Archive

Tag Archives: Georgia Tech

"I’ll see your 10,000 and raise you….

August 16, 2011
Loose Cannons, Practice, Technology
3 Comments

In “Dancing with the Stars” I talked about what a classroom with 10,000 students might be like. The transformation of higher education has begun, and the pace of that change is accelerating.

Dick Lipton’s blog Godel’s Lost Letter has since attracted tens of thousands more.  It is a virtual seminar that, for example, coordinated a global effort to referee an important paper in the theory of algorithms.  At times, the number of viewers topped 100,000. Now Stanford’s Peter Norvig and Sebastian Thrun are offering an online course in artificial intelligence that will enroll 58,000 students.

On September 12, I will join with 60 or so colleagues to offer a MOOC for tens of thousands of students.  Georgia Tech  students will get credit, and others will get badges that could be convertible to credit if they ever enroll at Tech.  Other institutions will announce their approaches to certifying achievement in the course. A MOOC is a Massive Open Online Course, a style of college-level teaching that was pioneered by George Siemens and Stephen Downes. The first MOOC, offered in 2008 by George and Stephen was devoted to the subject of their research, a style of learning called connected connectivism. It attracted 10,000 students.

The 2011-12 MOOC is all about transforming university learning and the organizers hope it will attract a much wider global audience.  They are calling it the Mother of all MOOCS.

The course will also be a C21U experiment on self-certification, a concept I discussed in my book. Where will this all lead?  It’s far too soon to predict an outcome, but within the last year, the number of experiments in higher education has exploded.  If you believe like me that innovative change is just what traditional colleges and universities need, that’s a good thing. The way to innovate is to try out lots of ideas.

Big Animal Pictures

April 4, 2011
Practice
2 Comments


I’ve been spending more time with alumni.  Zvi Galil, the new dean of computing at Georgia Tech — my successor — has been on a national tour to get acquainted with recent graduates. I accompany him whenever I can to make introductions and to generally help smooth his transition.  Not that he needs it.  Zvi was dean of engineering at Columbia for many years and knows how to get alumni to talk honestly about their undergraduate experiences. We were having lunch with a group of recent graduates when I heard Zvi ask someone at the end of the table, “What’s the one thing you wish we had taught you?”

The answer came back immediately: “I wish I had learned how to make an effective PowerPoint™ presentation!”  If the answer had been “more math” or “better writing skills” I would have filed it away in my mental catalog of ways to tweak our degree programs. It’s a constant struggle in a requirement-laden technical curriculum — even one as flexible as our Threads program — to get enough liberal arts, basic science, and business credits into a four year program, so I was prepared to hear that these young engineers wanted to know more about American history, geology, or accounting. After all, I am a former dean.  I had heard it all before.

But PowerPoint? Everything came to a stop.  Zvi said, “PowerPoint!” It was an exclamation, not a question.  Here’s how the rest  of the conversation unfolded” “Look, the first thing I had to do was start making budget presentations. I had no idea how to make a winning argument.”  From the across the table: ” Yeah, we learned how to make technical presentations, but nobody warned us that we’d have to make our point to a boss who didn’t care about the technology.”  “It’s even worse where I work,” said a young woman. “Everybody in the room has a great technology to push.  I needed to know how to say why mine should be the winner.”  And so it went.  This was not a PowerPoint discussion.  We were talking about Big Animal Pictures. If you understand Big Animal Pictures, you understand  how to survive when worlds collide.

David Stockman directed Ronald Reagan’s Office of Management and Budget (OMB) from 1981 to 1985.  He was a technician.  A financial engineer. He had a Harvard MBA, and spent the early part of his career on Wall Street with Solomon Brothers and Blackstone. It was a checkered career, and if you take seriously the accounts in his memoir of the Reagan years, he never really understood that he was caught between colliding worlds. Which brings me to Big Animal Pictures.

Stockman was a conservative deficit hawk who thought his job was to restore fiscal sanity.  Reagan had beaten Jimmy Carter in part by painting the Democrats as financially irresponsible.  David Stockman’s job was to fix that, and that meant budget-cutting.  Defense Secretary Caspar Weinberger thought that Reagan had been elected to restore America’s military might. Weinberger’s job was to pump more money into defense budgets.  Stockman and Weinberger were on a collision course, and for a year they traded line-item edits to the federal budget. This was a technical duel. Stockman and Weinberger both had considerable quantitative skills. It was a bureaucratic game that Weinberger had learned to play when he worked for Reagan in California, but there was a deepening recession. In the end, it appeared that DoD would have to make do with the 5% increase that the White House was proposing. It was a spending increase that Stockman believed was unwise and unaffordable.

Weinberger’s proposal was 10%.  Stockman could barely contain himself. It set up a famous duel in the form of a budget briefing with Reagan playing the role of mediator. It was going to be a titanic debate.

Stockman showed up with charts, graphs and projections.  The stuff that the OMB Director is supposed to have at his fingertips. Weinberger came armed with a cartoon, and walked away with his budget request more or less intact.

Weinberger’s presentation was a drawing of three soldiers. On the left was a small, unarmed, cowering soldier — a victim of years of Democratic starvation. The  bespectacled soldier in the middle — who bore a striking resemblance to Stockman — was a little bigger, but carried only a tiny rifle. This was the army that David Stockman wanted to send to battle. The third solder was a  menacing fighting machine, complete with flak jacket and an M-90 machine gun. It was the soldier that Weinberger wanted to fund with his defense budget.  Weinberger won the budget debate with Big Animal Pictures.

Stockman was appalled:

It was so intellectually disreputable, so demeaning, that I could hardly bring myself to believe that a Harvard educated cabinet officer could have brought this to the President of the United States. Did he think the White House was on Sesame Street?

Stockman and many analysts concluded that the episode revealed something deep about Reagan’s intellectual capacity. Maybe so, but I think it revealed more about Weinberger’s insight into what it takes to carry an argument when the opposing sides can each make a strong technical case for the correctness of their position: argue for the importance of the end result, not for the correctness of how you will achieve it. It is a classical colliding worlds strategy.

Michael Dell’s 1987 private placement memorandum for Dell Computer Corporation was a Big Animal Picture. Buying computers was a hassle when Dell started his dormitory-based business in 1984.  By 1987, PC’s Limited had sold $160M worth of computers based on a simple strategy: eliminate the middle man, get rid of inventories, and give customers a hassle-free way to buy inexpensive, powerful IBM-compatible computers.  In the midst of a stock market crash, Michael Dell managed to raise $21M based on a short document that ignored the conventional view that private placement business plans had to be highly technical:

Dell has sold over $160 million of computers and related equipment on an initial investment of $1,000. The Company has been profitable in each quarter of its existence, and sales have increased in each quarter since the Company’s inception.

Tacked onto the memorandum, almost as an afterthought were letters from customers — inquiries from people who were interested in buying computers from Michael Dell and testimonial from owners of his made-to-order PCs who wanted to buy more of them.  It was short (45 pages with the letters attached) and, aside from a few pro-forma financials to explain what would be done with the new money, it was almost entirely devoted to painting a picture of what success looked like to Michael Dell.

A copy of the original Dell memorandum wound up on my desk in late 1998.  At the time, my Bellcore department heads were struggling to define businesses that could either be spun out of the company or funded as internal startups. I was drowning in  highly technical market forecasts and details of patent disclosures. Each new spreadsheet screamed: “Idiot! Just look at this equation.  It’s obvious why our approach is better than everyone else’s.”  One afternoon, in exasperation,  I threw Michael Dell’s private placement memorandum on my conference table and said “Make me a presentation that looks like this.” The room got very quiet as they realized what was going on.  I was asking for Big Animal Pictures.

We started four businesses within 18 months.  Three were spun out  and made a modest amount of money for the company and the founders.  We ran one as an internal start-up. It did not do nearly so well. One of the key factors was that we could not duplicate Michael Dell’s Big Animal Picture.

This is not a lesson that engineers and scientists learn easily. In fact, when presented with overwhelming evidence that business decisions are seldom made on the basis of technical elegance and correctness, engineers retreat to the safer ground staked out by David Stockman: “Do you think we are on Sesame Street?” The answer is “Yes!”  Successful engineers and scientists know all about Big Animal Pictures.

Paul R.  Halmos was one of the great mathematicians of the 20th century. He studied the most abstract topics imaginable. One of his crowning achievements, for example, was to create an entire algebraic theory to describe mathematical logic, which was itself an abstract mathematical theory to explain symbolic logic. Symbolic logic was, in turn, an abstract explanation of the kind logic used by Aristotle, and Aristotle’s logic was the formalization of correct patterns of human  inference. Halmos did not deal in uncomplicated matters.

How did Paul Halmos counsel young mathematicians to present their work in public?

A public lecture should be simple and elementary; it should not be complicated and technical. If you believe you can act on this injunction (“Be Simple”) you can stop reading here, the rest of what I have to say is, in comparison, just a matter of minor detail.

The mistake, Paul Halmos noted in his essay How to talk Mathematics is thinking that a simple lecture talks down to the audience. It does not. Halmos (or PRH as he sometimes called himself) seems to have understood worlds in collision.   Of course, a simple lecture in PRH world might open with the phrase “…as far at Betti numbers go, it is just like what happens when you multiply polynomials,” so it’s a sliding scale.

No matter what you’re doing in the technical world, learning how Big Animal Pictures work is a valuable thing.  I sometimes sit on review panels to decide on research funding.  I recently advised a young scientist to use Big Animal Pictures.  She had five minutes to present her work and I knew that the competition would be strong.  Her first instinct was to jump into the technical meat of her research to give the reviewers a feeling for why her approach was better than other approaches. My advice was to not do that.  I wanted her to literally give a BAP presentation that would inform the panel about the importance of her research and why they should care about it.  I later found out that other colleagues had given her identical advice, which she apparently followed with great success.

And it doesn’t matter which of the colliding worlds you are on.  BAPs are always a good idea. My colleague Wenke Lee was recently called upon to give a presentation on the state of computer security research to a  group of mathematicians.  It was all about how powerful mathematics can be used to exploit security flaws and vulnerabilities. Wenke resisted the temptation to dive into the technical details of botnet attacks.  It is, after all, a subject he knows well and he probably would have had fun demonstrating his prowess. But here is how Wenke began his lecture.

He went on for another twenty minutes, but he really didn’t need to. Everyone got the point in the first thirty seconds.

Climate Change, Ivory Towers and The Journal of Irreproducible Results

December 8, 2009
Policy
Leave a comment

There’s a kerfuffle on the eve of the United Nations Climate Change Conference in Copenhagen. 1,700 email messages  that were supposed to be stored on a secure server somehow found their way to open servers and were rapidly picked up by bloggers and others, who jumped on the opportunity to use the sometimes embarrassing messages to discredit  the overwhelming consensus of climate scientists that the earth is warming at an alarming rate and that human activity is the most likely cause. Aside from the shocking coincidence of events — what are the chances that a massive, worldwide fraud would be exposed at the same time the conspirators are getting together to impose their new world order? — and the uproar among climate scientists — who are launching ad-hominem attacks at every skeptic who pokes his head above ground — are there other lessons to be drawn from this shameless bit of theater?  My Georgia Tech colleague, climate scientist Judith Curry, hit the nail on the head when she  pointed out that: (1) there is really nothing in the released messages that discredits published scientific results and (2) scientists are being incredibly counterproductive by retreating into their Ivory Towers and passing up the opportunity to educate and engage both skeptics and the public.  Her Open Letter to Graduate Students and Young Scientists should be required reading for everyone interested in how to keep worlds from colliding:

…even if the hacked emails from HADCRU end up to be much ado about nothing in the context of any actual misfeasance that impacts the climate data records, the damage to the public credibility of climate research is likely to be significant. In my opinion, there are two broader issues raised by these emails that are impeding the public credibility of climate research: lack of transparency in climate data, and “tribalism” in some segments of the climate research community that is impeding peer review and the assessment process.

For “climate science” you can substitute “innovation” and the message is the same. If you’ve circled the wagons and are shooting at anything that moves, the easy target is public understanding of not only science but innovation in general.  The American public is not interested in the long-term thinking required to make sense out of squabbles like this. There are simply not enough people like San Diego Florist Steve Boigon, who — according to the New York Times — downloads MIT physics lectures because he  finds that:

I walk with a new spring in my step and I look at life through physics-colored eyes.

Curry did not go after the easy targets. Instead, she talked honestly to students about the importance of climbing down from the Ivory Tower. The interactive relationship between basic science, technological innovation and public policy — what Donald Stokes calls Pasteur’s Quandrant —  is a hot topic these days, because  so many important societal issues can only be resolved at their intersection.

There’s a veil that conceals the inner workings of creative science and engineering  from the lay public, and attempts to lift it sometimes produce  bizarre reactions.  I was once struck speechless  at an all-hands meeting when one of my engineers stood to scold  the  CEO for making product decisions because he knew “nothing about electronics.”  A prominent member of my Board of Advisers at the National Science Foundation once countered criticism of his particularly cumbersome approach to software development by angrily proclaiming,  “Programming is like playing a piano.  Only virtuosos should do it!”  A world-renowned engineer once responded to an essay critical of his methods by widely distributing a letter entitled “On a Political Pamphlet from the Middle Ages.”  I was one of the young authors who was at the receiving end of that one.  When  outsiders try to lift the veil, the best course is to repair to the upper reaches of the Ivory Tower, hope that the hubbub goes away, and shoot down if it doesn’t.

It is a world view that is somehow wired into university training. The Medieval regalia, semi-religious icons,  and murmured  incantations that convey special status on the conferees reinforce the impression at every college commencement that something mystical has taken place. Science textbooks are uniformly silent on how science is done, presenting instead the subject as a linear, completed work — orderly in progression and tidy in its use of knowledge.  Nearly every engineering textbook guides  readers through well-rehearsed exercises to successful completion of design tasks. Why would anyone want to learn how to build a bridge that falls down?

Insiders, of course, know differently. What takes place behind the curtain is as important as the finished product.  Some of the best technical books ever written lift the veil.  Proofs and Refutations by Imre Lakatos describes  the centuries-long frustration of mathematicians  trying — and repeatedly failing —  to precisely define polyhedra.  The process led some of  the greatest mathematical results of all time. Why Buildings Fall Down by Mario Salvatori and To Engineer is Human by Henry Petrosky are both compelling arguments that progress in  engineering is inextricably tied to understanding engineering failure.  Insiders know that failure is part of the package.  That’s exactly what makes the most outrageous of the climate change attacks so improbable.

There is a sub-genre of humor devoted to obvious, boundlessly incompetent scientific failure, real or imagined.  The Journal of Irreproducible Results is perhaps the defining publication that holds technical vanity up to ridicule. An article entitled Peaceful Use of Nuclear Explosives helpfully noted that

Development of hydro power in the desert of North Africa awaits only the introduction of water

My personal favorite medical discovery was an announcement entitled The Incidence and Treatment of Hyperacrosomia in the United States:

Some very famous Americans  have indeed been afflicted with Acute Hyperacrosomia, among them Abraham Lincoln, George Washington and Lyndon Johnson.  Their condition is readily apparent upon comparison with normal individuals such as Napoleon Bonaparte, Truman Capote  and Dick Cavett…..Since the male population does express the condition to a higher degree, it falls primarily to the female population to objectively consider the risks of involving themselves with hyperacrosomic males…

The jokes are so well-known that Henry R. Lewis apparently had not second thoughts when he wrote The Data Enrichment Metho d:

The following remarks are intended as a non-technical exposition of a method which has been promoted (not by the present author) to improve the quality of inference drawn from a set of experimentally obtained data.  The power of the method lies in its breadth of applicability and in the promise it holds in obtaining more reliable results without recourse to the expense and trouble of increasing the size of the sample of data.

I have a hazy understanding of the data manipulation charges that climate skeptics are leveling at researcher, but I am pretty sure that The Data Enrichment Method was not involved.  There is also the issue of transparency that is specific to climatologists, but Curry handles that well. And then there are the charges that editors of journals were unduly influenced by political considerations.  Like the Inspector in Casablanca, I would be shocked — truly shocked — to hear that hundreds, perhaps thousands, of smart, educated, and highly ambitious people make decisions based on self-interest. The secret that Curry reveals is that it may be regrettable, but  it doesn’t matter in the long run.  Science is not an orderly, axiomatic progression of knowledge. It is a social process.

Even a brief peek under the veil would be enough to convince many fair-minded skeptics that if there were another, compelling, contradictory analysis of the same data, it would have by now appeared in a reputable scientific journal.  Why?  Because it would be a career-making result.  The article would write itself.  What editorial board could long resist publishing an epochal article?  History teaches that political manipulation is much more likely to focus on who gets priority as multiple groups rush to publish simultaneously.  It’s a to maintain a conspiracy when everyone is looking out for himself.  None of this means that everything that has been published is correct. It just means that it’s very unlikely that the shrill cries of  systematic fraud have any validity.



So strong is the urge to seek out systematic scientific fraud, that there is a magazine devoted to the subject. The Skeptical Inquirer (SI) is a kind of companion to The Journal of Irreproducible Results. It specializes in debunking academic myths and scientific hoaxes.  It has over the years exposed magicians, perpetual motion charlatans, creationists, and hundreds of scientific frauds.  Who are these crusaders?  They are the very power brokers that would have to be co-opted if the climate change conspiracy theorists were right.  Here’s a partial list of SI Fellows:

If there is  a less easily manipulated group under one banner, I have not seen it.

Judy Curry’s Open Letter does not only apply to climate scientists. It applies to every boardroom that squashes the discussion of how innovation takes place and every executive suite where technologists are too busy innovating to engage seriously with corporate management.  Of course, it also applies to the easy targets — facile business leaders who confuse near term planning with technical progress and are too quick to jump to the “bottom line” — but that discussion will have to wait for another post.

Another Seat at the Table: Grace Hopper and Diversity

September 30, 2009
Comment and Review
2 Comments

George Heilmeier was a master  of getting a seat at the table for R&D (see “Guess Who’s Coming to Dinner“), but worlds can also collide when they expand, and as the 2009 Grace Murray Hopper Celebration of Women in Computing opens this week in Tuscon, I thought it would be a good time to mention the role that diversity plays in defining the role of innovation and innovators and to recall that Grace Hopper was a pioneer at the table.

Grace Hopper was already a legendary figure in computing when I met her as a graduate student in 1969.  Retired from the Navy but recalled to active duty (eventually with the rank of Admiral) she lectured widely on the history of computers and programming languages.  Her role in developing the first compiled programming languages for Harvard’s Mark I Computer and later with Eckert-Mauchley Computer Corporation inspired early computer scientists — including this one — to concentrate on programming languages.  Her lectures were always packed. She took her seat at the table in the days when there were only a handful of women in the computing industry and although she counseled young women entering the field, she did not talk about her role in expanding the boundaries of computing in any of the interactions I had with her. She preferred to concentrate on technology and where it was heading.

The celebration that carries her name is a series of conferences designed to highlight the contributions of women to the field of information technology.  It has special significance to me because I spent much my tenure as Dean of Computing at Georgia Tech expanding the boundaries of computing.

This idea of expanding boundaries was born  in 2002 when computing education was in a downward spiral after the dot-com bust.   This may not not sound like  colliding worlds, but in fact our solution to the problem of declining enrollments  was to remake undergraduate computing education with a new face that was more inclusive — more open to broader participation.  That meant lifelong, relevant education focused on combining student interests and real-world needs and impact. It meant ignoring disciplinary boundaries, which was something not all universities did.

It also meant, literally, “new faces” in classrooms — on both sides of the podium — and laboratories.  Shortly after I became dean, Beth Mynatt, now Director of Georgia Tech’s GVU Center, said to me, “You know, you can’t have a more diverse college without intellectual diversity.”  So our initiative  A New Face of Computing at Georgia Tech became defined in part by the New Faces who were brought into the field by an explosion of new programs and degrees.

As the Grace Hopper Celebration kicks off this week in Houston, it’s a good time to be reminded that there’s always room for more at the table.  It keeps worlds from colliding, expands social networks, and promotes innovation.  Let’s celebrate that.