the Internet where the distribution of a story largely depends on readers to share the narrative with their friends through social media. The standard climate change narratives are not shareable.

But to me the most interesting stories to tell about climate change have never been attempts to elucidate the worst-case scenarios. As an organizing narrative, what climate change offered me was a reason to rediscover and reimagine the world's basic infrastructure. Want to radically improve the efficiency of the transportation system? Well, first you have to understand how and why Americans built the system that we have. You have to ask: What problems were our forebears trying to solve?

For people who grew up in the 1980s and 1990s, this is a fascinating topic because we came into a world that had effectively covered its tracks. By the logic of the system, making the industrial processes that power the world opaque was good, so we don't see them in our daily lives. As an early 21^st century American, it is easy to be completely ignorant of the basic systems -- food, water, energy -- that make modern life possible. You just don't have to know.

I think there's a perception that people don't want to read stories about the innards of industrial life, but I've never had a hard time getting people to look at and share these narratives. Take a look at Reddit's Today I Learned section. Among the miscellany, you often find factlets about how the 20^th century's big technological systems work.  Which makes sense because there is just so much to know about the complex networks that deliver what we need. When you really think about everything that needs to happen for a piece of coal in Wyoming to become the electricity that flows into your phone, it's stunning. It doesn't make me mad or depressed, even though burning the coal emits carbon. Rather, I'm filled with awe about the achievements of previous generations, and maybe with some hope that our generation can accomplish something equally ambitious.

Readers aren't stupid. They know when your product is cheap.

For a time, people measured site 'traffic' by the number of page views on that site. So, any time someone opened a page on that publication, it counted as one. Shortly thereafter, people started juicing the pageview stats by throwing up a bunch of pictures and asking people to click through them. It was a lot easier to generate 20 pageviews with 20 photos than it was to bring 20 people to the site by other means.

Of course, the fact that these pageviews are not all worth the same is obvious to everyone: readers, writers, editors, advertisers, advertising agencies, etc. So, many forward-looking media companies like Gawker went away from pageview metrics back in early 2010. The company's head Nick Denton wanted to focus on unique visitors to his site. Many of us have followed suit.

And yet still, today, nearly halfway through 2012, we find this story on The Atlantic Wire. The president of the Washington Post, Steve Hills, told his team that "awards 'don't matter' [and] urged more traffic-driving slideshows."

Now, I've got nothing against slideshows. At their best, I see them as a kind of horizontal storytelling. They are a tool you can deploy to tell certain stories. In fact, as storytelling widgets, I think they're actually underexploited. You can embed them as a sidebar to convey some complicated set of ideas without interrupting the main flow of a narrative. And I've got nothing against a well-curated set of images a la our own In Focus or BuzzFeed's random weirdness.

But that's not what the WaPo's slideshows are all about. Instead, they are seen as a cheap and fast way to produce "traffic." The problem is that they are not producing "traffic" -- which in any other context would mean the number of people in a space -- they are producing page views. This is not a simply academic distinction. The company's president is calling on his workers to juke the stats, effectively. These companies want you to think that more pageviews equal a larger, more engaged audience, but that's a quantitatively and qualitatively shaky proposition.

Quantitatively, sites vary widely in their page views to visitor ratios, and I can tell you from experience that it is much, much easier to drive up the former than the latter. So, when companies are in trouble, what do you think they try to do?

If you're trying to juice page views, your staff will ineluctably be forced to make galleries. Where else can they get a 10x or 20x multiplier on their work? I can guarantee you that will not help you break the kinds of stories or do the kinds of analysis that will keep people coming back. Not only that, but it's demoralizing to your best people, the ones who want to be out there producing their best work.

Worse, readers may click through your slideshow, but they'll hate you a liiitttle bit more than they did when they got to the site. And I bet they'll feel the same way about whatever advertiser was unlucky enough to get stuck on the page with some stupid thing that a reporter did with a little bit of hate in his heart and fingertips.

That won't be visible to you in your analytics, but what reader of the Internet has not felt that pang: "This site doesn't really value me or my time." You can get a page view spike that's actually a negative for your brand. And the more the slideshow spreads because of a clever headline or just because the topic is hot, the farther that brand damage spreads. Congratulations! You juiced the stats with an invisible poison!

I'm sympathetic to the business concerns of the media industry. I really am. But this myth that slideshows are the path to salvation has got to be put into a rocket and sent hurtling into the sun. People know when your product is cheap; there is no "trick" of the web. The sad truth is that to win on the Internet you have to do good reporting and analysis, write great headlines, empower individual staffers to embed themselves in communities that can serve up scoops and distribute finished stories, and understand the social ecosystems that bring visitors to your site.

Just one more pebble of evidence on a growing pile that Silicon Valley has been too focused on small ideas in the social space.

In a blog post today, an early investor in the warehouse robot company, Kiva, detailed what he learned from the venture.

If you haven't been watching the logistics space, Kiva makes squat little robots that work in vast teams in e-commerce fulfillment centers. Instead of humans wandering into vast stacks of merchandise, robots bring that merchandise to the workers. The robots carry out the work according to constantly evolving algorithms that maximize the efficiency of the operation.

They are a very big idea in logistics -- and one that founder Mick Mountz built a company around that Amazon purchased for $775 million in a deal that closed this week.

In today's blog post, Ajay Agarwal of Bain Capital Ventures noted that Mountz was unable to find funding in Silicon Valley, despite the idea's now-fulfilled promise.

There have been several blog posts, most notably by Peter Thiel and Founders Fund, discussing the venture community's lack of desire to fund transformational companies -those with disruptive technologies taking on big problems.  Mick saw this firsthand. When Mick first started Kiva shortly after the bubble burst, he was unable to raise funding on Sand Hill Road.  This ultimately caused him to move to Boston, where he raised his angel round and eventually his round from Bain Capital Ventures...

The truth is, Kiva simply wasn't a company that could be cranked out in weeks with some seed money, and the technical obstacles inherent in building a solution like this forced Kiva to invest years working on the solution pre GA. However, once they built a working and viable solution, they had the advantage of significant IP and few direct competitors.

There are echoes in Agarwal's post of the no-idea-too-small attitude that I discussed in my recent essay, "The Jig Is Up: Time to Get Past Facebook and Invent a New Future."

A small group of dolphins in Brazil have learned to steer fish into humans' nets, but no one knows why.

Let's cut to the chase here: although whales are majestic, dolphins are the best. I love that they work in teams and that they develop something like "culture," in which specific groups of dolphins behave idiosyncratically in ways that are intelligent.

Here's the latest case-in-point. Down in Brazil, a group of dolphins aid the local fishermen! This has been going on for generations, despite unclear benefits for the marine mammals. Researchers looked at the practice in a new paper in the journal Biology Letters entitled, "The structure of a bottlenose dolphin society is coupled to a unique foraging cooperation with artisanal fishermen."

For the record, this seems like the ultimate hipster foodie trump card: "My fish was caught by artisanal Brazilian fishermen with the help of a unique group of... intelligent dolphins." Here's how ScienceNow describes the situation:

Every autumn, lucky visitors to Laguna, Brazil, which is situated around a narrow lagoon on the Atlantic Ocean, catch an odd sight. Here, resident bottlenose dolphins (Tursiops truncatus) frequently turn sheepdog, herding schools of small, silver fish called mullets toward the shore--and, it turns out, toward lines of wading fishers. As soon as the dolphins get close to their human companions, they give the signal, slapping their heads or tails against the surf. In an instant, the fishers cast their nets, catching dozens of frenzied mullet.

Even more intriguingly, only a subgroup of the 50 or so dolphins have picked up the human habit. A third help out, while the rest steer clear of the people. Why do some dolphins help out? And why do others avoid the behavior?

To figure that out, the researchers spent two years watching the dolphins in the water, seeing which animals hung out together and for how long. They found that the animals that the social network of dolphins that help humans spends more time together and more tightly associated than the other animals. They still don't know why this is or how precisely younger dolphins figure out what to do, but the fact that there are identifiable characteristics to the human-helping social network is fascinating. In the graph below, the human helpers are the white circles.

A pioneering ship proves a point about the possibilities of renewable energy.

For the past 19 months, the Turanor PlanetSolar has been on a voyage around the world powered only by 38,000 SunPower solar cells. On May 4, the boat will return to Monaco, completing a trip that it began in September of 2010. The 537 square meters of solar panels power six banks of lithium-ion batteries. On a good day towards the end of the trip, they could charge up the batteries by a full 50 percent. On a bad day, they might only get 10 percent more juice before dusk.

The trip was not intended to prove that solar-powered ships are ready for the commercial primetime, but like many first-time journeys, to prove to future engineers that the feat could be accomplished and now need only to be optimized. Perhaps they also provided new (and green!) inspiration for a sequel to Kevin Costner's Waterworld.

The boat's crew recorded the trip in impressive detail, blogging and posting photos all along the way. Their adventures were funded primarily by the Swiss watchmaker, Candino, and the German energy company, Immosolar.

Dairy scientists are the Gregor Mendels of the genomics age, developing new methods for understanding the link between genes and living things, all while quadrupling the average cow's milk production since your parents were born.

Reuters.

While there are more than 8 million Holstein dairy cows in the United States, there is exactly one bull that has been scientifically calculated to be the very best in the land. He goes by the name of Badger-Bluff Fanny Freddie.

Already, Badger-Bluff Fanny Freddie has 346 daughters who are on the books and thousands more that will be added to his progeny count when they start producing milk. This is quite a career for a young animal: He was only born in 2004.

There is a reason, of course, that the semen that Badger-Bluff Fanny Freddie produces has become such a hot commodity in what one artificial-insemination company calls "today's fast paced cattle semen market." In January of 2009, before he had a single daughter producing milk, the United States Department of Agriculture took a look at his lineage and more than 50,000 markers on his genome and declared him the best bull in the land. And, three years and 346 milk- and data-providing daughters later, it turns out that they were right.

"When Freddie [as he is known] had no daughter records our equations predicted from his DNA that he would be the best bull," USDA research geneticist Paul VanRaden emailed me with a detectable hint of pride. "Now he is the best progeny tested bull (as predicted)."

Data-driven predictions are responsible for a massive transformation of America's dairy cows. While other industries are just catching on to this whole "big data" thing, the animal sciences -- and dairy breeding in particular -- have been using large amounts of data since long before VanRaden was calculating the outsized genetic impact of the most sought-after bulls with a pencil and paper in the 1980s.

Dairy breeding is perfect for quantitative analysis. Pedigree records have been assiduously kept; relatively easy artificial insemination has helped centralized genetic information in a small number of key bulls since the 1960s; there are a relatively small and easily measurable number of traits -- milk production, fat in the milk, protein in the milk, longevity, udder quality -- that breeders want to optimize; each cow works for three or four years, which means that farmers invest thousands of dollars into each animal, so it's worth it to get the best semen money can buy. The economics push breeders to use the genetics.

The bull market (heh) can be reduced to one key statistic, lifetime net merit, though there are many nuances that the single number cannot capture. Net merit denotes the likely additive value of a bull's genetics. The number is actually denominated in dollars because it is an estimate of how much a bull's genetic material will likely improve the revenue from a given cow. A very complicated equation weights all of the factors that go into dairy breeding and -- voila -- you come out with this single number. For example, a bull that could help a cow make an extra 1000 pounds of milk over her lifetime only gets an increase of $1 in net merit while a bull who will help that same cow produce a pound more protein will get $3.41 more in net merit. An increase of a single month of predicted productive life yields $35 more.

When you add it all up, Badger-Fluff Fanny Freddie has a net merit of $792. No other proven sire ranks above $750 and only seven bulls in the country rank above $700. One might assume that this is largely because the bull can help the cows make more milk, but it's not! While breeders used to select for greater milk production, that's no longer considered the most important trait. For example, the number three bull in America is named Ensenada Taboo Planet-Et. His predicted transmitting ability for milk production is +2323, more than 1100 pounds greater than Freddie. His offspring's milk will likely containmore protein and fat as well. But his daughters' productive life would be shorter and their pregnancy rate is lower. And these factors, as well as some traits related to the hypothetical daughters' size and udder quality, trump Planet's impressive production stats.

One reason for the change in breeding emphasis is that our cows already produce tremendous amounts of milk relative to their forbears. In 1942, when my father was born, the average dairy cow produced less than 5,000 pounds of milk in its lifetime. Now, the average cow produces over 21,000 pounds of milk. At the same time, the number of dairy cows has decreased from a high of 25 million around the end of World War II to fewer than nine million today. This is an indisputable environmental win as fewer cows create less methane, a potent greenhouse gas, and require less land.

At the same time, it turns out that cow genomes are more complex than we thought: as milk production amps up, fertility drops. There's an art to balancing all the traits that go into optimizing a herd.

While we may worry about the use of antibiotics to stimulate animal growth or the use of hormones to increase milk production by up to 25 percent, most of the increase in the pounds of milk an animal puts out over the pastoral days of yore come from the genetic changes that we've wrought within these animals. It doesn't matter how the cow is raised -- in an idyllic pasture or a feedlot -- either way, the animal of 2012 is not the animal of 1940 or 1980 or even 2000. A group of USDA and University of Minnesota scientists calculated that 22 percent of the genome of Holstein cattle has been altered by human selection over the last 40 years.

In a sense that's very real, information itself has transformed these animals. The information did not accomplish this feat on its own, of course. All of this technological and scientific change is occurring within the social context of American capitalism. Over the last few decades, the number of dairies has collapsed and the size of herds has increased. These larger operations are factory farms that are built to squeeze inefficiencies out of the system to generate profits. They benefit from economies of scale that allow them to bring in genomic specialists and use more expensive bull semen.

No matter how you apportion the praise or blame, the net effect is the same. Thousands of years of qualitative breeding on family-run farms begat cows producing a few thousand pounds of milk in their lifetimes; a mere 70 years of quantitative breeding optimized to suit corporate imperatives quadrupled what all previous civilization had accomplished. And the crazy thing is, we're at the cusp of a new era in which genomic data starts to compress the cycle of trait improvement, accelerating our path towards the perfect milk-production machine, also known as the Holstein dairy cow.

***

A botanical drawing of Mendel's pea plants. The Field Museum.

There are no more famous experiments in genetics than the ones undertaken by the Austrian monk Gregor Mendel on five acres in what is now the Czech Republic from 1856 to 1863. Mendel bred 29,000 pea plants and discovered the most basic rules of genetics without any knowledge of the underlying biochemical mechanics.

Smack dab in the middle of Mendel's experiments, Charles Darwin's Origin of Species was published, but we don't have any record of intellectual mingling between the two men. Even the idea of a gene as an irreducible unit of inheritance wasn't presented until 30 years after Mendel began his experiments. The term and field of genetics would not be fleshed out until William Bateson and company came along in the early 1900s. And its form, DNA, would not be proposed by James Watson and Francis Crick with indispensable help from Rosalind Franklin until 90 years after his last pea plant died. All this to say: Mendel was ahead of his time.

What he had going for him was a dedication to data, to quantification. His fundamental insight was statistical.

Here's the simple version of what he did. Mendel took pea plants that reliably produced purple or white flowers when they self-pollinated. Then he crossbred them, carefully controlling how the plants reproduced. Now, one might expect that if you breed a pea plant with a purple flower and a pea plant with a white flower, you'd get progeny that were sort of mauve, a mix of the two colors. But what Mendel found instead is that you either got purple flowers or white flowers. Even more amazingly, sometimes breeding two purple flowers would yield a white flower. Among the first generation of crossbreeds, the mix of flower colors occurred at a roughly constant ratio of about 3:1, purple to white. If the traits of two plants were being mixed to generate the next generation, how could two purple flowers yield a white flower? And why would this ratio arise?

Mendel took a conceptual leap and hypothesized that the plants had two possible copies of its plans (i.e. genes) to make flower color (or any of six other traits he analyzed). If the plant received two of the dominant plan (purple), the flowers would, of course, be purple. If it received one of each, the dominant plan would still reign. But if the plant received two recessive plans, then the flowers of that pea would be white.

The monk turned out to be right. For traits controlled by a single gene, things really do work as he predicted. Mendel's insights became part of the central dogma of genetics. You can use the statistical method he used to calculate how likely someone is to get sickle cell anemia from her parents. In most genetics classes, Mendel is where it all starts and for good reason.

But it turns out that Mendel's version of things doesn't actually give a very clear picture of the kinds of things we care about most. "Mendel studied a few traits that happened to be controlled by a single gene, making the probabilities easier to figure out," the USDA's VanRaden said. "Animal breeders for many decades have used models that assume most traits are influenced by thousands of genes with very small effects. Some [individual] genes do have detectable effects, but many studies of plant and animal traits conclude that most of the genetic variation is from many little effects."

For dairy cows -- or humans, for that matter -- it's just not as simple as the dominant-recessive single-gene paradigm that Mendel created. In fact, Mendel picked his model organism well. Its simplicity allowed him to focus in on the simplest possible genetic model and figure it out. He could easily manipulate the plant breeding; he could observe key traits of the plant; and these traits happened to be controlled by a single gene, so the math lay within human computational range. Pea plants were perfect for studying the basics of genetics.

With that in mind, allow me to suggest, then, that the dairy farmers of America, and the geneticists who work with them, are the Mendels of the genomic age. That makes the dairy cow the pea plant of this exciting new time in biology. Last week in the Proceedings of the National Academy of Science, two of the most successful bulls of all time had their genomes published.

This is a landmark in dairy herd genomics, but it's most significant as a sign that while genomics remains mostly a curiosity for humans, it's already coming of age when it comes to cattle. It's telling that the cutting-edge genomics company Illumina has precisely one applied market: animal science. They make a chip that measures 50,000 markers on the cow genome for attributes that control the economically important functions of those animals.

A snippet from Illumina's animal science fact sheet.

***

Mendel may have worked with plants, the rules he revealed turned out to be universal for all living things. The same could be true of the statistical rules that dairy scientists are learning about how to match up genomic data with the physical attributes they generate. The statistical rules that reflect the way dozens or hundreds of genes come together to make a cow likely to develop mastitis, say, may be formally similar to the rules that govern what makes people susceptible to schizophrenia or prone to living for a long time. Researchers like the University of Queensland's Peter Visscher are bringing the lessons of animal science to bear on our favorite animal, ourselves.

Want to live for a very long time? Well, we hope to discover the group of genes that are responsible for longevity. The problem is that you have genomic data over here and you have phenotypic data, i.e. how things actually are, over there. What you need, then, is some way of translating between these two realms. And it's that matrix, that series of transformations, that animal scientists have been working on for the past decade. 

It turned out they were in the perfect spot to look for statistical rules. They had databases of old and new bull semen. They had old and new production data. In essence, it wasn't that difficult to generate rules for transforming genomic data into real-world predictions. Despite -- or because of -- the effectiveness of traditional breeding techniques, molecular biology has been applied in the field for years in different ways. Given that breeders were trying to discover bulls' hidden genetic profiles by evaluating the traits in their offspring that could be measured, it just made sense to start generating direct data about the animals' genomes.

"Each of the bulls on the sire list, we have 50,000 genetic markers. Most of those, we have 700,000," the USDA's VanRaden said. "Every month we get another 12,000 new calves, the DNA readings come in and we send the predictions out. We have a total of 200,000 animals with DNA analysis. That's why it's been so easy. We had such a good phenotype file and we had DNA stored on all these bulls."

They had all that information because for decades, scientists have been taking data from cows to figure out which bulls produced the best offspring. Typically, a bull with a promising pedigree would reach sexual maturity and his semen would be used to impregnate a selection of about 50 test cows. Those daughters would grow up and start producing milk a few years later. The data from those cows would be used to calculate the value of that now "proven" bull. People called the process "progeny testing" and it did not require that breeders knew the exact genetic makeup of a bull. Instead, scientists and breeders could simply say: We do not know the underlying constellations of genes that make this bull so valuable, but we do know how much milk his kids will produce. They learned to use that data to predict who the best bulls were.

That meant that some bulls became incredibly sought after. The number two bull of the last century, Pawnee Farm Arlinda Chief, had more than 16,000 daughters, 500,000 granddaughers, and 2 million great granddaughters. He's responsible for about 14 percent of all the genetic material in all Holsteins, USDA scientists estimate.

"[In the past], we combined performance data -- milk yield, protein yield, confirmation data -- with pedigree information, and ran it through a fairly sophisticated computing gobbledygook," another USDA scientist Curt Van Tassel told a group of dairy farmers. "It spit out at the other end predicted transmitting ability, predicted genetic values of whatever sort. Now what we're trying to do is tweak that black box by introducing genomic data."

There are many different ways you could model the mapping of 50,000 genetic markers onto a dozen performance traits, especially when you have to consider all kinds of environmental factors. So the dairy breeders have been developing and testing statistical models to take all this stuff into account and spit out good predictions of which bulls herd managers should ultimately select.The real promise is not that genomic data will actually be better than the ground-truth information generated from real offspring (though it might be), but rather that the estimates will be close enough to real but save 3 to 4 years per generation. If you don't have to wait for daughters to start cranking out milk, then you can shave those years off the improvement cycle, speeding it up several times.

Nowadays breeders can choose between "genomic bulls," which have been evaluated based purely on their genes and "proven bulls," for which real world data is available. Discussions among dairy breeders show that many are beginning to mix in younger bulls with good-looking genomic data into the breeding regimens. How well has it gone? The first of the bulls who were bred from their genetic profiles alone, are receiving their initial production data. So far, it seems as if the genomic estimates were a little high, but more accurate than traditional methods alone.

The unique dataset and success of dairy breeders now has other scientists sniffing around their findings. Leonid Kruglyak, a genomics professor at Princeton, told me that "a lot of the statistical techniques and methodology" that connect phenotype and genotype were developed by animal breeders. In a sense, they are like codebreakers. If you know the rules of encoding. it's not difficult to put information in one end and have it pop out the other as a code. But if you're starting with the code, that's a brutally difficult problem. And it's the one that diary geneticists have been working on.

Their work could reach outside the medical realm to help us understand human's evolution as well. For example, Kruglyak said, human population geneticists want to figure out how to explain the remarkable lack of genetic variance between human beings. "The typical [genetic] variation among humans is one change in a thousand," he said. "Chimps, though they obviously have a much smaller population now, have several fold higher genetic diversity." How could this be? Researchers hypothesize that human beings once went through a bottleneck where there were very few humans relative both to the current human population and the chimp population. Few humans meant that the gene pool was limited at some point in the pre-historical but fairly recent past. We've never recovered the diversity we might have had.

***

The number-one ranked bull in the world. Kathy DeBruin.

It might seem that Badger-Bluff Fanny Freddie is the pinnacle of the Holstein bull. He's been the top bull since the day his genetic markers showed up in the USDA database and his real-world performance has backed up his genome's claims. But he's far from the best bull that science can imagine.

John Cole, yet another USDA animal improvement scientist, generated an estimate of the perfect bull by choosing the optimal observed genetic sequences and hypothetically combining them. He found that the optimal bull would have a net merit value of $7,515, which absolutely blows any current bull out of the water. In other words, we're nowhere near creating the perfect milk machine.

The problem, of course, is that genomes cannot really be cut and pasted together from the best bits. "When you go extremely far for one trait, you're going to upset some of the other traits," Vanraden said. Breeding is a messy (i.e. biological) process, no matter how technologically sophisticated the front end. After decades of breeding cows for milk production, people realized (to their dismay) that the ability to generate milk and the ability to have babies were negatively correlated. The more milk you tried to order up, the less babies your herd was likely to have. While we're nowhere near the hypothetical limit for Holstein bull value, we do now know that nature is not so easily transformed without some deleterious effects. We may have factory farms, but these machines are still flesh and blood.

Except for Badger-Fluff Fanny Freddie and his fellow bulls, that is. Freddie is a disembodied creature, an animal that is more important as data than as meat or muscle. Though he's been mentioned in thousands of web pages and dozens of trade industry articles, no one mentions where he was born or where the animal currently lives. He is, for all intents and purposes except for his own, genetic material that comes in the handy form of semen. His thousands of daughters will never smell him and his physical location doesn't matter to anyone. He will be replaced very soon by the next top bull, as subject to the pressures of our economic system as the last version of the iPhone.

It all started when I was looking at a photograph of an otter that was, BuzzFeed said, "disappointed that [I] never finished Infinite Jest." The merits of this otter's feelings aside (I got pretty far after all), the whole gallery of "extremely disappointed" animals caught my attention. Out of the 33 animal photos, only two of them had attributions. The rest seemed to emerge from some Internet froth only to be reassembled by Jack Shepherd into something that I have looked at least four times since it was posted two week ago.

It got me thinking: where were they getting all of these photos? And how were they using them without worrying about copyright? Naturally, I tweeted at @Buzzfeed about it, and received an email with an invitation to talk with the site's founder, Jonah Peretti.

First, Peretti told me this morning, BuzzFeed pays licensing fees to Reuters, AP, and Getty Images for the use of their libraries.

But a lot of what BuzzFeed traffics in -- the fun stuff, that is -- emerges on Tumblr or Pinterest or 4chan. Users of those sites surface photos that in some cases have been shared around the Internet for a decade. In those cases, even if BuzzFeed editors try to track down the creator, which Peretti assures me they do, they probably won't find whoever uploaded the photo of every obese cat.

Which leaves BuzzFeed in a bit of a bind. While Tumblr and Pinterest and any other site to which users upload are protected by the safe harbor provision of the Digital Millennium Copyright Act if their users upload stuff that they didn't create, BuzzFeed has no such protections. (Except in cases where BuzzFeed users upload content that they didn't create, which happens sometimes.)

So while Pinterest users are *explicitly encouraged* to snatch photos from all over the Internet, no matter who made them or under what hypothetical licensing agreement, BuzzFeed editors (as well as Atlantic editors) face a tougher set of circumstances. "I would love if every image contained some secret metadata and a way to license that image," Peretti told me. "But the practical reality is that it is pretty challenging, particularly in the web culture of animals and the images that spread on Pinterest and Tumblr."

And it's in cases like these that things really get interesting. With these kinds of posts, Peretti is willing to make a Fair Use argument that goes like this. First off, the Fair Use limitation and exception to exclusive copyright is notoriously fuzzy. Let's quote from Wikipedia on this one point because the explanation there is reasonable and understandable:

To justify the use as fair, one must demonstrate how it either advances knowledge or the progress of the arts through the addition of something new. A key consideration is the extent to which the use is interpreted as transformative, as opposed to merely derivative.

So, Peretti told me that he considers a BuzzFeed list -- its sequencing, framing, etc -- to be a transformative use of photos. That is to say, including that unattributed photo of the otter in that list was OK because its inclusion as an "extremely disappointed" animal transformed the nature of the photo.

"It's a question," Peretti said, "of when lots of little things add up to a transformation as opposed to a copyright violation."

The photo below shows an intelligent pug trying to understand how this theory might be interpreted in a court of law. Also I stole it from Buzzfeed but I renamed the dog in it Mr. Pookie Pants and I would like you to imagine that he, in fact, considers David Foster Wallace to be only a middleweight author within his generation. Furthermore, the dog's aunt had just died, with whom he was very close, so don't take his sadness as any sort of commentary on whether including photos in posts should be considered a transformation within Fair Use law. I genuinely don't know where the line is and am pretty sure that no one else does either.

What I do know is that it's very strange that Pinterest and Tumblr users don't have to play by the same rules that media editors do. And I also know that I don't want the DMCA safe harbor provisions to go away so much as I want Fair Use law to have some bounded space in which we can all work on this here Internet.

Or as Peretti put it, "Is it good for the world to have a broad definition of Fair Use where people can create new things that are transformative or that people can enjoy? I think it is good for the world."

Harvard's library appears set to push open access harder than its Ivy League rival.

Earlier this month, a special council to the Harvard library system sent a note to the school's entire faculty encouraging a broad range of measures to support open access journals, which are free and freely available. "Many large journal publishers have made the scholarly communication environment fiscally unsustainable and academically restrictive," they wrote.

Perhaps the strongest recommendation was that faculty should "consider submitting articles to open-access journals, or to ones that have reasonable, sustainable subscription costs." This would, in effect, "move prestige to open access," and away from the traditional, paid publications.

Earlier this week, Yale university student, Emmanuel Quartey, posted a video interview with the school's librarian, Susan Gibbons, in which he asked her about open-access publishing. Her response was far more ambivalent than the Harvard faculty council's. Though she noted that open-access journals are more accessible, she worried that asking younger faculty to publish in open-access (presumably less prestigious) journals could jeopardize their chances to attain tenure. In essence, prestige would stay put but tenure would move away from younger Yale professors. So, the library would continue to support both open and closed-access journals. You can read her full answer below or check out the video interview above.

Open access is a very complicated topic. On the one hand, from a faculty member or researcher's perspective, they go through all this work to make a discovery, to write about it. You'd want the broadest possible audience for your article or your book. You want people to read what it is you've discovered. So on the one hand, you want the greatest possible audience. On the other hand, if you think about the tenure process, how do you get tenure? You get tenure by, in part, publishing in the best journals. And until those journals are interested in an open access model, which really takes away their revenue stream, we have this tension going on. On the other hand, you want everyone to be able to read your research, but on the other hand, you need to go through the current steps that are necessary to reach tenure and achieve your promotion.

So the faculty have to make this decisions along the way to publish in an open access journal and give up perhaps some of the prestige that's associated with one of the more established journals. So, sometimes what you'll see is some of the junior faculty who are less inclined to publish in open access journals because they are focused on the career path and tenure track process. But once they get tenure, they feel like they have more freedom in participating in the open access movements going around.

So it's not just a scholarship issue. It's not just an issue between libraries and publishers. There is a whole other element to it that often people forget about: which is that the tenure process is tightly intertwined with the promotion process and publishing. Until that gets settled out, it isn't clear what is the best way to go. And you want a professor to be able to get tenure if possible. And you don't want to put something in the way that jeopardizes it. So from the library's perspective, we support both kinds of journals. We subscribe to those that still require payment but if there are open access journals, we'll make sure they are in the Orbis catalog as well. We don't have the mechanisms to preserve those open access journals that are out there on the web unless we downloaded every article, printed it, and bound the journal we don't have that archival copy, which is something that we're also very concerned about. There are a lot of issues at play here. I think too often people think of this in the dichotomy of pro-open access or against it. I think there are subtleties in the middle that need to be explored more.

Via Nicholas Bramble

Today, Google announced that their translation engine, which is premised on simple machine learning techniques multiplied by vast volumes of data, now receives 200 million users per day. The scale of the service spins out some crazy stats about Google's role in language today. Here's Franz Och, a research scientist at the company:

In a given day we translate roughly as much text as you'd find in 1 million books. To put it another way: what all the professional human translators in the world produce in a year, our system translates in roughly a single day. By this estimate, most of the translation on the planet is now done by Google Translate.

Of course, Och pays lip service to human translators "for nuanced or mission-critical translations," but when you just want to know the basics of what someone is saying, Google does the trick. Machine translation really is an amazing service and something it's easy to underestimate now that we have it.

A key question over the next six years is how far Google's current techniques can take them. The strategy for the last six years has been constant: MORE DATA. But even Peter Norvig, head of Google Research, admits that there are declining returns to the more-data game. Certainly, it doesn't appear that just adding more data is going to yield Gary Snyder's translations of Chinese poetry. Eventually, it seems to me, Google (or any other translation software) will have to start understanding (in some way) the semantic content of the words it is arranging. And that's a much harder AI problem to solve than the one that's brought you the wonders of Google Translate.

The government's use of drones in other countries has drawn scrutiny, but there are plenty of drones flying in American skies too.