Michael Nielsen

Click here for all of my del.icio.us bookmarks.

• The story of the WorldWide Telescope « Jon Udell
  • Quoting Jim Gray in 2002: “Most scientific data will never be directly examined by scientists; rather it will be put into online databases where it will be analyzed and summarized by computer programs. Scientists increasingly see their instruments through online scientific archives and analysis tools, rather than examining the raw data. Today this analysis is primarily driven by scientists asking queries, but scientific archives are becoming active databases that self-organize and recognize interesting and anomalous facts as data arrives. “
• Nascent: Social Not Working?
  • A stimulating talk by Timo Hannay about Science 2.0.
• Peter Norvig: Presidential Election 2008 FAQ
  • A great deal of useful information about both campaigns.
• xkcd - Height
  • xkcd does “Powers of Ten”. Very cool.
• Jay Walker’s Library
  • Lust.
• Media Bias: Going beyond Fair and Balanced: Scientific American
  • A clever way to test for bias: “Groeling collected two different data sets: in-house presidential approval polling by ABC, CBS, NBC and FOX News and the networks’ broadcasts of such polls on evening news shows from January 1997 to February 2008. Groeling found that, with varying degrees of statistical significance, CBS, NBC and ABC showed what Groeling calls a pro-Democrat bias. For instance, CBS was 35 percent less likely to report a five-point drop in approval for Bill Clinton than a similar rise in approval and was 33 percent more likely to report a five-point drop than a rise for George W. Bush. Meanwhile FOX News showed a statistically significant pro-Republican bias in the most controlled of the three models Groeling tested: its Special Report program was 67 percent less likely to report a rise in approval for Clinton than a decrease and 36 percent more likely to report the increase rather than the decrease for Bush.”
• Adding Noughts in Vain: Shock: Global Warming Still Happening!
  • Useful discussion of the last 30 years of data on global temperatures.

Click here for all of my del.icio.us bookmarks.

• …My heart’s in Accra » Cass Sunstein’s “Infotopia”
  • An excellent summary of an excellent book.
• Anne Truitt Zelenka » Wikified Lectures
  • “I hit upon the idea of outsourcing my lecture to the students — crowdsourcing it, wikifying it, though they didn’t get to choose which part of the lesson they were to teach which would really be required for true wikification. I assigned each student one test and one problem using that test. They had 10 minutes to study their test and figure out their problem. Then they each presented the test with its demo problem.”

Click here for all of my del.icio.us bookmarks.

• Kevin Kelly — Where Attention Flows, Money Follows
  • “Almost anything else except attention can be manufactured as a commodity. Luxury goods are only luxuries temporarily. They quickly are counterfeited and commodified. Premium brands are only premium because they garner a surplus of attention. Maintain an incoming flow of attention and money will follow. That is really all you need to know.”
• Ten Thousand Cents
  • Crowdsourced art project - a US 100 dollar note, image created by 10,000 people at Amazon’s Mechanical Turk.
• Review of the Australian National Innovation System
  • A report prepared for the Australian Federal Minister of Innovation, Kim Carr. It contains extensive recommendations related to the need for a more open scientific system, especially from page 93 on. Via Joshua Gans.
• Open access in Australia
  • Favourable remarks on open access from Innovation Minister Kim Carr.
• Aaron Sorkin Conjures a Meeting of Obama and Bartlet - Op-Ed - NYTimes.com

Click here for all of my del.icio.us bookmarks.

Two years after the breakup of the Soviet Union, British economist Paul Seabright was talking with a senior Russian official who was visiting the UK to learn about the free market. “Please understand that we are keen to move towards a market system,” the official said, “But we need to understand the fundamental details of how such a system works. Tell me, for example: who is in charge of the supply of bread to the population of London?” [1]

The familiar but still astonishing answer to this question is that in a market economy, everyone is in charge. As the market price of bread goes up and down, it informs our collective behaviour: whether to plant a new wheat field, or leave it fallow; whether to open that new bakery you’ve been thinking about opening on the corner; or simply whether to buy two or three loaves of bread this week. The price thus aggregates an enormous amount of what would otherwise be hidden knowledge from all the people interested in the production or consumption of bread, that is, nearly everyone. By using prices to aggregate this knowledge and inform further actions, the market produces outcomes superior to even the brightest and best informed individuals.

Unfortunately, markets don’t always aggregate knowledge accurately. When participants in a market are mistaken in systematic ways, markets don’t so much aggregate knowledge as they aggregate misunderstanding. The result can be an enormous collective error in judgement; when the misjudgement is revealed, the market crashes.

My subject in this essay is not economics, it’s science. So what’s all this got to do with science?

The connection involves the question of what it means to understand something. In economics, many basic facts, such as prices, have an origin which isn’t completely understood by any single person, no matter how bright or well informed, because none of those people have access to all the hidden knowledge that determines those prices.

By contrast, until quite recently the complete justification for even the most complex scientific facts could be understood by a single person.

Consider, for example, astronomer Edwin Hubble’s discovery in the 1920s of the expansion of the Universe. By the standards of the time, this was big science, requiring a complex web of sophisticated scientific ideas and equipment - an advanced telescope, spectroscopic equipment, and even Einstein’s special theory of relativity. To understand all those things in detail requires years of hard work, but a dedicated person like Hubble could master it all, and so in some sense he completely understood his own discovery of the expansion of the Universe.

Science is no longer so simple; many important scientific facts now have justifications that are beyond the comprehension of a single person.

For example, in 1983 mathematicians announced the solution of an important longstanding mathematical problem, the classification of the finite simple groups. The work on this mathematical proof extended between 1955 and 1983, and required approximately 500 journal articles by 100 mathematicians. Many minor gaps were subsequentely found in the proof, and at least one serious gap, now thought (by some) to be resolved; the resolution involved a two-volume, 1300-page supplement to the proof. Although mathematicians are working to simplify the proof, even the simplified proof is expected to be exceedingly complex, beyond the grasp of any single person.

The understanding of results from the Large Hadron Collider (LHC) will be similarly challenging, requiring a deep knowledge of elementary particle physics, many clever ideas in the engineering of the accelerator and the particle detectors, and complex algorithms and statistical techniques. No single person understands all of this, except in broad detail. If the discovery of the Higgs particle is announced next year, there won’t be any single person in the world who can say “I understand how we discovered this” in the same way Hubble understood how he discovered the expansion of the Universe. Instead, there will be a large group of people who collectively claim to understand all the separate pieces that go into the discovery, and how those pieces fit together.

Two clarifications are in order. First, when I say that these are examples of scientific facts beyond individual understanding, I’m not saying a single person can’t understand the meaning of the facts. Understanding what the Higgs particle is requires several years hard work, but there are many people in the world who’ve done this work and who have a solid grasp of what the Higgs is. I’m talking about a deeper type of understanding, the understanding that comes from understanding the justification of the facts.

Second, I don’t mean that to understand something you need to have mastered all the rote details. If we require that kind of mastery, then there’s no one person who understands the human genome, for certainly no-one has memorized the entire DNA sequence. But there are people who understand deeply all the techniques used to determine the human genome; all that is missing from their understanding is the rote work identifying all the DNA base pairs. The examples of the LHC and the classification of the finite simple groups go beyond this, for in both cases there are many distinct deep ideas involved, too many to be mastered by any single person.

Science as complex as the LHC and the classification of finite simple groups is a recent arrival on the historical scene. But there are two forces that will soon make science beyond individual understanding far more common.

The first of these forces is rapid internet-fueled growth in the number of large-scale scientific collaborations. In the short term, these collaborations will mostly just crowdsource rote work, as is being done, for example, by the galaxy classification project Galaxy Zoo, and so the results will pose no challenge to individual understanding. But as the collaborations get more sophisticated we can expect to see many more online collaborations that delegate large amounts of specialized work, building up to a whole whose details aren’t fully understood by any single person.

The second of these forces is the use of computers to do scientific work. A nascent example is the proof of the four-colour theorem in mathematics. A small group of mathematicians outlined a proof, but to complete the proof, they had to check a large number of cases of the theorem, more than they could check by hand. Instead, a computer was used to check those cases. This isn’t an instance of science beyond individual understanding, though, because mathematicians familiar with the proof feel the computer was simply doing rote work. But the people doing computational science are getting cleverer in how they use computers to make discoveries. Machine learning, data mining and artificial intellgience techniques are being used in increasingly sophisticated ways to produce real insights, not just rote work. As the techniques get better, the number of insights found will increase, and we can expect to see examples of science beyond individual understanding generated this way: “I don’t understand how this discovery was made, but my computer and I do together”.

More powerful than either of these forces will be their combination: large-scale computer-assisted collaboration. The discoveries from such collaboration may well not be understood by any single individual, or even by a group. Instead, it will reside inside a combination of the group and their networked computers.

Such scientific discoveries raise challenging issues. How do we know whether they’re right or wrong? The traditional process of peer review and the criterion of reproducibility work well when experiments are cheap, and one scientist can explain to another what was done. But they don’t work so well as experiments get more expensive, when no one person fully understands how an experiment was done, and when experiments and their analyses involve reams of data or ideas.

Might we one day find ourselves in a situation like in a free market where systematic misunderstandings can infect our collective conclusions? How can we be sure the results of large-scale collaborations or computing projects are reliable? Are there results from this kind of science that are already widely believed, maybe even influencing public policy, but are, in fact, wrong?

These questions bother me a lot. I believe wholeheartedly that new tools for online collaboration are going to change and improve how science is done. But such collaborations will be no good if we can’t assess the reliability of the results. And it would disastrous if erroneous results were to have a major impact on public policy. We’re in for a turbulent and interesting period as scientists think through what’s needed to arrive at reliable scientific conclusions in the age of big collaborations.

Acknowledgements

Thanks to Jen Dodd for providing feedback that greatly improved an early draft of this essay. The essay was stimulated in part by the discussion during Kevin Kelly’s session at Science Foo Camp 2008. Thanks to all the participants in that discussion.

Further reading

This essay is adapted from a book I’m currently working on about “The Future of Science”. The basic thesis is described here, and there’s an extract here. If you’d like to be notified when the book is available, please send a blank email to the.future.of.science@gmail.com with the subject “subscribe book”. You’ll be emailed to let you know when the book is to be published; your email address will not be used for any other purpose.

Subscribe to my blog here.

You may enjoy some of my other essays.

Footnote

[1] “Who is in charge of the supply of bread to the population of London?” - see Paul Seabright’s The Company of Strangers.

• Enhancing Multitasking to Enhance Our Minds. « Essays by Danielle Fong
  • Great essay about information architecture.
• Evolutionary Acceleration Research Institute Ready to Start “Squirrel Smasher”
  • “Scientists from the Evolutionary Acceleration Research Institute (EARI) announced that the first test of the Giant Animal Smasher (GAS) will begin on December 19, 2008, the 41st anniversary of the premiere of Dr. Dolittle.”
• User-generated science | The Economist
• Kevin Kelly: The Next Fifty Years of Science
  • “Landmarks in the history of the scientific method are the invention of libraries, indexes, citations, controlled experiments, peer review, placebos, double blind experiments, randomization, and search among others. At the core of the scientific method is the structuring of information. In the next 50 years, as the technologies of information and knowledge accelerate, the nature of the scientific process will change even more than it has in the last 400 years. We can’t predict what specific inventions will arise in the next 50 years, but based on long-term trends in epistemic tools, I believe we can speculate on how the scientific method itself — that is, how we know — will change in the next five decades.”
• That first step can be a doozy « Jon Udell
  • Jon Udell on machine readability and crowdsourcing of data analysis.
• Obama Campaign Reveals Science Advisors | Wired Science from Wired.com
  • Includes Harold Varmus, cofounder of PLoS.

Click here for all of my del.icio.us bookmarks.

• Sergey Brin’s blog
• Stanford offers 10 free online Computer Science courses
• Hal Varian: Copying and copyright (pdf)
• The Long Tail: Another Harvard professor helpfully suggests that we make hits
  • Heh: “What is it about Harvard Business School professors and their embrace of the grindingly conventional pitched as fresh contrarianism? The latest is HBS marketing professor John Quelch, who bravely argues for, well, successful products”
• The Long Tail: A passionate amateur almost always beats a bored professional
  • “Amateurs self-select for the job. Professionals are selected. For most jobs, volunteers beat draftees.”
• Schneier on Security: The NSA Teams Up with the Chinese Government to Limit Internet Anonymity
  • Charming: “A United Nations agency is quietly drafting technical standards, proposed by the Chinese government, to define methods of tracing the original source of Internet communications and potentially curbing the ability of users to remain anonymous… The U.S. National Security Agency is also participating in the “IP Traceback” drafting group, named Q6/17, which is meeting next week in Geneva to work on the traceback proposal. Members of Q6/17 have declined to release key documents, and meetings are closed to the public. “
• Jason Fried - 10 Things We’ve Learned at 37Signals | Kris Jordan
  • Excellent list from a company that’s done a lot of bold experimentation.
• A Proposed Standard for the Scholarly Citation of Quantitative Data
• Ellen Roche
  • Heartbreaking: “Ellen Roche was a healthy 24 year old lab technician at the Johns Hopkins (JH) Asthma Center. She volunteered to take part in an experiment to understand the natural defenses of healthy people against asthma. Roche was part of a group that inhaled hexamethonium, a drug which induced a mild asthma attack. Physicians stood by in case of complications and to measure how the subjects responded to the asthma attack. Within 24 hours of inhaling the drug, Roche had lost one-third of her lung capacity. Within a month she was dead… Dr. Alkis Togias, the director of the
    experiment, apparently limited his hexamethonium research to one contemporary textbook and PubMed… PubMed is a premier example of FOS, a contender for FOS at its best. So does the Ellen Roche case prove that FOS is inadequate, even hazardous? How just is this interpretation? What are the lessons of this case for FOS?”
• Abū Rayhān Bīrūnī - Wikipedia, the free encyclopedia
  • Extraordinary Persian polymath of the 11th century.
• Caveat Lector » What do we want from IRs, and what are we doing to repository rats?
  • Dorothea Salo on the future of Institutional Repositories.
• Confessions of a Science Librarian: Science in the 21st Century reading list
  • Really great list of books to read from John Dupuis.
• A Blog Around The Clock : ScienceOnline’09 - Registration is Open!
  • Building on the very successful Science Blogging 2007 and 2008 events.
• John Graham-Cumming: Dear Nature
  • “…you want to sell it [the paper] to me for $32. How do you justify selling a PDF of a 76 year old paper that contains just over 700 words for $32?”

Click here for all of my del.icio.us bookmarks.

• Ellen Roche
  • Heartbreaking: “Ellen Roche was a healthy 24 year old lab technician at the Johns Hopkins (JH) Asthma Center. She volunteered to take part in an experiment to understand the natural defenses of healthy people against asthma. Roche was part of a group that inhaled hexamethonium, a drug which induced a mild asthma attack. Physicians stood by in case of complications and to measure how the subjects responded to the asthma attack. Within 24 hours of inhaling the drug, Roche had lost one-third of her lung capacity. Within a month she was dead… Dr. Alkis Togias, the director of the
    experiment, apparently limited his hexamethonium research to one contemporary textbook and PubMed… PubMed is a premier example of FOS, a contender for FOS at its best. So does the Ellen Roche case prove that FOS is inadequate, even hazardous? How just is this interpretation? What are the lessons of this case for FOS?”
• Abū Rayhān Bīrūnī - Wikipedia, the free encyclopedia
  • Extraordinary Persian polymath of the 11th century.
• Caveat Lector » What do we want from IRs, and what are we doing to repository rats?
  • Dorothea Salo on the future of Institutional Repositories.
• Confessions of a Science Librarian: Science in the 21st Century reading list
  • Really great list of books to read from John Dupuis.
• A Blog Around The Clock : ScienceOnline’09 - Registration is Open!
  • Building on the very successful Science Blogging 2007 and 2008 events.
• John Graham-Cumming: Dear Nature
  • “…you want to sell it [the paper] to me for $32. How do you justify selling a PDF of a 76 year old paper that contains just over 700 words for $32?”
• Paint your roof white, save the planet - Machinist
  • The effects of making urban surfaces white are, apparently, significant. I’m not sure I buy this - the same argument should should that road surfaces have a major greenhouse effect, but I haven’t run the numbers.
• Uncertain Principles: A Longitudinal Study of Blogging Traffic
  • Chad compares blog traffic for science vs non-science posts, and how they do over time. Do the science posts have greater staying power or not? The data are ambiguous, but if there’s an effect, it’s not large.
• BarCamp Africa
• iamelgringo: Mechanical Turk: Now with 25 percent more Awesome.
  • Using crowdsourcing to do data analysis.
• gapingvoid: good ideas have lonely childhoods
• Spore’s Piracy Problem - Forbes.com
  • Well, I was all set to go buy Spore this morning. But judging from the reviews the DRM in the game is getting, I think not.
• Dancing death
  • “Sometime in mid-July 1518, in the city of Strasbourg, a woman stepped into the street and started to dance. She was still dancing several days later. Within a week about 100 people had been consumed by the same irresistible urge to dance.”
• Peter Suber: More on the arguments to overturn the NIH [open access] policy
• Internet Bots: Anatomy of a Stock Selling Frenzy
  • Much more about the massive automated United Airlines stock selloff triggered by Google News. Fascinating.
• How long would it take the LHC to defrost a pizza?: Scientific American Blog
  • Scientific American tackles the big questions.
• Google News and United Airlines’ share price
  • UAL lost 75% of its market cap over 15 minutes. It’s unclear what happened, but it looks like Google News may have played some role in driving the crash.
• Has the Large Hadron Collider destroyed the earth yet?
  • Helpful.
• Uncertain Principles: Micro-Blogging Conference Talks
  • Chad Orzel on the benefits of multiple people simultaneously micro-blogging conference talks.
• FriendFeed room for “Science in the 21st Century”
• Science in the 21st Century Talks
  • Video and slides for the talks.
• Prospects in Theoretical Physics (PiTP) - 2008 | Video Lectures
• When Academia Puts Profit Ahead of Wonder - NYTimes.com
  • About the Bayh-Dole act, one of the most important pieces of legislation in the 20th century.
• Mememoir: Wiki For Science
• Terry Tao’s blog book
• Twitter / cern
  • Guess who has a Twitter feed?
• PLoS ONE: Targeted Development of Registries of Biological Parts
  • An analysis of useage patterns in the MIT’s Registry of Standard Biological Parts, which is a prototype for open source science.
• CIA, FBI push ‘Facebook for spies’ - CNN
  • ‘”It’s a place where not only spies can meet but share data they’ve never been able to share before,” Wertheimer [assistant deputy director of national intelligence for analysis] said. “This is going to give them for the first time a chance to think out loud, think in public amongst their peers…’
• Cosma Shalizi: Collective Cognition
  • A wonderful collection of links on collective cognition.
• Mark Newman: The first-mover advantage in scientific publication
  • “Mathematical models of the scientific citation process predict a strong “first-mover” effect under which the first papers in a field will, essentially regardless of content, receive citations at a rate enormously higher than papers published later. Moreover papers are expected to retain this advantage in perpetuity - they should receive more citations indefinitely, no matter how many other papers are published after them. We test this conjecture against data from a selection of fields and in several cases find a first-mover effect of a magnitude similar to that predicted by the theory. Were we wearing our cynical hat today, we might say that the scientist who wants to become famous is better off — by a wide margin — writing a modest paper in next year’s hottest field than an outstanding paper in this year’s. On the other hand, there are some papers…that buck the trend and attract significantly more citations than theory predicts despite having relatively late publication dates…”
• Rob Carlson :: “Biology is Technology”
  • Draft chapters of Rob Carlson’s book on synthetic biology.
• Brad DeLong and Michael Froomkin: Speculative Microeconomics for Tomorrow’s Economy
  • “[the paper deconstructs] Adam Smith’s case for the market system. It points out three assumptions about production and distribution technologies that are necessary if the invisible hand is to work as Adam Smith claimed it did. We point out that these assumptions are being undermined more and more by the revolutions currently ongoing in data processing and data communications. “
• LiveScience: Era of Scientific Secrecy Near End
  • An article on open science for a general audience.
• Augmented Social Cognition: Long Tail of user participation in Wikipedia
  • The (very well-known) blog post which describes the distribution of user edits in Wikipedia. Based on an academic paper, but this observation seems to have been made _after_ the paper the author wrote was finalized, so it’s not actually in the paper.
• William James - The PhD Octopus
  • “America is thus a nation rapidly drifting towards a state of things in which no man of science or letters will be accounted respectable unless some kind of badge or diploma is stamped upon him, and in which bare personality will be a mark of outcast estate. It seems to me high time to rouse ourselves to consciousness, and to cast a critical eye upon this decidedly grotesque tendency. Other nations suffer terribly from the Mandarin disease. Are we doomed to suffer like the rest? “
• The world needs more foxes and fewer hedgehogs
  • Philip Tetlock, who has spent 20 years asking pundits to predict who will win elections, what countries will acquire nuclear weapons or enter the European Union and how the first Gulf war would end… his respondents are not very good. They do better than a chimp who answers at random, but not much, and worse than simple forecasting rules based on extrapolation. But some pundits are better than others. A little knowledge is helpful. Dilettantes – people with the information you will acquire from diligent reading of this newspaper – do much better than undergraduates who based their judgment on a one-page summary of the issues. But experts have little advantage over dilettantes. The reputation of the experts is a guide to which are worth following. But not in the way you might expect. Bad forecasters are consulted more frequently than good ones. The more famous the expert, the worse his prognostications. “
• Freebase Parallax
  • Very interesting application capable of extracting complex information from Freebase. The video demo is worth watching.
• Peering into PLoS One comment stats : Deepak Singh
  • Lots of statistics about PLoS One’s experiment with commenting.
• Vernor Vinge’s View of the Future - Is Technology That Outthinks Us a Partner or a Master ? - John Tierney
• Offloading Cognition onto Cognitive Technology: Itiel Dror and Stevan Harnad
  • “Cognitive technology allows cognizers to offload some of the functions they would otherwise have had to execute with their own brains and bodies alone; it also extends cognizers’ performance powers beyond those of brains and bodies alone. Language itself is a form of cognitive technology that allows cognizers to offload some of their brain functions onto the brains of other cognizers. Language also extends cognizers’ individual and joint performance powers, distributing the load through interactive and collaborative cognition. Reading, writing, print, telecommunications and computing further extend cognizers’ capacities. And now the web, with its distributed network of cognizers, digital databases and sofware agents, has become the Cognitive Commons in which cognizers and cognitive technology can interact globally with a speed, scope and degree of interactivity that yield performance powers inconceivable with unaided individual cognition alone. “
• American lawbreaking: Tim Wu - Slate Magazine
  • “The importance of understanding why and when we will tolerate lawbreaking cannot be overstated. Lawyers and journalists spend most of their time watching the president, Congress, and the courts as they make law. But tolerance of lawbreaking constitutes one of the nation’s other major—yet most poorly understood—ways of creating social and legal policy. Almost as much as the laws that we enact, the lawbreaking to which we shut our eyes reflects how tolerant U.S. society really is to individual or group difference. It forms a major part of our understanding of how the nation deals with what was once called “vice.” While messy, strange, hypocritical, and in a sense dishonest, widespread tolerance of lawbreaking forms a critical part of the U.S. legal system as it functions. “
• Dani Rodrik’s weblog: Why the econ-blogosphere is here to stay
  • “one of the unexpected scholarly benefits of having a blog is that it is like keeping an intellectual journal. You get an idea, you jot it down in your blog. Some months later, you vaguely remember having had the idea and you google your own blog to recover it. I am not kidding: I google my own blog all the time… “
• The Value of Openness in Scientific Problem Solving
  • “Openness and free information sharing…are supposed to be core norms of the scientific community… these norms are not universally followed. Lack of openness and transparency means… problem solving is constrained to a few scientists… who typically fail to leverage the entire accumulation of scientific knowledge… We present evidence of the efficacy of problem solving when disclosing problem information. The method’s application to 166 discrete scientific problems from the research laboratories of 26 firms is illustrated. Problems were disclosed to over 80,000 independent scientists… approach solved one-third of a sample of problems that large…R & D-intensive firms had been unsuccessful in solving internally… success was…associated with the ability to attract specialized solvers with…diverse scientific interests…. successful solvers solved problems at the boundary or outside of their fields of expertise, indicating a transfer of knowledge from one field to others. “
• The Quantum Pontiff : Self-Correcting Quantum Computers, Part I
  • The first of Dave Bacon’s excellent multi-part series about how quantum computers can correct themselves.
• Uncommon Knowledge and Open Innovation - john wilbanks’ blog - john wilbanks’ blog on Nature Network
  • “We are seeing the transformation of knowledge from something that is primarily conveyed in paper formats into something else: a computable graph, in which the knowledge is written in formats that computers can understand and interconnect, based on the same technologies that underlie the internet and web. Paper technology simply contains expressions of ideas, but the very technology of paper makes integration of ideas very difficult, if not impossible… the idea of “the paper” as the core container for knowledge is dying, and technology will be the killer. This transformation is happening first, like the transformation of documents to the Web, in the sciences.”
• PolishMyWriting.com
  • Checks your writing against more than 7000 rules of plain language. I put a couple of draft essays through it, and found about half the suggestions helpful, which is a pretty good batting average.
• Upton Sinclair: “It is difficult to get a man to understand something when his salary depends upon his not understanding it.”
  • It’s curious that often the last people to really grok that a profession is disappearing is people within the profession itself. This quote of Sinclair’s summarizes a part of why that is.
• Dorothea Salo: Innkeeper at the Roach Motel
  • ‘Trapped by faculty apathy and library uncertainty, institutional repositories face a crossroads: adapt or die. The “build it and they will come” proposition has been decisively proven wrong. Citation advantages and preservation have not attracted faculty participants, though current-generation software and services offer faculty little else. Academic librarianship has not supported repositories or their managers. Most libraries consistently under-resource and understaff repositories, further worsening the participation gap. Software and services are wildly out of touch with faculty needs and the realities of repository management. These problems are not insoluble, but they demand serious reconsideration of repository missions, goals, and means.’

Click here for all of my del.icio.us bookmarks.

“Can you give me a simple, concrete explanation of how quantum computers work?”

I’ve been asked this question a lot. I worked on quantum computing full time for 12 years, wrote 60 or so papers, and co-authored the standard text. But for many years the question stumped me. I had several pat answers, but none really satisfied me or my questioners.

It turns out, though, that there is a satisfying answer to the question, which anyone can understand if they’re willing to spend some time concentrating hard.

To understand the answer, let’s back up and think first about why big media outlets like the New York Times and the Economist regularly run stories about quantum computers.

The reason is that quantum computer scientists believe quantum computers can solve problems that are intractable for conventional computers. That is, it’s not that quantum computers are like regular computers, but smaller and faster. Rather, quantum computers work according to principles entirely different than conventional computers, and using those principles can solve problems whose solution will never be feasible on a conventional computer.

In everyday life, all our experience is with objects which can be directly simulated by a conventional computer. We don’t usually think about this fact, but movie-makers rely on it, and we take it for granted - special effects are basically just rough computer simulations of events that would be more expensive for the movie makers to create in real life than they are to simulate inside a computer. Much more detailed simulations are used by companies like Boeing to test designs for their latest aircraft, and by Intel to test designs for their latest chips. Everything you’ve ever seen or done in your life - driving a car, walking in the park, cooking a meal - all these actions can be directly simulated using a conventional computer.

Because of this, when we think in concrete terms we invariably think about things that can be directly simulated on a conventional computer.

Now, imagine for the sake of argument that I could give you a simple, concrete explanation of how quantum computers work. If that explanation were truly correct, then it would mean we could use conventional computers to simulate all the elements in a quantum computer, giving us a way to solve those supposedly intractable problems I mentioned earlier.

Of course, this is absurd! What’s really going on is that no simple concrete explanation of quantum computers is possible. Rather, there is an intrinsic quantum gap between how quantum computers work, and our ability to explain them in simple concrete terms. This quantum gap is what made it hard for me to answer people’s requests for a concrete explanation. The right answer to such requests is that quantum computers cannot be explained in simple concrete terms; if they could be, quantum computers could be directly simulated on conventional computers, and quantum computing would offer no advantage over such computers. In fact, what is truly interesting about quantum computers is understanding the nature of this gap between our ability to give a simple concrete explanation and what’s really going on.

This account of quantum computers is distinctly at odds with the account that appears most often in the mainstream media. In that account, quantum computers work by exploiting what is called “quantum parallelism”. The idea is that a quantum computer can simultaneously explore many possible solutions to a problem. Implicitly, such accounts promise that it’s then possible to pick out the correct solution to the problem, and that it’s this which makes quantum computers tick.

Quantum parallelism is an appealing story, but it’s misleading. The problem comes in the second part of the story: picking out the correct solution. Most of the time this turns out to be impossible. This isn’t just my opinion, in some cases you can mathematically prove it’s impossible. In fact, the problem of figuring out how to extract the solution, which is glossed over in mainstream accounts, is really the key problem. It’s here that the quantum gap lies, and glossing over it does nothing to promote genuine understanding.

None of my discussion to now actually explains how quantum computers work. But it’s a good first step to understanding, for it prepares you to expect a less concrete explanation of quantum computers than you might at first have hoped for. I won’t give a full description here, but I will sketch what’s going on, and give you some suggestions for further reading.

Quantum computers are built from “quantum bits”, or “qubits” [1], which are the quantum analogue of the bits which make up conventional computers. Here’s a magnified picture of a baby quantum computer made up of three Beryllium atoms, which are used to store three qubits:

(credit)

The atoms are held in place using an atom trap, which you can’t see because it’s out of frame, but which surrounds the atoms, holding them suspended in place using electric and magnetic fields, similar to the way magnets can be used to levitate larger objects in the air.

The atoms in the picture are about three micrometers apart, which means that if you laid a million end to end, they wouldn’t quite span the length of a living room. Very fine human hair is about 20 micrometers in diameter - it’d pretty much cover the width of this photo.

The atoms themselves are about a thousand times smaller than the spacing between the atoms. They look a lot bigger in the picture, and the reason is interesting. Although the atoms are very small, the way the picture was created was by shining laser light on the atoms to light them up, and then taking a photograph. The particles making up the laser light are much bigger than the atoms, which makes the picture come out all blurry; the photo above is basically a very blurry photograph of the atoms, which is why they look so much bigger than they really are.

I called this a baby quantum computer because it has only three qubits, but in fact it’s pretty close to the state of the art. It’s hard to build quantum computers, and adding extra qubits turns out to be tricky. Exactly who holds the record for the most qubits depends on who you ask, because different people have different ideas about what standards need to be met to qualify as a genuine quantum computer. The current consensus for the record is about 5-10 qubits.

Okay, a minor alert is in order. I’ve tried to keep this essay as free from mathematics as possible, but the rest of the essay will use a little high-school mathematics. If this is the kind of thing that puts you off, do not be alarmed! You should be able to get the gist even if you skip over the mathematical bits.

How should we describe what’s inside a quantum computer? We can give a bare-bones description of a conventional computer by listing out the state of all its internal components. For example, its memory might contains the bits 0,0,1,0,1,1, and so on. It turns out that a quantum computer can also be described using a list of numbers, although how this is done is quite different. If our quantum computer has n qubits (in the example pictured above n = 3), then it turns out that the right way to describe the quantum computer is using a list of 2ⁿ numbers. It’s helpful to give these numbers labels: the first is s₁, the second s₂, and so on, so the entire list is:

What are these numbers, and how are they related to the n qubits in our quantum computer? This is a reasonable question - in fact, it’s an excellent question! Unfortunately, the relationship is somewhat indirect. For that reason, I’m not going to describe it in detail here, although you can get a better picture from some of the further reading I describe below. For us, the thing to take away is that describing n qubits requires 2ⁿ numbers.

One result of this is that the amount of information needed to describe the qubits gets big really quickly. More than a million numbers are needed to describe a 20-qubit quantum computer! The contrast with conventional computers is striking - a conventional 20-bit computer needs only 20 numbers to describe it. The reason is that each added qubit doubles the amount of information needed to describe the quantum computer [2]. The moral is that quantum computers get complex far more quickly than conventional computers as the number of components rises.

The way a quantum computer works is that quantum gates are applied to the qubits making up the quantum computer. This is a fancy way of saying that we do things to the qubits. The exact details vary quite a bit in different quantum computer designs. In the example I showed above, it basically involves manipulating the atoms by shining laser light on them. Quantum gates usually involve manipulating just one or two qubits at a time; some quantum computer designs involve more at the same time, but that’s a luxury, it’s not actually necessary. A quantum computation is just a sequence of these quantum gates done in a particular order. This sequence is called a quantum algorithm; it plays the same role as a program for a conventional computer.

The effect of a quantum gate is to change the description s₁, s₂,… of the quantum computer. Let me show you a specific example to make this a bit more concrete. There’s a particular type of quantum gate called a Hadamard gate. This type of gate affects just one qubit. If we apply a Hadamard gate to the first qubit in a quantum computer, the effect is to produce a new description for the quantum computer with numbers t₁, t₂,… given by

and so on, down through all 2ⁿ different numbers in the description. The details aren’t important, the salient point is that even though we’ve manipulated just one qubit, the way we describe the quantum computer changes in an extremely complicated way. It’s bizarre: by manipulating just a single physical object, we reshuffle and recombine the entire list of 2ⁿ numbers!

It’s this reshuffling and recombination of all 2ⁿ numbers that is the heart of the matter. Imagine we were trying to simulate what’s going on inside the quantum computer using a conventional computer. The obvious way to do this is to track the way the numbers s₁, s₂,… change as the quantum computation progresses. The problem with doing this is that even a single quantum gate can involve changes to all 2ⁿ different numbers. Even when n is quite modest, 2ⁿ can be enormous. For example, when n = 300, 2ⁿ is larger than the number of atoms in the Universe. It’s just not feasible to track this many numbers on a conventional computer.

You should now be getting a feeling for why quantum computer scientists believe it is infeasible for a conventional computer to simulate a quantum computer. What’s really clever, and not so obvious, is that we can turn this around, and use the quantum manipulations of all these exponentially many numbers to solve interesting computational problems.

I won’t try to tell that story here. But if you’re interested in learning more, here’s some reading you may find worthwhile.

In an earlier essay I explain why conventional ways of thinking simply cannot give a complete description of the world, and why quantum mechanics is necessary. Going a little further, an excellent lay introduction to quantum mechanics is Richard Feynman’s QED: The Strange Theory of Light and Matter. It requires no mathematical background, but manages to convey the essence of quantum mechanics. If you’re feeling more adventurous still, Scott Aaronson’s lecture notes are a fun introduction to quantum computing. They contain a fair bit of mathematics, but are written so you can get a lot out of them even if some of the mathematics is inaccessible. Scott and Dave Bacon run excellent blogs that occasionally discuss quantum computing, and their blogrolls are a good place to find links to other quantum bloggers.

Finally, if you’ve enjoyed this essay, you may enjoy some of my other essays, or perhaps like to subscribe to my blog. Thanks for reading!

Acknowledgements

Thanks to Jen Dodd and Kate Nielsen for providing feedback that greatly improved early drafts of this essay.

About the author

Michael Nielsen is a writer living near Toronto, and working on a book about The Future of Science. If you’d like to be notified when the book is available, please send a blank email to the.future.of.science@gmail.com with the subject “subscribe book”. You’ll be emailed to let you know when the book is to be published; your email address will not be used for any other purpose.

Footnotes

• Connectedness: Teaching executives to see social capital, by Ron Burt and Don Ronchi
  • “we aren’t very good at recognizing bridging opportunities (aka ’structural holes’), and hence we miss out all the time on chances for innovation and profit… people can learn to recognize structural holes. Surprisingly, even a little learning goes a long way in this regard.”
• Repository Support Project Blog Directory
  • Useful list of blogs related to academic repositories.
• Wired Campus: When Professors Create Social Networks for Classes, Some Students See a ‘Creepy Treehouse’
  • The comments are well worth reading.

Click here for all of my del.icio.us bookmarks.