jep.jpg (13389 bytes)

Chaos Manor Special Reports

Intellectual Capital (2)

Wednesday, December 11, 2002

Email Jerry


Chaos Manor Home

View From Chaos Manor

Reader Mail



Special Reports

Book &; Movie Reviews

Picture Gallery


Table of Contents

What's New

The BYTE Fiasco


This is a Part Two of a collection of the columns I wrote for Intellectual Capital during the all too brief years that publication existed. Published by Pete DuPont (former Governor of Delaware), it was a splendid collection of ideas and thoughts.  It was also a victim of the dot bust.

These are presented more or less in the order they were written. The final as it appeared in Intellectual Capital was somewhat different: the editor, Bob Kolasky, managed to make them shorter, and, I hate to admit it, better.  Not many editors can do that with my work. In most cases the titles were supplied by the editors.

Part One of these reports.

Part Three

Part Two June - December 1998





Doomsday? Not Just Yet.

Jerry Pournelle

We have always had doomsayers. There were more in the seventies, but we have our share today. Some are the same people. It was a great deal more fashionable to pretend to take Paul Ehrlich seriously in the 70's than now, but he's still around. You'll recall that Ehrlich predicted mass famines and rising natural resource prices first in the 70's, then in the 80's. I don't know what he's predicting now, but it's likely to be doom and gloom; it generally is, but taking him seriously is a luxury available mostly to tenured professors.

One of those is MIT economist Paul Krugman, who continues to predict rising commodity prices and a great deal more. In a recent issue of The Red Herring  he argues forcefully that inflation will return, there will be raw materials crunches, and America will lose confidence, then slide into a recession if not depression. Perhaps. That could all happen, but if so, it won't be some economic law of nature driving the system, just good old fashioned politics as the American nomenklatura tries to gain control over more and more of our lives. As the late Julian Simon used to observe, economic freedom will generally take care of purely economic matters. Of course Paul Ehrlich likes to say that Julian "Simon is the absolute equivalent of the flat-earthers," but that hasn't stopped Simon from being right. The only real question is whether the politicians will grant the American people the necessary economic freedom.

Of more interest is Krugman's prediction that the growth of the Internet will slow drastically, so that "by the year 2005 or so it will become clear that the Internet's impact on the economy has been no greater than the fax machine." He also says that you shouldn't study computer science, because the rate of technological change is slowing, and "the number of jobs for IT [Information Technology] specialists will decelerate, then actually turn down; ten years from now, the phrase information economy will sound silly."

So what are we to make of this? First, it's clear that Krugman doesn't understand modern technology at all. Far from slowing, the rate of technological change is accelerating. Technology moves by a series of S curves, slow growth at first, then very steep progress over time, then slowing to a plateau; but modern information technology growth comes from the sum of a lot of those S curves, so that as one slows another is taking off. Moore's law !<I could put a link to a previous definition of that on the IC web page if you like, or you can> has a number of cycles to go, and that's just chip technologies.

The biological sciences are just taking off. Sixty Minutes is talking about transplants of organs from pigs genetically altered so that they carry a few human genes, thus fooling human immune systems into not rejecting a new heart, kidney, lung, spleen… We grow drought resistant vegetables through genetic alteration. And that too is only the beginning. If you thought technological change was bewildering in the last decade, you ain't seen nothing yet.

Krugman's denigration of the Internet is no wiser, but it is of a piece with conventional views. I was recently at a conference in which Robert Wussler, former President of CBS TV, laughed off the Internet as any threat to conventional entertainment business. The cable companies and big networks will continue to dominate for years to come, "because people don't want to watch TV on a computer." Dream on. First, with Moore's law inexorably making electronics faster, more complex, and cheaper, for all practical purposes your TV will BE your computer. Big screen high resolution monitors are now priced down where color TV sets were not long ago. They'll get cheaper, and believe me, once you have seen a movie in digital format on a high resolution monitor, you will simply hate the crummy low resolution junk TV has got you accustomed to. Second, what do the cable companies and networks have that you don't have? Primarily they've got a lot of money invested in expensive equipment, most of which is rapidly becoming obsolete. Play Inc.'s ( Trinity box for under ten grand will give your computer professional TV editing capabilities that no network or cable company had not long ago, while Sony will sell you a digital video camera good enough for professional work. The equipment is there for anyone interested in producing a TV show. The other thing the cable companies have is a distribution network. Uh - isn't the Internet a distribution system? It may not be fast enough for real time full screen video, but satellites, cable modem, ISDN, and general upgrading of the net connections are changing that even as you read this.

Krugman has one point, but I don't think he understands it. It's probably not true that the need for IT specialists will actually decline, but it's probably not the fastest growing part of the new high tech job corps either. Just as most TV network employees aren't engineers and technicians, most of the new information company employees won't be programmers and computer scientists. Teaching the machine to do elementary accounting was enough to make Dan Bricklin rich with VisiCalc, but that was then: now the big bucks come to people who think up new things for the computer to do. Printers and book binders once made as much money as authors and publishers, but that hasn't been true for a long time.

We aren't doomed. We aren't running out of natural resources, witness that their prices keep falling despite all predictions that they'll soar. Oil is now about the same price in constant dollars as it was before OPEC tried to create a crunch and Saddam Hussein set fire to a quarter of the world's supply. We're all right, Jack.

What of the rest of the world? Now it's certainly true that if we are to make the rest of the world as rich as the US, we're going to need new sources of raw materials. And just last Friday, a private rocket company, Rotary Rocket !( broke ground for their new facilities in Mojave, California. Ninety percent of the resources easily available to mankind aren't on the Earth, and without NASA dragging us down American enterprise will go get them-just as Julian Simon would have predicted. Incidentally, Rotary Rocket expects to be able to let CNN charter an orbital vehicle to send up reporters for live coverage of the opening of Space Station. I see no reason why they can't do that.

In the academic world we're still doomed, technology progress is grinding to a halt, and we're going to run out of resources if we haven't polluted ourselves to death. In the real world, the gross world product is higher than ever, economic problems kill fewer people than politics, resource prices are falling, and the air and water are cleaner than they were a decade ago.

Also in the real world, Paul Ehrlich got a lot of miles out of shouting "Doom!", Donella and Dennis Meadows got rich out of a book called The Limits to Growth, and the millionaires only Club of Rome got a lot of publicity by telling the rest of us to get used to poverty. But then the world has never been a particularly fair place.

- 30 -


Just What Is the Business of America?

Jerry Pournelle

And why beholdest thou the mote that is in thy brother's eye, but considerest not the beam that is in thine own eye? Thou hypocrite, first cast out the beam out of thine own eye; and then shalt thou see clearly to cast out the mote out of thy brother's eye. Matthew, 7:2-5

The United States seeks to plant democracy and secure human rights throughout the world. While these may be admirable ambitions, they may also be impossible; indeed they may be incompatible. In Indonesia, democracy apparently means the freedom to burn alive Indonesian citizens of Chinese ancestry without regard to how long they may have lived in Indonesia. In Kosovo, democracy means the freedom to expel the Serbs; in the wider Yugoslavia that includes Kosovo, democracy means the freedom to suppress the Albanians.

We don't yet have troops in Indonesia and Kosovo, but one wonders how long it will be. We do have troops in Haiti, with remarkably little success; and we had them in Somalia, a region starkly similar to Albania. Of course in Somalia we endured a small but quite sharp defeat and left with all haste and not a little shame.

Why are we doing these things?

This isn't a trivial question, nor is it intended as an attack on the current administration. The fact is that few in Washington seem to have any idea of what the United States is about. What should our policies be? Why do we, as a nation, exist?

Rome had a statement of purpose: "To protect the weak, and make humble the proud." The Roman Republic and later the Empire may not always have adhered to that purpose, but few doubted that was the ideal policy. The business of Rome the superpower was the imposition of order on a disorderly world.

Some would argue that is the purpose of the United States; but I wonder. I doubt you could get the Senate, much less both Houses of Congress, to adopt that statement of purpose, and while the present White House might be a bit more inclined to the view, I doubt again you could get a consensus in the Cabinet.

The Constitutional Framers and their successors would have rejected the notion entirely. World order might have been desirable, but it wasn't something for the United States to impose. "We are the friends of liberty everywhere, but we are guardians only of our own," Adams said; a statement that expressed and governed American foreign policy for more than a century.

More common still was the notion that America would be the "city on a hill", a shining example to the rest of the world. The business of the United States was to be the best and most free land we could be; to set an example for the world, an example so compelling that all would want to follow it; to convert the world to liberty and order by showing the great benefits of our system. The business of America is freedom; which, for many of us, translates into something less lofty: the business of America is business.

Specifically, the best thing the United States can do is continue to be prosperous; to continue the Long Boom which fuels the world growth economy. We've been doing that, and a careful economic analysis indicates we can still do it. We have capital values at 50 and more times earnings, something unsustainable without steady growth, but the growth potential is there. The worst estimates of Internet commerce have it growing from some $20 billion to $200 billion in a decade, and even those figures may be low. Better yet, the Internet helps sustain the boom while reducing inflation by making the market a great deal more efficient. No longer do you get piled up inventories of goods not wanted and growing demand for goods not available; or at least that doesn't happen as much, and as the Net grows ought to be rarer still.

Surely, continuing the Long Boom will do more for the world -- not to mention for us -- than getting ourselves bogged down in places where we know neither the language nor the history? Now, true, the level of effort put into Bosnia, Kosovo, Haiti, and everywhere else we feel compelled to poke our noses isn't all that large and isn't likely to be much of a drain; but it's not resource drain we need to worry about.

Only two things threaten the Long Boom. One is government interference: high tech industry is extraordinarily sensitive to bureaucratic regulation. It has to be, because it is a highly flexible industry. Our high tech people never know what they'll be doing next; they only know that if they have to get permission from bureaucrats they probably won't be doing it.

There's one other thing that can threaten the Long Boom: a shortage of workers with the skills and flexibility needed to sustain it.

And that, alas, is more than probable given our wretched schools. Certainly there are pockets of excellence in the school system, and certainly many of our kids are bright enough to go make something of themselves despite a high school education an 8th grader would have been ashamed of in 1950; but that's hardly the point. Just as we reach a time when the new economy requires both skills and flexibility, we have a school system more concerned with equality of misery than excellence of results.

By no odd coincidence we can trace a great decline in school results to just after massive infusions of Federal aid to education. The mechanisms seem obvious to me: a great deal of Federal money on 'educational research' goes into 'diagnostic tools', which sounds like a good idea, but on reflection translates to "sophisticated excuses for not getting a good result." I say this seems obvious to me, but it's not necessary to my argument: which is that the analogy between Washington and national education, and the US and world order, is, I think, near perfect. Just as the best way to induce the world to follow our advice would be to show how well off we are for taking it, the best way for Washington to improve the national education system would be to perfect its own.

There certainly is no Constitutional bar to that. While one may debate whether Congress has any Constitutional power to make grants in Humanities and the Arts, or, given the entire want of the word 'education' in the Constitution, to impose Federal restrictions or hand out Federal money for education, the question doesn't arise within the District of Columbia. Within the District, Congress is as sovereign as any state within its own borders. Congress can build a national opera house, a national Shakespeare theatre, museums and cultural centers, and perhaps it should. Even Ludwig von Mises, hardly a fan of government action, argued in favor of the legitimacy of the Vienna State Opera. Specifically, though, Congress should make the DC school system a model for the nation, and thus for the world.

Begin simply: teach every DC child to read by the end of second grade. We know this is possible because it used to be done routinely in both city and rural school districts across the country; and surely our national protoplasm is not worse now than it was fifty years ago? As to how to do it, I have some ideas, and Mrs. Pournelle, who was the teacher of last resort in the Los Angeles juvenile justice system has more (see ); but even if I had none, it can't be that hard to do. We did it once before, and what man has done, man can aspire to.

In the worst case, just experiment. Try anything likely to work. Try some things not likely to work. Keep track of what doesn't work, and do less of that. Keep track of what does work, and do more of that. Try a whole bunch of stuff, reward success, and starve failure.

It certainly would be no more expensive than imposing democracy on Haiti and Kosovo; it would be a lot more clearly constitutional; and in the long run, showing the states how to do the job works far better than sending out a bureaucracy to impose solutions known not to work, which is what we do now.

Let Congress come down from its hill and do the job the Constitution imposes. Make the District the city on a hill. Pluck out the beam from Washington's eye, and then we can see clearly how to deal with the states; and when we've done that, the world will be eager to imitate our success.

- 30 -



The Millennium Bug, AKA Y2K: Don't Panic

(Note, October, 2002: You cannot believe the hate mail I got over this column. Irresponsible was the least of the names I was called. I wasn't taking this terrible threat seriously enough...)

Jerry Pournelle

Most of us have seen it coming for years, but suddenly everyone is afraid of the "Y2K" bug. We hear stories of airplanes crashing, missiles launching, water mains bursting, and a general series of disasters that bring down civilization. Some anarchists are looking forward to the Millennium.

There are three parts to this problem. The first one will happen on September 9, 1999: some computer operating systems believe that the number 9999 means "end of file". The result of unexpectedly encountering this number isn't predictable. Most computers won't care, and it doesn't look as if this bug is more than a mild annoyance, but it will be a good test of just which critical systems use really obsolete software.

The second problem is February 29, 2000. Although 2000 is divisible by 100, and thus we might think it should not be a leap year, it is also divisible by 400--and thus IS a leap year under Pope Gregory's revisions to the calendar. This has confused a number of people, including Microsoft: current Microsoft software shows 29 days in February, 2000, but some earlier versions did not. Here and there you still find computer software (and paper calendars, for that matter) that get it just plain wrong. Again this is more annoying than frightening: the worst that happens is that the day/date is off until someone notices it. Most computers will get it right anyway.

Finally, there's the actual Year 2000 or Y2K bug. Briefly stated, the problem is that early computer systems were short of memory, so programmers got in the habit of representing the year by 2 digits, thus tacitly assuming that the other two were "19". The habit endured even after memory became cheap, and even after it became obvious to many that this was a poor thing to do given that the century was ending. Worse, even after programmers broke that bad habit, a lot of the older software was simply brought over to new systems without rewriting, and that software has two digit dates. The result is that on January 1, 2000, some computers are going to think it's 1900.

So what? Well, that depends on the program. Computers take instructions very literally, so if they've been told to send out checks on January 3, 2000, and the machine sees the date as 1900, it may happily wait until forced by an operator to act, and then it may issue checks dated January 3, 1900. Another program may issue airline tickets dated 100 years early. Neither of these events is a tragedy. It just means more work for human clerks and flight agents.

That's true for most of these systems, including banks. After all, we ran the country nicely without computers and redi-tellers for most of my lifetime; if some of them stop working it may be a bit expensive to go hire clerks to do their jobs, but it won't be _that_ costly. What man has done, man may aspire to.

Some experts say the real problems will come from so-called embedded systems: computers installed in quite unexpected places. VCR's. Clock radios. Even my toaster and coffee maker know what the date is. The toaster doesn't care what the year is, but one of the VCR's does, and after January 1, 2000, I may not be able to program it properly: I might even miss taping an episode of Buffy The Vampire Slayer. Once again hardly a tragedy. Elevators are a more serious matter: some elevators are programmed to stop working when there's been too long a time since the last maintenance inspection. What the elevator will do when it believes there has been a negative 100 years since it was inspected isn't predictable, but chances are, it will conclude that it's confused, and do whatever it's programmed to do by default: generally that will be to go to a home floor and stay there. Like banks, though, elevators worked nicely before computers, using a wonderful invention known as the elevator operator. They can again.

We could go through a long list of devices that may cease to work when they discover the date is 00. In just about every case the remedy is the same: get a human operator. After all, most of this stuff was designed by people who know that computers fail, electricity fails: there has to be a way for humans to intervene when the computer isn't working. Employing more people may cost money, but it's not going to end the world. Even air traffic control has human override: this mode requires greater separation of airplanes, and thus many flights will be delayed or cancelled, but the planes won't fall out of the sky or crash into each other. They'll just be held pending greater separation and a fix to the computer problem. The fix, incidentally, may be to fool the computer into thinking that the year really is 1900: that works in a surprising number of instances.

So: if the Y2K problem is likely to produce major annoyances and unexpected expenses, but not disaster, why do some see it as the potential end of civilization?

Well, look at who does. One group are religious theorists, including my friend Gary North, who see the Millennium as a Biblical event. I'm unable to understand that reasoning: why does God care about round numbers? Modern scholarship demonstrates that the conventional Year One is unlikely to have been the year of Christ's birth, so we're talking about 2000 years from an arbitrary date. North has more reasons to believe civilization will collapse in 2000, but they're all more religious than logical.

Then there are the computer professionals, many of whom have formed partnerships and corporations to hire out their services surveying and correcting the Y2K problem in your organization. It would be astonishing if they didn't predict dire, but avoidable, consequences that only they can avert. Finally, there are the professional doomsayers who have been predicting the end of civilization for the past 30 years. We can take them as seriously as we always have.

Of course it is sensible to take some precautions. I certainly intend to fill my cars with gasoline a day or so before New Year's Day, 2000, and I intend to have a few hundred dollars in cash in my wallet. (And there IS a danger: if enough people decide the banks will fail in January, 2000, and decide to get their money out in December, there really could be bank failures: caused not by the Y2K bug, but by panic at the thought of it.)

I won't be booking any airline flights that week, but then I wouldn't have anyway. If I have to schedule surgery, it won't be for the first week of that January. I may even do extra heavy grocery shopping just after Christmas next year.

What I won't do is activate my old survival company or run for the hills. Neither should you.

- 30 -

Jerry Pournelle has written about computers and civilization for twenty years. He is also a former leader of the survivalist movement and former contributing editor for SURVIVE Magazine. He intends to stay home for New Year's Day, 2000



The Millennium Bug, AKA Y2K: What to do

Jerry Pournelle

Roosevelt once said the only thing we have to fear is fear itself.

 That's the best mantra for considering the Year 2000 problem. For those who read my previous commentary on this problem as trivializing it, my apologies: it wasn't what I intended. On the other hand, it's vitally important to understand that many proposed remedies are cures worse than the disease. Widespread selling off of stocks, withdrawing money from savings, and exaggerated hoarding of food and fuel and inventories can transform the Year 2000 bug into a major disaster. Yes, the problem is serious. Yes, it's going to contribute to the economic problems we're already seeing in Asia and elsewhere. Yes, there will be some unexpected shutdowns in computer-controlled industrial processes like oil drilling and shipments and chemical manufacturing; but having said all that, I return to the main point: far too many act as if Y2K will be a disaster like war or an asteroid strike, and that it will not be.

Consider: the Republic has endured hurricanes, drought, the Great Freeze -- remember when power plants weren't working because the piles of fuel coal were frozen and had to be blasted apart with dynamite? -- and worse. These were real events of real destruction. Property was destroyed and people died. Computer failures are hardly in the same league.

In a word, the situation is serious, yes; but it won't be grave unless we make it so.

Now: what should you do?

First, don't panic. Expect inconveniences, some major, and plan around them. Don't schedule critical events for January 2000.

Then go to and read the testimony of Bruce Webster to the House Subcommittee on Government Management. Webster has the most realistic and balanced assessment of the problem I have yet seen.

Then test your computer systems.

That's not difficult, but it is sufficiently inconvenient that many haven't done it. What you must do is schedule a time when it will not be a disaster if your computer systems don't work for a couple of days. (If there is no such time you have a larger problem than Y2K). Then crank the date up to 23:30 of 31 December 1999, on all the computers in your place. Let the date roll over and keep operating at that false date for another 24 hours. See what happens. If something fails, you know what you'll face: plan on taking care of it, either by replacing the system with another, or by hiring human workers to do the job the machine was doing.

Next, have standby equipment. For most businesses, today's desktops and laptops--which do understand the year 2000--are far more powerful than the old iron that may not be aware that years have 4, not 2, digits. You can today buy for under $2000 a network capable machine far better than anything that existed when most of the Y2K problem code was written. Take advantage of that. If you're planning on replacing computer equipment but holding off because of accounting practices, reconsider that decision now: the sooner you have modern equipment installed the better.

Regarding software: Despite the protestations of purists who hate modern programming tools that use the computer design tools known as "wizards", the fact is that it is now fairly easy to create business programs tailored to your needs. Use them. If, like me, you are using older software (I wrote most of mine in CP/M days) because it's a pain to convert, this is a good time to modernize and get it over with. Incidentally, if the programmer you are thinking of hiring seems more interested in programs as art forms than as tools to get the job done, reconsider. You're not trying to advance the state of the computer arts, you're trying to get past a problem. Stay focussed. Try to use off the shelf software rather than custom designed programs.

That's the general run of businesses. Now the embedded processor problem:

Webster and others have concluded that it's between unlikely and impossible that all the processor chips controlling manufacturing processes can be replaced before 2000, so there will be failures. A few failures may be critical: city water supplies overloaded with chlorine, that sort of thing. These are just going to have to be dealt with the same way city authorities deal with any other disaster. Most are critical in the other direction: things stop working. Oil isn't pumped, fertilizer isn't made. No one knows how widespread this will be, and that's what scares people.

Yet, once again, consider: every one of those chips will fail some day. The death rate for both people and computer chips is the same over time, one to a customer. Presumably the world goes on when computer chips fail: any business designed on the premise that their control systems were immortal and would always be failure free is a good candidate for bankruptcy: you wouldn't want the people who designed that to work on anything else. In the vast majority of cases there is a contingency plan. Consider what you would do in the face of a real disaster: earthquake, fire, flood, tornado: something that caused widespread power failures predicted to last a couple of weeks. If that would most likely cause you to run in circles flapping your arms like a chicken, it's not too early to consider alternatives.

Do I discount the Y2K problem? Discount from what? From survivalists who advocate running for the hills and returning to rebuild civilization from the ashes? Yes, of course I do. From those who believe we will have a worldwide Great Depression? Certainly. From those who believe we are headed for a recession? Not really, but then I have never thought we could continue with stocks valued at 50 to 70 times earnings without a strong correction. Y2K will exacerbate the effects of that correction, but anyone reading Intellectual Capital must already be aware that Asia is undergoing economic problems right now. It can't be long before some of that comes home.

After World War II, Germany lay in ruins, caused not by computer failures but by tons of explosives and firebombs. The Marshal Plan headed off widespread starvation, but the recovery planned by the experts wasn't working very well. Then Konrad Adenaeur and the American Proconsul Lucius Clay proclaimed the end of economic regulation. In effect they said "If you have an idea, try it. If you want to hire someone, for any wage, do so." The result of this economic freedom is now known as the German Economic Miracle. Note that it was done without computers, and without government expertise. Most of the world was built without computers. In the words of the old song, "I got along without you before I met you, gonna get alone without you now…" Do not underestimate unfettered human ingenuity.

The organizations least prepared and least likely to be prepared for the Y2K crashes are the bureaucracies. If you believe we can't live without bureaucratic services, if you believe that only government can save us from the Year 2000 Disaster, then you are right to predict doom. Government will not save us. If you believe America was built by free people, the prospect of widespread government impotence may have a different message.

- 30 -

Jerry Pournelle has written about computers and civilization for twenty years. He is also a former leader of the survivalist movement and former contributing editor for SURVIVE Magazine. He intends to stay home for New Year's Day, 2000



September 1998


Phonetaps and Encryption

Jerry Pournelle (1600 words)

The government wants to listen in to anything you say on the telephone or send by email.

From the beginning, the Clinton Administration has tried to get laws that would make it very easy to tap any telephone in the country. They claim they need this capability for law enforcement. A White House science advisor told me "There are some bad people out there, and we have to be able to move fast to stop them."

By move fast they mean fast indeed. The days when you needed to go to someone's office to bug his telephone are long gone: today, the phone company uses computers and can tap into almost any phone from nearly any technical administrator's office. They use the same technique to set you up with a voicemail system. A government agent with a court order can get the phone company to apply a tap in minutes.

However, the White House wants to require the telephone companies to set things so that any government agent can use a desk computer to tap any telephone line in the country simply by typing in the phone number to be tapped and some special code. The privilege would never be abused, they would always get a court order before actually doing that, and I am the Queen of Rumania. So far Congress has refused to make this requirement; the last I heard the Feds were trying to persuade the phone companies to do it voluntarily.

The government doesn't make its case solely on the need to listen to telephone conversations. They also want instant access to any messages sent, whether email, bank transfers, stock market data, or encrypted messages. Particularly encrypted messages.

The government doesn't want you to send messages in codes they can't break, and wants to punish you if you do it. They'll also punish you for writing programs that produce encrypted messages they can't read, and they have got some goofy export control laws to keep you from exporting encryption programs that work. This hasn't kept foreign competitors from getting strong encryption programs. It also endangers American commercial security.

It's a paradox: A government that insists that we have a constitutional right to privacy wants to be sure we have no private communications or even private records. Their case is simple: if you give them probable cause to get a warrant to search your house and business and subpoena your records, they want to be sure they can read all those messages and records and notes and memoranda and account books. You have no right to record things so that the government can't read it. To enforce that, they want to make it physically impossible for you to do it.

Their method is to insist that when you encrypt a message, the government has access to a key that can decode it. In one scheme, this government secret key would be built right into the hardware of your computer: you encrypt your message, using strong encryption, and only you and the recipient can decode it; but the government has a hardware device that takes advantage of a deliberate flaw built into your encryption system, so that the government can decode any message it gets hold of. Since this scheme is pretty vulnerable - a foreign agent or a business competitor need only compromise the government "keeper" to have access to every encrypted message sent by any person or business in the United States - there are usually refinements. One refinement would let you have strong encryption without the built-in government back door, but you'd have to deposit the key to your message with some third party: say a firm of accountants, or a special bonded 'codekeeper'; someone who would deliver that key to an official with the proper warrant.

Until some such scheme is in place the government is determined not to let U.S. citizens have access to strong encryption.

Now what is strong encryption? And for that matter, what is encryption?

Encryption is a means of taking a message and scrambling so that only the intended recipient can read it. It differs from a code. A code is a translation: in one of Peter Ustinov's movies, the diplomatic code used the word "Darling" to mean "You are the stupidest idiot in the diplomatic service". Codes require code books, and so long as the books are secure and messages aren't repeated too often, codes are secure in the sense that no mathematical techniques can be used to break them. Even if you know what some of the words mean, you have no way to learn the others. The problem with codes is that it's pretty hard to talk about a subject you didn't think to put in the code book. They're also slow and hard to use.

Encryptions or ciphers, on the other hand, take any text whatever and translate that into something else. The something else is then read by means of a key. Julius Caesar used ciphers in his reports to the Roman Senate. Edgar Allen Poe's story "The Gold Bug" is not only about a cipher, but describes fairly well the technique used to break the cipher. The important point about ciphers is that you can encrypt any text, but if someone uses powerful mathematics to discover the key, he can read anything that uses that key.

Poe's simple substitution cipher would be broken by the computer on your desk in microseconds. Today's encryptions are much more sophisticated. They generally involve large numbers which are themselves the products of two large prime numbers, and while the details aren't important, the thing to remember is that we're dealing with mathematical techniques, and what one mathematician using a computer can hide, teams of mathematicians using many computers can discover. Every now and then we hear announced a new encryption system that is "proved" to be unbreakable. Inevitably a few years later we hear that it has been broken. Consequently, we no longer talk about "unbreakable" codes. Instead, we speak of weak and strong encryption, with the obvious implication that "strong" encryption produces code that's safe at least for a few years, and possibly longer.

Strong encryption is a moving target. What was strong a few years ago is very weak now, and what's strong today may be vulnerable tomorrow. The important thing is that we have programs to encrypt messages in ways that make them very hard to break. These programs operate quickly - they'd be pretty useless if they didn't - and can be implemented in hardware or software. They're in wide use; but in the US they're at the edge of illegality if not downright illegal.

The real problem, though, is there are a number of legal encryption programs floating around that are said to be strong, but may not be. Now they're certainly strong enough to stand up to casual attacks. You can aim your desktop machine, or your company's mainframe, at one of those messages, and it will take years before you get a result even with luck. Turn the job over to a supercomputer and there may be a different result. More interesting, though, is that scientists have now found ways to break the decryption task into hundreds to thousands of small parts and distribute those parts to hundreds or even thousands of small computers. This can reduce the decryption times from years to hours.

The hundreds of small computers need not be expensive: 300 small Pentium systems with the requisite hardware wouldn't cost $200,000, and could be used to break, say, the bank codes which transfer billions of dollars. If you could siphon off $50 million to a Cayman Islands bank one afternoon it would be a pretty good return on a half million dollar investment in computers. I may write a science fiction story about this: meanwhile, if you hear of an anonymous purchaser of 300 Gateway 2000 systems with networking cards, you might worry.

And that brings me to the point: modern commerce needs strong encryption, and many of our overseas competitors have it; while the United States government hasn't begun to make it easy for American companies to have strong encryption because the government is determined that you shall not be able to send messages the government can't read. They don't want you to buy any encryption device without some kind of government back door, while few companies trust the government with (or to protect) that back door and thus don't want to buy government approved devices. And there we stand.

The government's argument is that they need these powers to protect against organized crime, drug dealers, and child pornographers. Once at a meeting of the American Association for the Advancement of Science I found after arguing for strong encryption I was accused of aiding criminals and child abusers. "There are whole networks of child abusers out there. They build victim data bases for use by visiting perverts. Do you want to protect those people?" a White House spokesman asked me. She was quite serious.

The answer, of course, is that the cure is worse than the disease. I don't want to protect those people, but I don't know how to take away their right to privacy without taking away everyone else's. Moreover, since the strong encryption programs exist, the result isn't that the child abusers don't have strong encryption - they do - but that legitimate business don't. And every time I go to DC to testify on this issue, I wear my button that says

"When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl."

- 30 -  Top


October 1998

Jerry Pournelle

The news today was quite weird. First, we heard that English police have arrested a Chilean former chief of state, on a warrant from a Spanish judge over crimes supposedly committed in Chile while Pinochet was President of that country. Second, the Disney Corporation has persuaded Congress to extend the life of copyright to over a hundred years. Third, the Microsoft Monopoly Trial begins today.

My wife first noticed that these all have common elements: they are attempts to change the rules. As she explains, when you are the last of eleven children, you become very conscious that rules are important. When you are at the bottom of the pecking order, you very much want to be sure of the conditions under which you can be pecked.

On of the greatest moments in the history of human freedom was when the Plebians required the Roman Patricians to write down the Twelve Tables of the Law and place them were all could see them. The Plebians were willing to obey the Law: but they wanted to know what the Law was.

In the three cases above one would be hard pressed to find rules that can be written down and published. Without rules there is no rule of law; and recent experience shows that the rule of law is probably the most important element of prosperity. There is much at stake here.

But, some will cry, these are cases that cry out for justice, and if the law gets in the way of justice, then so much the worse for the law. Pinochet deserves to be tried for his crimes. Copyrights ought to be extended for long periods. Microsoft is evil and ought to be punished, broken up, nationalized if possible. This is what justice demands, and let justice be done though the heavens fall!

But what is justice? We have been contending over that since men began rational disputation. Plato thought to end the question when he wrote The Republic. Isaac Asimov rather more recently had his robot policeman say that "justice is the condition that prevails when all the laws are enforced;" which while not as well known as Plato's inquiry, at least has the merit of being comprehensible, as well as moving us closer to the Rule of Law.

While some may contend that Justice demands Pinochet's trial and execution for overthrowing the Allende regime before it could be converted into a full copy of Castro's Cuba, there are plenty of other Chileans who every night thank God for Pinochet's coup. There are more who were satisfied that the Civil War was ended and Chileans no longer breaking things and killing people in the name of ideology. Whatever the demands of justice, the British Government has helped Spain reopen old wounds, and we will not soon hear the last of this.

The copyright case is different in that it's not likely to cause a civil war, but it's a violation of principle all the same. The Constitution is quite explicitly against monopolies, and for good historical reasons: the English Crown was accustomed to granting monopolies to its friends. One might have a monopoly over importing Sherry, another a monopoly in matches, or salt, or trading with a particular country; and all of them would be enforced by Crown officers. The Framers were determined that this wouldn't happen in the new republic. On the other hand, they could see a need for protection of some rights in intellectual properties.

Article One, Section 8 of the Constitution says: Congress shall have Power: To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." Other grants of power in that section include the power to declare war, establish courts inferior to the Supreme Court, raise and support Armies, provide and maintain a Navy; which is to say they didn't consider this a trivial grant. Nor should we. But unlike the other grants of power, this one states its purpose, as if the Framers were especially concerned about it; and they hedged it by including the words "limited Times" as a restriction on what could be granted. Whatever was meant by "limited Times," it was certain that monopolies, even those earned by Authors and Inventors, could not be perpetual.

Understand, as an Author, I am certainly not against copyright enforcement, and assuming that any of my properties have any lasting worth, I suppose my children may have reason to be thankful for the new copyright extension; but this is one more assault on the Rule of Law. If life plus 100 years is not perpetual, it is close enough to it. I doubt you could find one of the original Framers who thought he was granting the power to extend a monopoly for a hundred years and more. Why should he? Authors will not write more vigorously for 100 years of protection than they did under the old rule of 26 years renewable for another 26. This new extension of copyright isn't going to "promote the Progress of Science and useful Arts," it's to benefit a corporation, and more specifically to benefit the executives of that corporation, no one of whom to my knowledge is an Author or Inventor. Politics as usual, and one more blow at the Rule of Law.

And finally we have the Microsoft case. But Microsoft is a monopoly! Surely this is in accord with the Rule of Law? Hardly. First, whether or not Microsoft is a monopoly, it certainly is not a legal monopoly created by law and enforced by public agents. Indeed, it's hard to see how it's a monopoly at all. Macintosh users are quick to point out that there's nothing you can do with a Microsoft Windows machine that you can't do as well with a Mac; even if that's not strictly so, it is certainly not Microsoft's fault. One of the most useful programs you can run on a Macintosh is Office 98, which has features not available in the Windows version; while some of the most useful Windows software, Excel and Power Point, were for years available only on the Macintosh. Indeed, I recall noticing when Windows was launched that all the briefing charts to promote Windows were generated in Macintosh Power Point and printed on an Apple LaserWriter.

Most computer users do choose Microsoft Windows, but they don't have to, even in the desktop market. When you get to larger systems and big servers, Microsoft hasn't got anything like a monopoly, and what dominance it has in net servers is being challenged by Linux - which is free, source code and all, and daily gains adherents. Microsoft is said to be driving Netscape out of business by restricting their access to the market: an odd statement for a company that recently boasted of shipping 100 million copies of its product.

Another criticism of Microsoft is that hardware is getting dramatically cheaper, but software, specifically the operating system, is not. This too is an astonishing statement: today I can buy both Windows 98 and Office 98 for under $500. Not that many years ago, in order to get anything like the functionality I have with just those two products I needed: DOS, whether IBM or Microsoft; Lantastic or some other networking software package; Quarterdeck DesqView; Data Storm ProComm Plus; Lotus 1-2-3 or Borland Quattro; Borland Sidekick; Word Perfect; and I still didn't have a drawing program or a net browser, or for that matter Solitaire and FreeCell, even though the total value of all those programs was well over $1,000.

I may get angry when a Microsoft program crashes, but then I remember the problems I had getting all that stuff to run. Windows does more, and is more reliable.

The truth is that Microsoft competes hard to ship programs, while other companies compete hard to hire better lawyers. For a look at how Microsoft really operates, see
, one of the best researched articles on corporate culture I have ever read. Microsoft runs scared, because Gates understands something others seem to have trouble grasping: there's only so much that marketing can do to cover product inferiorities. In today's fast moving electronics markets, giants appear and vanish regularly. It truly is a red queen's race.

Microsoft plays rough, and losers always want to change the rules. After all, doesn't fair play mean that someone else gets to win sometimes? Don't we want fairness and justice, not rules?

Adam Smith said that the worst enemy of capitalism is a successful capitalist, who will sabotage the market by buying political favors and thus change the rules. That is as true today as it ever was. The Long Boom is fueled by expectations of continued growth: nothing else can justify the dizzying heights price to earnings ratios have reached.

No one expects growth where there is no Rule of Law. We forget that at our peril.

- 30 -   Top


Unrecoverable Attitudes: Foreign Policy

November 1998

Jerry Pournelle

Walter Lippman compared national policy commitments to drafts against a bank account: eventually you must have funds in the bank to cover those drafts. You can borrow - that is make threats - for a while, but eventually you need the military power to make good. We're seeing this with Saddam Hussein and Iraq. For the six years of the Clinton Administration we have tried to make Saddam allow our weapons inspectors to do their work, while simultaneously reducing the military bank account toward the dangerous levels of the Hollow Army. We have plenty of Admirals and Generals, but the number of troops and sailors is falling. Tours of duty aboard ship get longer and longer due to a dwindling number of surface warfare officers. All services are short of pilots. In six years our splendid military establishment has been cut back. This would make sense if we also cut back our foreign policy commitments, but the fact is we've extended them. We have forces in Haiti and Bosnia and now Kosovo, we're supposed to be monitoring the peace between Israel and the Palestinian Authority, and now we're threatening Iraq.

And so what, some would say. We jawboned Kosovo, and maybe the result wasn't perfect, but we got through that crisis. For the moment. As for Saddam Hussein, surely we have the means to bomb Iraq back to the Stone Age. We'll just use Tomahawk missiles and blow things up. It's expensive, but no lives will be risked. At least not American lives. And our weapons are so accurate that we'll do it all with little damage, and our information warfare people will then go in and finish things, and pretty soon Saddam will go away.

There may be one or two flaws in that scenario. First, the number of Scud missile launch sites we detected and destroyed during the Gulf War was precisely zero, and that was with active hunting by manned aircraft. Scuds aren't very accurate, but they're good enough to hit Israel. Now presume that Saddam has some weapons of mass destruction. How many Israeli lives do we care to risk? We sure can't protect Tel Aviv. Second, if information warfare were as effective as its advocates say, why is Saddam Hussein in charge after all these years? And perhaps more to the point, why do we need inspectors to find Iraq's weapons factories? Why don't we know where they are now? Will we know better after we break things and kill people?

The last time we used Tomahawk missiles we killed some civilians and blew up a pharmaceutical factory in Sudan. The strike itself wasn't flawless. One of the birds hit a candy factory almost a mile off target, an unfortunate accident, and particularly unfortunate for the families of those we killed. Now it appears that the 'evidence' we had for the target plant being used to make nerve gas was flawed, and we were suckered by bin Laden's people using cell phones. More dead Sudanese, alas. The strike against the Afghani camps was no more successful. If it ever was a secret that our satellites can listen to cell phone conversations, it hasn't been for years; certainly every terrorist organization not only knows but counts on it as a means to feed us false information, and it worked just fine this time.

We hanged Japanese officers for planning the Pearl Harbor strike. How was this different? Will a Sudanese judge issue warrants for the arrest of our Fleet officers? The Secretary of Defense? Clinton, when he becomes a former President? What will it cost us to make those warrants a joke?

And joke or not, our information warfare experts were suckered. The President of the United States ordered the bombardment of Sudan, a country with whom we are not at war, and Afghanistan, whose Taliban rulers had until then no real reason to dislike us, and whose record of cooperation with anti-US terrorists was at worst unclear. Sudan hasn't fallen, and the Taliban is stronger and now hates us. So much for information warfare and remote bombardment as a means of changing governments.

T. R. Fehrenbach's THIS KIND OF WAR makes the point that you can fly over a country. You can bomb it into the Stone Age. You can drop nuclear weapons on it and slaughter all the inhabitants: but you don't own it until you can stand a 17 year old kid with a rifle on top of it. If we really want Saddam Hussein out, there's only one sure way to do that, and while our electronic surveillance systems can help, it's going to cost blood and treasure to get him.

And in fact we're about to go through meteor showers that can't make our satellite observation systems better, and will probably make them worse. This won't be the last meteor shower, and we don't have the launch capacity to replace the satellites that will be damaged.

While we're on that subject, we know about the coming meteor showers, which we can't do anything about; suppose we knew about a Dinosaur Killer? It is now estimated that your chances of being killed by a meteor are about the same as your chance of being killed in an airplane accident. That's not a very high number, but it's not zero, either. It's as high as it is because while the number of individuals who have been harmed by meteors is nearly zero, there's also a small but quite real chance that we'll all be killed by a big rock coming at 30,000 feet per second.

 When I did the research for Lucifer's Hammer, the first of the best selling books about big rocks hitting the earth, the chance of a big meteor smashing us was estimated at about half what it is now. After the Cold War ended, the Air Force declassified some of the data gathered by our satellite detectors. It turns out there are about twice as many near-Earth large rocks as we thought there were in 1970. Of course if we knew that one of those rocks would hit us in two years, there wouldn't be anything we could do about it. We can't even get a space station in orbit.

Once years ago when I was in aerospace I was assigned to a crash investigation team. A warplane had augured in under odd circumstances, clear skies, no turbulence. There was just time for some of the crew to eject. The pilot stayed aboard to hold the ship steady while the others got out, and rode her in. Eventually we found that the trim tab controls had been connected backwards, and by the time the pilot discovered that, the plane was in an unrecoverable attitude: that is, while nothing had happened to the ship yet, there was no way to get it back to straight and level. It was going in and there was nothing to be done.

Suppose we saw an asteroid coming, as happened in this summer's popular movies? Without better capability to get into and work in orbit, there would be nothing to do, just as there isn't much we can do about the coming erosion of our existing satellite assets. It will happen, and we'll try to keep secret just how badly we were hit. With luck your cell phones will still work by the time you read this. The coming asteroid is a low probability way of getting into an unrecoverable attitude, but there are others more probable. We can continue to make commitments we can't keep. Much more of that and we'll be faced with wars on several fronts, and nothing like the capacity to fight all of them; possibly not any of them.

Then there's our education system: while our graduate schools are splendid, there's no supply of educated young to feed into it; and no one is likely to go broke selling short the American primary education system. These are real problems, likely to get worse, and if anyone in authority is concerned about them to do more than apply deep layers of rhetoric, I haven't noticed. It's much easier to ignore them. The United States is not yet in an unrecoverable attitude, but the concept applies to countries as well as to aircraft.

- 30 -  top


Operation Impeachment Storm

December 1998

Jerry Pournelle

What my Pentagon friends called Operation Impeachment Storm (a few preferred Operation Free Willie) ended a few hours after the House voted a Bill of Impeachment, leaving many to wonder what firing off all that ordinance was in aid of. Leave cynicism about the "true purpose" aside and take it at face value: what did Operation Desert Fox accomplish?

The official estimate is that we have set Iraq's missile construction capability back by a year, destroyed some of the command and control infrastructure, and damaged several chemical plants which may or may not have been involved in manufacture of chemical weapons. We damaged or destroyed an oil factory, and we killed or disabled something like 100 military people and about that many civilians. To the extent that we destroyed military facilities that Saddam Hussein considers necessary, we have increased the misery of the Iraqi people: that is, retaining control of the country and posing a strong threat to the region's peace takes first priority with Saddam, so money that might have found its way into the civilian economy won't get there until the military is satisfied.

We have also ensured Saddam's rule until the day he is dead. Far from encouraging an uprising, we have regenerated a spirit of patriotism and loyalty among the Iraqi people, with many of them rallying to the support of a leader they may not like, but who stood up to the Great Satan. Of course there never was much of a chance of an uprising, because our last blunders in that region encouraged the boldest of the resistors to surface and be killed. It would take a foolish Iraqi indeed to believe the Americans will actually take risks to help those they have encouraged to rise up in rebellion. We'll send them fair words, but no buttered parsnips.

We've not only shored up Saddam at home, but we've made him into something of a regional hero. Whatever the ruling elites think, the Arab in the streets of Jordan, Arabia, Palestine, Syria, or Iran sees Saddam as a genuine hero, an Arab who defied the New World Order. We have pretty well busted the economic blockade of Iraq. That always was a rather leaky deal, but within a year Iran and Syria at least, and possibly Jordan as well, will more or less openly trade with Iraq, with or without the disguise of "humanitarian" aid.

We've ended UNSCOM inspections. Given that seven years of ground inspections complete with cameras monitoring many of Iraq's factories did not eliminate his weapons capabilities, it's not likely that satellite inspections and covert operations will be more effective in finding out what kind of weapons Saddam is building, and where he's building them. We've spent about a billion dollars, and used up a good part of the inventory of Tomahawk missiles.

One hopes they'll be replaced. We may need them for a rational purpose some day. Precision weapons are valuable when used as part of a strategic plan. Accurate artillery always is. But North Viet Nam didn't fall to bombing, and neither did Iraq. No surprises there. We didn't overthrow Ho Chi Minh, nor could we without invading North Viet Nam. We didn't overthrow Saddam Hussein, nor were we likely to; the only chance of that would be a lucky hit that killed him. If we want Saddam out, we'll have to send in the Marines. We won't do that.

So: was it all worth while? We didn't weaken Saddam, but we did show some resolve. We showed the world we are willing to shed blood. Not our own - there was nothing sillier than the Congressional speeches about our fighting men and women being at risk - but blood none the less. We conducted a live fire exercise, a weapons test against real targets willing to fight back.

 Surely that is valuable? Perhaps. There were a few surprises. In the Baghdad area at least the Iraqi's were able to shoot down more cruise missiles than we had officially expected. I say officially, because some of us always did have doubts about just how many of those expensive birds will get through. They aren't stealthy, nor are they supersonic. The current generation has to fly to get bearings from landmarks and check points, and an intelligent enemy will put flak batteries at those points. He will also look at the paths the missiles can take to reach targets important to him, and line those paths with more guns.

I concluded from the flak patterns during the live broadcasts from Baghdad that they'd done precisely that; this wasn't the random firing of all the guns in the city that we saw back in Desert Storm. The Iraqis learned something from the last desert war. It would be astonishing if they hadn't: I would imagine that the top military priority of every third world dictatorship on Earth is to find a way to deal with American cruise missiles. It takes no genius to realize that American Presidents aren't likely to risk American lives. Cruise missile let us expend treasure without blood. Our blood, anyway.

And that brings us to the real tragedy of Operation Desert Fox: it's one more wound to the constitution. In 1787 the King of England had the undoubted prerogative of making war on whomever he pleased, without let or consent of Parliament. He couldn't pay for a war without Parliament; but he could start one, and often did as a way to distract the public attention from domestic problems. To this day the right of making war remains with the Crown, which is to say, the Cabinet.

The Framers of the Philadelphia Constitution deliberately took that power away from the Executive. Only the Congress can declare war. Of course we haven't had a declared war since World War II, but up to now we have always had some kind of Congressional approval: the Tonkin Gulf Resolution, a joint resolution of support for Desert Storm: something. Not this time. On the day Monica Lewinsky was to testify to the grand jury we bombarded Afghanistan and Sudan. On the day Congress was scheduled to debate articles of impeachment, we bombarded Iraq. In neither case was there any consent of Congress.

Now the timing of those events may have been mere coincidences, lucky breaks for the President; assume that's true; it still cannot escape some potential Caesar that he can, with little political risk, bombard any nation on Earth as a way to distract the citizens from domestic problems. It costs money, but it's money already spent; no new appropriation is required until it comes time to replace the expended missiles. When Presidents had to risk American blood to make war they had to be more cautious. No longer. The real effect of Desert Fox, then, was to strengthen executive power in both Iraq and the United States. That may not have been intended, but it's what we did.

- 30 -


Part One of these reports.

Part Three