The New York Observer calls Madrigal "for all intents and purposes, the perfect modern reporter." He co-founded Longshot magazine, a high-speed media experiment that garnered attention from The New York Times, The Wall Street Journal, and the BBC. While at Wired.com, he built Wired Science into one of the most popular blogs in the world. The site was nominated for best magazine blog by the MPA and best science Web site in the 2009 Webby Awards. He also co-founded Haiti ReWired, a groundbreaking community dedicated to the discussion of technology, infrastructure, and the future of Haiti.
He's spoken at Stanford, CalTech, Berkeley, SXSW, E3, and the National Renewable Energy Laboratory, and his writing was anthologized in Best Technology Writing 2010 (Yale University Press).
Madrigal is a visiting scholar at the University of California at Berkeley's Office for the History of Science and Technology. Born in Mexico City, he grew up in the exurbs north of Portland, Oregon, and now lives in Oakland.
Sears! Once the catalog king, then an eminent brick-and-mortar retailer, and now, perhaps, a real-estate holding company that leases out space for computers that power the cloud.
Data Center Knowledge reported today that Sears had created a new unit -- Ubiquity Critical Environments -- to look into repurposing its shuttered stores as datacenters, starting with this one in Chicago.
Yes, this is this week's sign that the 21st century is upon us.
Sears Holdings has a portfolio of 2.5 million square feet of retail space. Not all of it will be suitable for housing server farms, but some percentage of it will be. Ubiquity is tasked with figuring out which stores could be converted.
Right now, mall sites are out, but you never know. "I don't think the industry is yet ready for a mall-based data center," Ubiquity's manager told the site. "That may take some time. The stand-alone location is optimal."
(Imagine wandering an empty mall. Closed, closed, closed. But behind each grate, you hear the whir of a shard of the Internet. Finally, you come to an open storefront. It's a Starbucks and there are 30 IT guys at makeshift standup desks.)
Sears is considering other options for its closed stores: renting them out as "disaster recovery facilities (euphemistically: "business continuity centers") or leasing their roofs to wireless carriers.
Marc Whitten stands in front of an army of servers (Microsoft).
Watching the reveal of the Xbox One this week, one particular claim about Microsoft's new console caught my ear. Marc Whitten, the executive in charge of Xbox Live, the company's online gaming network, charted its historical progression.
"When we launched Xbox Live in 2002, it was powered by 500 severs. With the advent of the 360, that had grown to over 3,000," Whitten said. "Today, 15,000 servers power the modern Xbox Live experience."
Then Whitten said something extraordinary, "This year, we will have more than 300,000 servers for Xbox One, more than the entire world's computing power in 1999."
Now that's impressive! The statement even turned into an (unplanned?) applause line that tripped up Whitten's presentation. Because 1999 is not some distant date. It was the height of the dot-com bubble, after all. On an average home computer, you could play complex 3D games and download MP3s, edit video and mess around in Photoshop. Tens of millions of people had computers in their homes and Microsoft Office was nearly universal in business. Deep Blue had already beaten Garry Kasparaov!
And now, not even 15 years later, that same amount of information processing -- all the nuclear physics and climate simulations and videogames and spreadsheets and databases -- was being dedicated to running just one entertainment network, just one videogame network.
Stoners the world over were mumbling, vaguely, "Moore's Law, man. Moore's Law."
But then I started thinking: how did they figure this out? And can it be true? I contacted Microsoft's press people, who told me, "We do not have information to provide on these calculations."
Luckily, last year, Martin Hilbert and Priscila López estimated the world's information processing capacity at various points in time for a paper published in the journal Science: "The World's Technological Capacity to Store, Communicate, and Compute Information." I got a hold of Hilbert over email and asked him if Microsoft's assertion held up to inspection. Let me walk you through his answer.
First, because Hilbert and López were dealing with historical terrain, they used a measure called MIPS, short for million of instructions per second. They broke out the world's computation resources into two broad categories: 1) General-purpose computing tracks mostly the CPUs in personal computers and videogame consoles (see chart) and 2) Application-specific computing, which is composed of digital signal processors (say in a DVD player), microcontrollers, and GPUs. By those two measures, in 1999, the world had 180 billion MIPS in general-purpose computing power and
800 billion MIPS in application-specific computing power for a total of 980 billion MIPS. With me? OK.
So, nowadays, most measurement of computers is done in FLOPS, or floating-point operations per second. So, Hilbert had to use a conversion of 1 FLOP to 0.0141 IPS (alternatively, about 71 FLOPs per IPS). This operation, Hilbert admits, is a "questionable" (his word) assumption in these calculations, but it allows us to make some meaningful comparisons.
Hilbert said, let's create an upperbound, by taking the average performance of the bottom 100 supercomputers on the Top 500 supercomputer list, and imagine that Microsoft has 300,000 of them. Under those (improbable) conditions, they'd reach 300 billion MIPS, more than the general-purpose computing power of 1999, but not even a third of the total processing power available.
Of course, we know Microsoft is not deploying 300,000 top supercomputers, so their claim is very likely an exaggeration.
But here's the weird thing: It's not that big of an exaggeration, according to Hilbert. "Realistically, since they are using less powerful (but specialized) servers, and orienting ourselves on the computing powers that are common in the gaming industry," he said, "I think the reality is rather that the computing power of this cluster is equal to the world's total computing power in 1994 or the world's general-purpose computing power in 1996."
As he summed it up, "I'd say they are some 5 years off... but nevertheless very impressive!"
Because 1995 is less than 20 years ago. More than a quarter of American households already had a computer. This is not a comparison to the Apollo guidance computer or some IBM machine that used punch cards.
And now all of that power, all of it, resides in some cluster of computers served up by one company in Redmond, Washington, so that we can all play Call of Duty and watch movies together. How strange exponentiality (re)makes the world.
Now, the Science Channel has created a livestream of of a terrarium filled with cicadas crawling and sitting and flapping their wings and crawling some more.
Yick, ack, ugh! It is gross. But it is also impossible to stop watching. I warned you.
In late 1999, Microsoft created an ad for its upcoming 'Microsoft Reader' software. The headline blared, "This is a story about the future of reading," and underneath the story about the company's actual product, the marketers inserted a timeline based on "the best estimates of Microsoft researchers and developers" of what was going to happen to books in the future.
The Microsoft Reader product was unremarkable and did not drive a revolution in the book marketplace. But the predictions! They're fascinating, particularly in how they attempt to anticipate the backlash and counterarguments to the increasing ubiquity of e-books that they forecast. It will not surprise you that they were overly optimistic, but interestingly, these are some of the few specific predictions that seem to have gotten better as they reached further into the future. Normally, the opposite is true. Let's go through them.
"2001: Electronic textbooks appear and help reduce backpack load on students."
"2002: PCs and eBook devices offer screens almost as sharp as paper: 200 dpi physical resolution is enhanced even further with ClearType."
People still argue about whether iPad resolution is as good as print. And there weren't really usable e-book devices in the market until 2004 when the Sony Librie was released.
"2003: eBook devices weigh less than a pound, run eight hours, and cost as little as $99."
Kindles didn't start selling for $99 or less until 2011. Though they were right that e-books would be very light. Even the aforementioned Sony Librie weighed less than half a pound.
"2004: Tablet PCs arrive with eBook reading, handwriting input, and powerful computer applications."
Though tablet devices have existed for a looong time, it's generally acknowledged that tablets "arrived" when Apple released the iPad in 2010. (Gartner called the 2009 tablet market sales "next to zero.")
"2005: The sales of eBook titles, eMagazines, and eNewspapers top $1 billion."
Most estimates for the e-book market in 2005 were between $10 and $100 million. The ebook market in the United States (the most developed market) only started to take off with the release of the Kindle and iPad (and increasingly big smartphones with good screens).
"2006: eBook stands proliferate offering book and periodical titles at traditional bookstores, newsstands, airports -- even in mid-air."
No. This was the old retail model applied too directly the sales of books. In reality, we buy digital books online.
"2009: eBook titles begin to outsell paper in many categories. Title prices are lower, but sales are higher."
This is the closest prediction yet: it was in 2011 and 2012 that ebook sales started to rival and overtake print sales in certain categories (in the United States). Still, not bad for a ten-year-out projection.
"2010: eBook devices weigh half a pound, run 24 hours, and hold as many as a million titles."
Pretty much nailed it. In 2010, Amazon unveiled the Kindle 3G, the first Kindle to weigh under a pound. It had about 30 hours of active battery life. Interestingly, the one miss is in the storage capacity of the device. Real Kindles hold a thousand or so books, but, of course, you can delete and then download more as you go, so you have *access* to a million or so titles.
"2012: Electronic and paper books compete vigorously. Pulp industry ads promote 'Real Books from Real Trees for Real People.' "
This one is my favorite prediction. The implication here is that the cultural pushback on e-books would focus on the authenticity of paper books and the people who read them. And if you look around, physical books, in fact, have come to signal authenticity ("real people").
Take Nicholas Carr's argument in the Wall Street Journal. "Readers of weightier fare, including literary fiction and narrative nonfiction, have been less inclined to go digital," he wrote on December 31, 2011. "They seem to prefer the heft and durability, the tactile pleasures, of what we still call 'real books'--the kind you can set on a shelf." The physical weight of the book instantiates the heaviness of its ideas. And setting such tomes on the shelf is an indication that the purchaser is a reader of important things.
He concludes his essay with a vague idea that has a lot of currency among certain intellectuals: "There's something about a crisply printed, tightly bound book." Note the construction: "there's something about." I personally suspect that if anyone were to spell out precisely what that something was, it would sound kind of silly, like talking about why you like your favorite t-shirt: It would reveal too much about the qualities we want our objects to impart to us. (I say this as someone with full bookshelves and hundreds of Kindle books. I want my books to say the right things about me, too.)
"2015: Former high-tech rivals unite to fund the conversion of the entire Library of Congress to eBooks."
Get on it, dudes! You've got two years left, and I think it's a great and worthy notion.
I'll stop there as we can't really say much about the predictions for 2018 and beyond, but they're definitely worth looking at. You can find the original ad here, thanks to Flickr user catablogger.
Yesterday, the National Weather Service's Rick Smith posted a briefing to YouTube at 11:30am, which laid out a scenario for the day's weather events that was eerily precise. Specifically, he mentioned schools as an area of concern and highlighted the potential for an EF-4 tornado in the area south of I-40 and east of I-44 between 3 and 6pm. Shortly after 3pm, an EF-4 (or stronger) tornado hit Moore, which is located just south of I-40 and east of I-44.
It's crushing to realize that this disaster's rough outlines were predicted four hours ahead of time and yet know that this did not stop lives from being lost.
Here are three excerpts from the sadly prescient forecast:
We'll be talking about our increasing concerns for significant severe weather this afternoon and into this evening. We are expecting more significant severe weather today. The highest impacts we expect will be along and south of Interstate 44. Tornadoes and giant damaging hail are likely today. Something that's a little different today than yesterday is we are on a Monday and we do have schools in session and people driving home from work and that is a big, big concern for us as we expect severe weather potential to peak in that 3-6pm timeframe today...
Supercell storms are expected to develop in this area very quickly between 1 and 2 o'clock this afternoon. They will become severe fast, just like yesterday. We had storms that went from virtually nothing to producing large hail and tornadoes in less than an hour in some cases. So it's gonna go fast today.
If you're south of I-40 and east of I-44, you need to have a heightened state of awareness and be super alert to severe weather. We're expecting conditions today to be just as volatile if not even more so than they were yesterday for tornadoes. We've already had one EF4 tornado confirmed that occurred yesterday near Shawnee. I would not be at all surprised to have similar tornadoes occurring south of I-40, east of I-44. We're not trying to freak you out and scare you, we want you to be prepared. We're not guaranteeing a pinpoint forecast this is definitely going to happen, but you need to plan as if it is and be ready for what you're going to do.
A massive and powerful tornado hit Moore, Oklahoma this afternoon, causing widespread destruction, including at least 51 deaths. It's the deadliest tornado since 2011, and one of the worst in the last 20 years. This evening, President Obama signed a disaster declaration for Oklahoma.
Moore has a deep and tragic tornado legacy. The town could probably lay claim to being the very center of Tornado Alley, an area roughly defined from north Texas to South Dakota, and west of the Mississippi river. On May 3, 1999, Moore was hit by one of the worst tornadoes on record. That storm's winds were indirectly measured at 302 miles per hour, according to the National Weather Service, which called them "the highest winds ever found near earth's surface by any means." That tornado killed 36 people, the most deadly tornado in over 20 years, although several storms have surpassed that number of deaths since then, including 2011's Joplin storm, which killed 158.
The Red Cross is accepting donations via text message. Text REDCROSS to 90999 and you'll be billed $10. The municipal government of Moore is posting updates to a Facebook page.
We've assembled some common questions and research-based answers about the scientific and historical context of the disaster. If you have others you'd like answers to, email me at alexis.madrigal[at]gmail.com.
The simple answer is that warm, moist air from the Gulf of Mexico gets sandwiched between war, dry continental air and cold, dry air from drifting down from the Rockies. The combination creates the perfect conditions for thunderstorms to form. A more detailed explanation of the regional air movements is available here.
As these air masses collide, they can generate a type of particularly dangerous thunderstorm called a "supercell." They are characterized by their very strong, rotating updrafts accompanied by strong downdrafts. Tornadoes tend to occur at the interface between these two air movements.
While meteorologists are not quite sure why some supercells spawn tornadoes while others do not, it's clear that having strong thunderstorms makes having supercell more likely and having more supercells makes tornadoes more likely. Hence, the areas with the most strong thunderstorms tend to have the most tornadoes.
75 percent of all tornadoes on Earth occur in North America. Per square mile, Oklahoma has the fifth-most tornadoes of any state, and the fourth-most tornadoes on an absolute basis after Texas, Kansas, and Florida.
National Weather Service
National Weather Service
It is also worth noting, however, that even in the very center of Tornado Alley, it is rare for any particular area to be hit by a tornado. As you can see in the graphic below, any location in the reddest part of the map could expect that a violent storm like the one today would touch down within 25 miles only four times per century.
The science of tornado formation is complex. In the video below, NASA Goddard's Tim Samaras explains what we do and don't know about the mechanics. While scientists understand the conditions that make tornadoes more likely, the final steps that lead to tornadogenesis remain mysterious.
The "Enhanced Fujita" scale attempts to quantify tornado strength.
There is some controversy about the scale, particularly because wind speed is often inferred from damage, which is dependent on local construction practices, and not as precise as researchers would like.
Other factors affect the destructiveness of a tornado aside from its winds: how wide it is, where it touches down, and how long it remains on the ground.
The main thing to keep in mind is that anything EF4 or EF5 is likely to be very destructive if it hits near a human population center and stays near the ground for any period of time. Larger tornadoes do more damage for obvious reasons, too.
The preliminary reports are that the storm today was an EF4 and that it was large. Many think it will be upgraded to an EF5 after further damage assessment.
Only a couple percent of tornadoes are EF4 or larger.
The short answer is no, according to the National Weather Service, especially when it comes to more violent storms. "There has been little trend in the frequency of the stronger tornadoes over the past 55 years," the service notes. The longer answer is that more tornadoes are now reported than in years past, but that's probably due to an increase in the number of eyes scanning for tornadoes than the number of tornadoes. "Today, nearly all of the United States is reasonably well populated, or at least covered by NOAA's Doppler weather radars. Even if a tornado is not actually observed, modern damage assessments by NWS personnel can discern if a tornado caused the damage, and if so, how strong the tornado may have been," the Service explains. "This disparity between tornado records of the past and current records contributes a great deal of uncertainty regarding questions about the long-term behavior or patterns of tornado occurrence."
The short answer, again, is no (the 2011 Joplin event notwithstanding). Although there are more people, and therefore a greater number of injured parties, in the United States in the regions where tornadoes strike, better forecasting and warning systems have greatly reduced the rate of fatalities per million people and the overall number of fatalities, as you can see in the following two charts. The fall was particularly dramatic before the tragedy in Missouri in 2011. (Note that the second chart shows the period from 1875-2000.)
No. Once loss data from the past is normalized for
increasing wealth, population and building stock and also adjusted for inflation, the damage tornadoes cause has not increased and actually shows a hint of having decreased since 1950, according to a
study by Kevin Simmons, Daniel Sutter and Roger Pielke, Jr. published late last year. The process of normalization is fairly
statistically complex and the historical data varies in
quality, but the methods used to normalize loss data are
well-established.
Normalized tornado damage (Roger Pielke Jr).
The normalization methodology shows the consequences of more humans and human infrastructure concentrated in areas at high-risk for tornado strikes, so when a bad storm strikes, worse things can happen. As the American Meteorological Society's Bill Hooke wrote in 2011, "Tornadoes
hitting downtown areas in the past? Rare - almost unheard of. But
tornadoes hitting downtown areas in the future? Increasingly common."
He gave the game Battleship as an analogy: if the board is filled with
ships, then it is easier to hit them. The ships are human
infrastructure and the opponents are natural disasters. (Still, Hooke
would note that any particular home in tornado alley has a very, very,
very small chance of being hit by a tornado in any given year.)
The short answer is: early warning systems worked as expected.
The National Weather Service's Norman, Oklahoma office says that a tornado warning went into effect 16 minutes before the storm hit. That's three minutes faster than the current average lead time for a warning, according to NOAA.
In Moore, our Emergency Management staff works closely with the meteorologists at the National Weather Service Forecast Office located in Norman. NWS personnel will generate warnings based upon not only radar information, but also information from our Moore severe weather spotters that are in the field. Therefore, when a warning is issued for Cleveland County, our EOC staff normally have played a part in the decision process. If our EOC staff determine that tornadic conditions will directly affect Moore, we will activate our local warning system, consisting of an outdoor warning siren system and a cable television interrupt. The NWS office will be triggering the alarm on NOAA Weather Radio, and local media will be broadcasting the warning as well.
The general problem with the warnings is that people might not hear or see them immediately, reducing the amount of time they have to find shelter. The second problem is that the tornado hit during early rush hour, a little after 3pm, when many people were on the move.
Because there was a large tornado in 1999, we have detailed information on some of the problems with the construction methods in the area.
Engineer Timothy Marshall, in a report on the 1999 tornado, noted that much of the construction in the area does not follow best practices for resisting a tornado's winds. Marshall's damage report found that homes that affixed their walls to their foundations in certain ways had serious problems during the storm. Yet, when he returned to see the rebuilding process, he found many homes rebuilding with the same methods that had failed before. That's the housing stock that was hit by the storm today. Here's how Marshall put it in his 2002 report:
The author revisited the disaster area three months after the tornado to check the quality of new house construction. A total of 40 houses were examined in Moore and southern Oklahoma City on sites at which houses previously had been destroyed. The author found that the quality of new home construction generally was no better than homes built prior to the tornado. Most newly built homes were attached to their concrete foundations with tapered cut nails or shot pins as had been noted in homes destroyed by the tornado.
Marshall is back on the ground now investigating the current tornado damage, so we expect to find out more about what happened soon.
Statistically, the worst months for tornado formation are May and June, though they can happen all year round. Historically, they've hit most often between 5 and 6 pm local time.
A better way of asking this question might be this: on what time scale, can we predict a tornado might form? The answer is that forecasters have a pretty good sense that a tornado is forming about 15 minutes before it touches down. On the day of a storm, forecasters have a pretty good idea of where severe storms might occur. Below, I've embedded the forecast update that the National Weather Service in Norman, Oklahoma released at about 11:30am. Generally speaking, they called the region and the time in which the tornado touched down: east of I-44 and south of I-40. Moore is less than 15 miles from where those two highways meet. But the region of worry that you see highlighted in the video is quite large: meteorologists can't provide anything like a pinpoint forecast.
On the much longer timescale, historical data, as in the map below (sent over by climate research Roger Pielke Jr), can tell you that a particular region is very prone to a tornado at a particular time of year, but not much more.
NASA's Landsat satellites have been snapping pictures of the Earth from orbit since 1972. The most recent iteration of the project, the Landsat Data Continuity Mission, arrived at its orbital resting place on April 12, and shot this series of 56 images shortly thereafter. NASA stitched the pictures together into one long strip, which you can tour in the video above.
As always, satellite images testify to the wonder of the biosphere. This particular set of pictures, though, is a simple meditation on the diversity of conditions on Earth, and the mark that humanity has left on the planet.
Yahoo announced they will acquire Tumblr for $1.1 billion this afternoon. The news comes about a year after Facebook snatched up the hot startup Instagram. In a post-Facebook world, that leaves two large independent social networks: Twitter and Pinterest, the oldest and youngest in the group, respectively.
I wanted to get a sense of the relative growth of these companies through time, so I put together this chart. DISCLAIMER: it's really hard to get exact numbers on these companies and even harder to get exact times for exact numbers. I used company announcements, stats geeks inferences, and some good old Business Insider aggregations. That is to say, the quality of the numbers varies here, too. So, take this all with a grain of salt, and know that while the curves you see are generally correct, this only a rough approximation.
Looking at the chart, you can see the remarkable success that all of these companies have had getting to 50 million users, even though their usage models are all very different. Twitter's the largest, Tumblr's second, and Instagram is third. But Instagram's growth stands out: building on the social graphs generated by earlier networks (and with a great product), they were in the big leagues within months, not years. Pinterest's graph looks a little different, but it's worth noting, the Pinterest and Tumblr numbers are the shakiest, and Pinterest is still early in exploring its own potential.
And just for some perspective, Facebook is more than five times larger than all these services and about twice as big as all of them combined.
