The world’s servers processed 9570000000000000000000 bytes in 2008

Posted by

0

We recently completed another major “How Much Information?” report. This one measures how much information servers processed, worldwide.  Answer: we estimated 9.57 ZB in 2008. The coverage was broad, although not as much general press as our previous report on information consumption. I was impressed to see us reach 130,000 Google hits in a few days (try searching on the string 9570000000000000000000 ), until I realized that most of them are spam sites – they just duplicate hot items in hopes that Google will point to them, instead of the original.

As always, we had to make a lot of judgment calls in deciding what to measure and how. See the report, and the forthcoming technical appendix.

The report is available here.

Researchers Study the World’s Information Explosion – The Numbers Guy – WSJ

Posted by

0

Arguments over what information to measure, and how are fundamentally irresolvable. There are too many good answers, and for most purposes measurement problems make it impossible to measure exactly what we want, in any case. The WSJ’s “numbers guy” collects some views on this, related to the new USC/Science article on the world’s computation.

My print column this week examines a recent paper in Science and other research efforts that attempt to quantify the world’s information. The studies generally compile dozens of pieces of data, such as hard-drive production and sales, and put all information into a single unit, such as bytes or words. “We can say things like ‘a 6 square-cm newspaper image is worth a 1000 words,’ ” the Science study’s authors, Martin Hilbert and Priscila López, wrote.

via Researchers Study the World’s Information Explosion – The Numbers Guy – WSJ.

Qualcomm and Intel: Why are industrial companies so bad at consumer products?

Posted by

0

Qualcomm is trying to bring out consumer products, and failing — Flo TV, and more on the way. There’s nothing wrong with initiatives that fail, especially when you are as rich as QC, but they are making systematic errors. Intel got burned early in its history trying to do digital watches, and learned from the experience – it has never had a direct-to-consumer product since then, though it has done plenty of brand-marketing campaigns. Qualcomm seems intent on repeating its errors, without learning from them.

I asked a deep-thinking friend, Jim Cook, about this, and here’s his response. He is even more negative than I was!

Industrial companies are built on rationality, non-commodity consumer companies are built on empathy.  These epistemological predilections invade and bias the functioning of  the corporation – R&D needs to shift from cost/performance to fad and fancy, Finance needs to shift from predictable to volatile, Manufacturing has to shift from plans to whim response.  Unless you either separate them completely (resource allocation becomes the major problem, not to mention confusing capital markets), you’re destined to have internal destructive struggles.  One can point to Apple as a counter example, until one realizes that Apple, from its origins, has been a cult satisfier (no price performance in almost any of the Apple products) and without a cult leader (Jobs) will not compete successfully with real consumer companies.  Motorola tried to do both industrial and consumer, eventually split up and both will die in the next decade.

JAMA — Ensuring Integrity in Industry-Sponsored Research: Primum Non Nocere, Revisited, March 24/31, 2010, DeAngelis and Fontanarosa 303 12: 1196

Posted by

0

The evidence on false conclusions from drug trials, and their publications, is mounting. I am forced to the conclusion that, most likely,

1) This has been going on for a long time; the big change is that it’s occasionally getting noticed now, and

2) The results of this over many years must be that  doctors are prescribing based on incorrect evidence. Specifically, lots of drugs are getting prescribed when they shouldn’t be.

According to the article by Nissen,1 the report of the Senate investigation,2 and published media accounts,8-9 the manufacturer of rosiglitazone exerted inappropriate influence during the conduct of a pivotal safety study of this drug, the RECORD Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycaemia in Diabetes clinical trial,10-11 which included undertaking nonprespecified unblinding of study data; attempting to undermine the authority and responsibilities of the study steering committee; expediting publication of an unscheduled interim analysis,10 specifically to counter2 the publication of a meta-analysis by Nissen and Wolski12 that suggested increased cardiovascular risk associated with rosiglitazone; having employees extensively involved in statistical analysis,11 and preparation of the manuscript10 reporting the results of the trial; and reportedly2 failing to fully acknowledge the significant cardiovascular risk associated with this drug.

via JAMA — Ensuring Integrity in Industry-Sponsored Research: Primum Non Nocere, Revisited, March 24/31, 2010, DeAngelis and Fontanarosa 303 12: 1196.

Airplane fires are still “craft” stage

Posted by

0

I just presented some of my research on how flying went from Art to Science, at the INFORMS conference. (Very short paper  at http://escholarship.org/uc/item/2sj362h9. Long working paper is coming soon.)  My colleague Erica Fuchs related a recent experience that demonstrates how certain activities in flying are still at what I call the “Rules + Instruments” stage, which got started in the 1930s.

During a flight to Chicago, she noticed a “bad smell.” One of the other passengers was an off-duty airline pilot who also noticed it, and went rocketing up to the cockpit to tell the pilots.  The result: they landed immediately, at an unused airport near their flight path.  A mechanic flew in and found that one of the light ballasts had overheated, producing the smell. He replaced it, and they proceeded.

Ten years ago, this would  have been treated much less urgently. “Bad electrical smell” is not very specific. But a few recent tragedies have focused attention on the problem of fuselage fires. The sobering insights were that 1) we still don’t have good instruments for detecting or locating fires in some parts of an aircraft  – the human nose is still the best we have,  2) most burning smells are not serious problems, but when there really is a fire, sometimes the only effective countermeasure is to land; 3) every minute counts.

Pilots used to follow an elaborate checklist, attempting to diagnose and solve the problem in flight, but in a couple of cases that gave the fire time enough to take hold and destroy the aircraft. So on her flight, the procedure apparently was “when you can’t identify the source, land as fast as possible.” Does anyone know what the “official” checklists look like now, for this situation?

Official report of F/A-18 crash

Posted by

0

On December 8, 2008 a Marine Corps jet crashed near my house, killing four civilians. The USMC did a thorough investigation and held a press conference about the results, complete with powerpoints. However, I was not able to find the actual report anywhere. I eventually filed a FOIA request, and got a redacted version of the report and all its attachments.

The crash fit a common pattern: a lot of small mistakes, by a variety of people, added up to a disaster. It was not just the pilot at fault, although he was certainly involved.

Because these reports were hard to get, and will be of interest to others, I’m putting them up on this site. If I can find a better (more relevant) repository, I’ll also post it there.

There is a lot of bureaucratic material in the reports. I will gradually post all of it. Questions are welcome – use the Comments section.

Continue reading →

Language Log » Tracking a factoid to its lair

Posted by

0

Our How Much Information? 2008 Consumer report continues to generate discussion. The comments that follow this blog entry in Language Log are quite interesting. I added my own comment – we’ll see how long it takes to go through their moderation process.

Matt Richtel, one of the leading current peddlers of the “technology is eating our brains” meme, is fond of this assertion:The average person today consumes almost three times as much information as what the typical person consumed in 1960, according to research at the University of California, San Diego.That version is the lead paragraph of the online site for his appearance on Fresh Air, “Digital Overload: Your Brain On Gadgets”, 8/24/2010. I was curious about what this sentence could mean, and more specifically, I wondered which UCSD researchers did the measurements, and what they they measured.

via Language Log » Tracking a factoid to its lair.