Bogonomicon

Looks like the economy only added 1,000 jobs in December. Oh, and the previous month got downgraded by 66,000 jobs while they were at it. Which means in a couple of months, when the inevitable downgrading of December's job market happens, how many jobs do you think we lost? Yeah, the economy grew at an (adjusted once already) rate of 8.6%- but this was all taken up by the 9.6% jump in productivity, and thus didn't create any jobs. The only reason the jobless rate went down was yet more people leaving the work force- now, one person in three who could be in the work force isn't even looking. This is the lowest it's been since 1991.

There is one bright spot to all of this. It's the economy, stupid.

It looks like Nader, although he won't be running on the Green ticket, thinking of running anyways. Presumably as generic third party. A fellow Dean supporter I met this afternoon told me she was "close to tears" when she heard this. Personally, I could care less. If Dean gets the nomination- as is the most likely scenario at this point- there will not be a signifigant challenge from the left, Nader or no.

I was a blue-green voter who strongly considered voting Nader in 2K. The reason I did was that the Democratic party was loudly moving to the right. They were telling us liberals, to our face, that a) we didn't matter, b) they wouldn't represent us at all if elected, and c) we were obligated to vote for Democrats no matter how bad because the Republicans were marginally worse. We were going to vote for Republican-lite and like it. And these campaign promises they actually abided by- the war in Iraq, Bush's tax cuts (both of them), PATRIOT I and II, on and on- every dumb idea that's come out of the Bush whitehouse has found strong and ready support among the Democratic party. This is causing justifiable anger among the liberal base of the party- if we wanted to vote Republican, we'd vote Republican for Pete's sake. The effect of this anger was Nader support in 2000. And the effect was even more pronounced in 2002.

But the situation is different in 2004 than it was in 2000. Dean, by moving the party back towards the left, is re-engaging the liberal/progressive base and bringing them back into the fold of the party. He's not a whole-loaf candidate- his position on the death penalty and gun control are at variance with mine among others. Almost all Dean supporters I know of always say "I don't agree with him on everything, but..." But his positions on gay marriage, health care, the war in Iraq, and other issues makes him much more than a half a loaf candidate for the liberal/progressives- three quarters a loaf at least, if not nine tenths a loaf.

There are a certain number of people who will vote Green no matter who the Democrats run- Liberman, Dean, or Kucinich, it doesn't matter. And I suppose there are some number of people who actually voted for Nader who will vote for Nader again, no matter what party he is running as. But most of the people who voted for Nader in 2000 were really just voting against Gore/Bush. With a Dean candidacy, those people will (by and large) be voting for Dean. Not to mention the surprising number of Perot/McCain voters Dean has been pulling in, who were the other base of voters Nader pulled from (I've seen reports that 30% of Nader's supporters came from the Bush camp). A Dean candidacy will prevent any mass exodus of liberals and progressives out the left hand side of the party. A Nader/Green ticket might get half a million votes. A Nader solo ticket will be struggling to get half of that.

In either case, I don't see Nader being an important player in the 2004 elections. He runs, he doesn't run- either way, he won't make much of a difference.

This post over on Berry's World is worth the read. I think it's satire. The trick with satire is that it has to be over the top enough to be obviously comedy, which is a trick lately.

Well, it looks like the media finally caught Dean saying something stupid. He's on the record four years ago not liking the caucuses because of their length driving people away. Which in some sense is understandable if you don't understand the caucus system, which keeps it from being boneheaded, but it's still stupid.

This stands in stark contrast to most of the time when the media plays "gotcha" games, where if you actually go back and listen to what Dean said, he's in fact saying very intelligent things. When Dean said that we weren't any safer after capturing Saddam, everyone was scandalized. 37 dead US soldiers and several weeks at Orange terror alert later, Dean was right. When Dean said that he should be the candidate for southerners who drive pickup trucks with confederate flags on them because they need health insurance and good schools too, he was right. People just left off the last part when they got on their high horses. And so on. Commenting that the emperor has no cloths has always been scandalous.

Actually, the entire Dean campaign can be viewed as his attempt to address the question "why should I go out on a cold and snowy january day to argue politics with my neighbors?" Is there more to political involvement that just spending fifteen minutes on the way home to cast a vote? In this sense, the Canada tape is simply a statement of the problem (although it's not phrased that way) without a solution. The solution is to realize that leaving your home on a cold and snowy day to talk to your neighbors is the reward. The we humans like getting together with others of our kind and talking/arguing. This is Dean before he read "Bowling Alone".

But long term I don't think this will matter. It's not like someone is going to write a book of Dean's boneheaded gaffes or something. Or if they did, it'd be a real short book.

And Iowans- especially the ones likely to leave their homes on a cold and snowy january day to go caucus- are serious enough about their votes that they are unlikely to be swayed by one gaffe. Now, Harkin's endorsement- that is big news, and will sway people. And I could actually see it working (subtly) in Dean's favor. Is this the worst thing they could dig up? This is praising with faint damnation (the opposite of damning with faint praise).

So it looks like AMD has some hardware "fix" for bounds checking. Sigh. At a certain level it's like what the hey- it's not like you are messing up some clean, neat architecture with unnecessary cruft. The x86 architecture is almost nothing but messy cruft. And it already has bounds checking support- look up the "bound" instruction in your handy x86 instruction manual.

The problem is that this is so totally unecessary. Even without hardware acceleration bounds checking is almost free with an optimizing compiler. It's called "strength reduction" and it's already widely done.

Here's how it works. Let's say you have a loop like:

    for (int i = 0; i < n; ++i) {

        a[i] = 0;

    }

A naive implementation without strength reduction would implement that code like you had written:

    for (int i = 0; i < n; ++i) {

        if (i < a.length) {

            a[i] = 0;

        } else {

            throw IndexOutOfBoundsException();

        }

    }

This requires you to do a branch every single loop iteration- slowing you code down, right? Except the compiler if it's at all smart can reorder the code like the following:

    for (int i = 0, limit = min(n, a.length); i < limit; ++i) {

        a[i] = 0;

    }

    if (n > a.length) {

        throw IndexOutOfBoundsException();

    }

Now, instead of having O(N) branchs, you only have 2 no matter how large the array is- one for the min and one to see if you throw the exception. Bounds checking has only added a very small constant factor to the runtime of your program- maybe a nanosecond or two.

Buffer overflow exploits are a direct result of the fact that we programmers use stupid languages. We're still using C to write applications in. And, unless you religously use templates, C++ has the exact same problems. We keep making the same mistakes over and over again. In 1988- so over 15 years ago- the Morris Worm exploited a program that used the C library routine gets(), which was the first big news exploit of a buffer overflow attack. The problem with gets() is that it doesn't bounds check its arguments. This behavior was considered bad enough that the GCC compiler was modified to produce a warning if you used the function gets() at all. But then C++ came along, and just a few years later wrote the iostreams library. And what do you know- it includes the function istream::operator << (char *). It's our old friend gets(), still not bounds checking it's values, still vulnerable to buffer overflow attacks, but now dressed up in fancy new operator overloaded cloths. Demonstratably we seem to be incapable of learning lessons from even a few years ago.

So they'll try to fix it in hardware. My prediction is that it won't work. The problem isn't hardware, it's between the keyboard and the chair. We can solve this problem any time we want- all we have to do is stop using broken languages.

Personally, I think it's exciting that we live in a time where we are finally starting to put some numbers into the Drake equation. But I'm not optimistic about our ability to talk to alien civilizations anytime soon- not because I don't think they're out there, but because I think it just takes a while after discovering radio before communication is established. This article is just a collection of my thoughts on the subject.

First off, people don't generally grasp just how frigging huge the galaxy is. Every if you put exceedingly optimistic numbers into the Drake equation and have millions of technological civilizations in the Milky Way galaxy alone, there are over 400 billion stars in the galaxy. This means the next nearest technological civilization will be one to two hundred light years away. We didn't start broadcasting signifigantly in radio bands that escape the atmosphere (radio doesn't) until the 1940's- we haven't been broadcasting all that long. Radio and TV signals travel at the speed of light- so if they're say 200 light years away, they won't hear us until 2140. If they immediately turn around and dash off a "We hear you! Let's talk!" signal back to us, we won't receive it here until 2340AD. More conservative numbers plugged into the Drake equation say that the nearest technological civilization would be tens of thousands of light years away- and the turn around time would be longer than the amount of time humans have engaged in agriculture.

Won't we be able to hear them first? After all, they've likely been around a lot longer than we have. Some dinosaurs, for example the Deinonychus, were well on their way to being intelligent, tool using creatures. If that pesky asteroid had just held off a few million years, the evolved descendents of the Deinonychus might have built space ships, gone out, and stopped the collision from ever happening. That would give Dinosaur civilization an easy 60 million year technological leg up over man. The Andromeda galaxy is less than 3 million light years away, so 60 million years is more than enough time for radio signals to propogate even from distant galaxies. Wouldn't we be hearing their radio signals?

This is assuming that they are broadcasting in a manner we could differentiate as random from even a few light years away. First off, the number of bits per time period that can be transmitted over any given frequency range is finite, and set by Claud Shannon's information theory. Every square meter of ground has only so many bits of information per time period that you can shove through the radio spectrum. And (I assume) the demand for information communication is infinite, especially over the tens of millions of years time frame we're talking about here. Big powerfull transmitters will be replaced by small, weak, local transmitters. Especially since land line communication is infinite, or much nearer. You can always lay another fiber optic cable if you need more bandwidth. So radio communication will more and more be local only, long haul heavy transfers will be done in cables, undectable from even two inches away, let alone two hundred light years. We're seeing this already with human civilization- big TV and radio stations blasting their signals out over hundreds of miles are being replaced by cellular telephones sending their signals out over thousands of feet, or wireless network hubs and cordless phones whispering their signals out over dozens of feet. A lack of strong, isolated signals and the cacophony of millions of weak signals interferring with each other would make it iffy to detect the planet earth as radiating intelligent signals from even a few light years away.

The second thing the limit on bits/sec that can be broadcast in radio spectrum implies is the increasing use of compression. Compression is the act of removing the redundancy from the communication- in effect, making the signal appear more random. Also, broadcast communication is public and hard to really control who receives it. Which is sort of the point from our perspective- we're not the intended recipient of the radio signal, but we'd like to hear it. But for me, while I don't mind aliens two hundred light years away hearing my cordless cell phone communications, I do have a problem with my neighbor evesdropping in. Or my goverment. If the alien race has the same or similiar desire for privacy, the use of encrypted communication will become standard. Already we're starting to get encrypted handsets, and while the US goverment isn't happy about this, they're trying to sweep back the tide. Sometime in the next few tens of millions of years, encrypted cordless phones will become the standard. So in addition to being low-powered, multiply reused frequency, and compressed information, alien radio communications will be encrypted as well.

And not this wussy-assed weak AES stuff either. We've only been doing this math thing seriously for a few hundred years, so it's not a surprise that a medium-hard problem like proving P != NP hasn't been solved yet. I'd expect the alien crypto algorithms to all be proven NP-hard, even if P != NP is undecidable. The only way that encryption wouldn't be standard is if the aliens were sitting on a proof that P = NP, and that encryption is a pointless exercise only technological backwards people do. But encrypting the communications makes them look even more random. In fact, the stupid, ignorant, backwards human cryptographers already beleive that if you can easily differentiate an encrypted stream from a random bit stream, the encryption algorithm is broken. I wouldn't be surprised if the aliens were sitting on a mathematical proof that their crypto system couldn't be differentiated from a random sequence in the expected lifetime of the universe. We're almost capable of doing that now.

What all this means is that I doubt our ability to detect an advanced technological civilization from their radio transmissions alone, even one as close as the moon (no, I don't think there is intelligent civilizations on the moon), let alone 200 light years away.

Another, more intriguing, conjecture presents itself here. Assume, just for a moment, that faster than light communication is possible. Yes, I know the physicists all say it's impossible. Or at least many do. But assume it is possible. In that case, radio communications would be a terrible way to communicate interstellar distances. Why send a message by radio to the next civilization over, when that takes hundreds of years, when FTL allows you to send a message there and back in a small fraction of the time, possibly even instantaneously?

In this case, the technological threshold for joining the intergalatic pen pal club would not be the development of radio, but instead the development of FTL communication. A threshold we haven't crossed yet. Which explains why we haven't heard anything, we aren't listening on the right channel. It's rather like the American Indian, dutifully looking for smoke signals of the horizon, not being able to detect the radio signals washing over him.

So why aren't they here? Even assuming a linear growth of population and sixty million years there should have been enough population growth that they would have gotten here if for no other reason than they have no where else to go. With even the slowest exponential growth curve and sixty million years they should be all over the galaxy. That American Indian in the previous paragraph didn't have to wait to invent radio to communicate with the europeans- the europeans came to him.

The problem is that I don't think humans will grow forever- the UN thinks human population will limit at less than 10 billion later this century, see this USA Today story. The main reason for this is the increasing wealth of society, not some Malthusian catastrophy (remind me to do an article someday about why Malthus was wrong). In poor socicities, children are both social security for your old age, and economic benefit right now (children can do chores). Add in a healthy dose of child mortality, and the impeteous for having lots of kids is there. In a rich, technological society, children are 100% capital drains. Yes, there are advantages to having kids, but no advantages to having more kids.

A personal example helps make this clear. My father was born on a farm. He was doing chores- feeding chickens and the like- and increasing his families wealth by age six. Not long after puberty he was working with his father and contributing almost as much to the family wealth. So there was an economic incentive for my grandparents to have lots of children. So my dad grows up, get's a Ph.D. in mathematics, and moves off the farm to be a computer programmer, and I come along. I wasn't really capable of working with my father in a productive way until after 12 years of grade school and 4+ years of college (even ignoring the detours I took). Call it 22 years before I could start contributing in a meaningfull way to my parent's wealth. By which time, naturally, I was on my own and still didn't contribute diddly squat to my parent's wealth. There were other factors involved, but you see the economic incentive towards or against having more than a minimum number of children changing. Statistically, the changing economic incentives change the population growth rates towards ZPG.

So assume there is never more than about 10 billion human being ever. With a finite limit on the human population, the number of stars we need to colonize drops radically. Once we start getting about a thousand or so star systems colonized, that means we only have 10 million people or so on average per star system. And a star system is a pretty big place for so few people. We only think 10 billion people is a lot because they all insist on living on the same planet. If we spread them out just among the planets of this solar system, that'd only be a billion or so people a planet- or the population of the world circa 1804. You know, back when the flag only had 17 stars. ~10 million people spread among ~10 planets is only a million people per planet. You're not even getting decent cities at that point. For humans, I doubt we'll ever get much beyond a thousand or so star system colonized if our population says around 10 billion. We humans simply don't like to be that spread out. We do like to clump if we can. There are more than a thousand stars within a 100 light year radius of earth.

So assuming the aliens have similiar limitations on their growth (fairly safe assumption) and a similiar tendancy towards clumping (a not unreasonable assumption- civilization is marked by it's cooperation and interaction, which requires clumping), our hypothetical alien civilization centered a mere 200 light years away has colonized out maybe a hundred light years or so in all directions- barely half way here. Possibly less- if the civilization s signifigantly more clumpy than humans are, they might have colonized only a few hundred systems, only a few dozen light years from their home star.

So why haven't they at least done a fly through? Sent a probe around for a look-see. Maybe even sent an exploration team for a meet and greet, and maybe do a little sexual experimentation on the side? First off, timing is everything. For the vast majority of the last sixty million years there hasn't been much of interest on the earth- especially if the panspermia theory is correct, and terrestrial life was seeded from outer space. This would make life common- both in the probabilistic sense (lots of planets with life on them) and in the sense of being basically the same everywhere. "Woo hoo," the aliens might say, "yet another planet with DNA-based multicellular life. Let us know when they develop sentience." Remember that humans have only existed for the last million years or so- only one chance in 60 of a probe sent any time in the last 60 million years to see humans. A probe sent through our system two million years ago might have noticed the primates as promising- if the aliens said "make a note to check up on these guys again in, shall we say three million years?", we'd still be a million years shy of their return. And technolocial humans are only about 10,000 years old or so- one chance in 6,000. It is quite likely, in fact, that they just called while weren't home.

This is why I'm not optimistic about SETI's chances anytime soon. But maybe we'll get lucky. And maybe I'm wrong.

We got around to the subject of war again and I said that, contrary to his attitude, I did not think that the common people are very thankful for leaders who bring them war and destruction.

"Why, of course, the people don't want war," Goering shrugged. "Why would some poor slob on a farm want to risk his life in a war when the best that he can get out of it is to come back to his farm in one piece. Naturally, the common people don't want war; neither in Russia nor in England nor in America, nor for that matter in Germany. That is understood. But, after all, it is the leaders of the country who determine the policy and it is always a simple matter to drag the people along, whether it is a democracy or a fascist dictatorship or a Parliament or a Communist dictatorship."

"There is one difference," I pointed out. "In a democracy the people have some say in the matter through their elected representatives, and in the United States only Congress can declare wars."

"Oh, that is all well and good, but, voice or no voice, the people can always be brought to the bidding of the leaders. That is easy. All you have to do is tell them they are being attacked and denounce the pacifists for lack of patriotism and exposing the country to danger. It works the same way in any country."

Goering had a point. Human beings, in general, simply aren't that ambitous most of the time. We much prefer to deal with issues closer to home- stick to our knitting and live and let live. It's not so much peacefull or lazy, as a matter of priorities- the problems in our face are always more pressing than the problems far distance. You have to goad people, of every race and era, into marching to war. And the easiest way to do that is to scare them- take the problem far away and shove it into their face. Say "if we don't go there and deal with them, they will come here and deal with us."

That requires, however, that you stay on message. And reality has a bad habit of not playing along with you.

That Goering quote lept to mind when reading this sotry. It seems the FBI got a lucky break- a letter delivered to the wrong address allowed them to break open a major terrorist plot. The terrorists were caught with the components to a bomb that would kill anyone in a 30,000 square foot facility, as well as 65 pipebombs and briefcase bombs, 500,000 rounds of ammunition, and even more disturbingly evidence that there is a nation-wide conspiracy of terrorists.

So why haven't you heard about this yet? Because the terrorists were white christians, not brown muslims. In this day of orange terror alerts, where the FBI is looking for people carrying the Farmer's Almanac, that's the only explanation I can think of.

One of the problems with christian terrorists is that they are, as Old Goering noted, useless for goading a nation to war. What are we going to do- invade Wyoming and Michigan? Brown muslims are much better for that.

Once you start viewing "security precautions" in terms of scaring people (aka "reassuring people") and not in terms many of them start to make more sense. As an example. When I took the train from Minnesota to New Hampshire, at one stop a guy from Immigration got on the train and asked everyone for IDs. In a scene remenescent of those old World War II movies, with the Nazis walking down the train intoning "Papers, please" in a fake german accent. Note that this train was taking the *southern* route around the Great Lakes, we never left the states. When travelling between states, remember to take ID with you. What do you think this is, a free country?

And it's not like the train could be hijacked and taken to Cuba or run into a building or anything. Hint: the train goes where the train tracks go, and nowhere else. The most you could do is use it as a crowd of people to attack, but in that case the local mall (especially my local mall) would present a more inviting attack. And it's not like terrorists wouldn't have good fake (or even legitimate) IDs anyways. And it's not like he was inspecting IDs close enough to tell a good fake- he had a whole train to inspect.

No- this guy's chance of catching a terrorist was precisely 0. No, his purpose there was to instill fear. "Be afraid- terrorists can be anywhere! The guy sitting next to you could be a terrorist!" he might as well been announcing.

The other problem with white christian terrorists is that they are uncomfortable close- in looks, in religion, in ideology, in politics- to
fine upstanding if slightly eccentric members of our community (IMHO, anyone who in this age of genetic engineering and human cloning still debates evolution is a few cans short of a six pack. But the rich are simply eccentric, not nuts). Or for that matter esteemed ex-Senators and senior administration officials. It's hard to seriously pursue terrorists whom you mostly agree with.

But this does present an interesting what-if scenario. There is a major terrorist act in the US in the comming year, killing hundreds of people. But after several rounds of Bush talking up the Muslim threat to the US as why he should get re-elected, the terrorists turn out to be white christian militia members, with ties to organizations that Ashcroft also has ties to. This added to the expected hue and cry that after three years, two PATRIOT acts, two wars, and hundreds of billions of dollars spent Bush still has not made us any safer would sink the Bush administration then and there. The question is wether the evidence that the terrorists are christian would come out in time to affect the election, or stop another invasion.

Researchers in the People's Republic of Berkley have managed to managed to build carbon nanotubes on silicon chips. This opens the doors to super cheap, super sensitive bomb and chemical detectors, not to mention computer memories 10K times denser than current. That 10K number, by the way, represents about 20 years of Moore's Law. Carbon Nanotubes- in addition to these nice features, also allow us to build space elevators, faster computer busses, higher res (and cheaper) flat screens, quantum cryptography devices, and more. They are bordering on an argument for intelligent design, IMHO. Forget plastics- invest in carbon nanotubes. The only problem is that everyone else is as well.

I'd like to comment on this article in City Pages, linked to from Dave's diary (hey man, don't forget you have a blog too). I have to agree with Dave that I disagree with a lot Steve Perry says.

First off, I think Dean will stick true to his principles. What he's saying he's saying because he beleives it to be true, not due to some strategy about appealing to some demographic or another. Which is why he's quite willing to torque people off, and take unpopular positions. As witnessed by his decision to oppose the war over a year ago, when all the rest of the Democrats were lining up to bow and scrape. This approach to getting elected stands in foursquare opposition to the theory held by the DLC. And the DLC's power came from them being the custodians of the magic formula which allows you to lose elections by small margins. Which means any success by Dean meant a dust-up with DLC was inevitable. He's attacking their base of power, intentionally or no.

And he's not stupid. There are two ways of winning elections, historically- and I'm talking about November general elections here. The first is to energize your base while simultaneously depressing you're opponents base. 2002 was a classic example- the Republicans got out, the Democrats stayed home. The second way is to lure a whole bunch of people- that vast, silent majority who doesn't traditionally vote- to get out and vote for you. Dean has hit on a strategy that does both. He'd be an idiot to give up on that strategy- and Dean is no idiot.

But a war with the DLC does NOT mean a war with Democrats in general. This is the number one mistake Steve Perry makes- the DLC is not the Democratic Party.

If Dean gets the nomination, there will be a "Democrats for Bush" movement- populated with the likes of Zell Miller and Joe Lieberman. DINOs- Democrats In Name Only- in other words. Note that I predict there will be one surprising face missing from the DINO party- Hillary Clinton. Personally, I think these DINOs should take a long hard look at their party affiliation. If they want to be Republicans, that party is the next door down. Zell Miller has already endorsed Bush for President- obviously even Holy Joe is just too dang liberal for his tastes.

I don't think Dean needs to apologize for the now infamous "confederate flag" remark. This is one case where context makes a huge difference- if you read his original words in context a completely different meaning.

Which leads to an interesting question. The Dean campaign is malicously and with malice of forethought reaching out to traditionally Republican constitutiacies- NASCAR dads, Eisenhower/Goldwater fiscal conservatives, civil libertarians (especially gun-owning civil libertarians), Perot/McCain populists, etc. As well as Nader greens and Wellstone progressives. Who he is not appealing to is the Neocons and Religous Right- two vote blocks he has no hope of gaining anyways, as Bush is their man. So why wouldn't he appeal to the conservative wing of the Democrats as well?

The answer is that he is. Notice the Gore endorsement. The DLC's fair haired boy, or at least willing dupe. The #2 man in the Clinton Whitehouse, the heir apparent (four years ago) to the DLC New (business) Democrats. So why did Gore jump ship? Because Gore didn't already have a horse in this race, and his loss in the 2K elections (and wise IMHO decision not to run this time around) has cut him out of the power-broker loop. Gore is DLC in name only. Which means his decision of whom to endorse wasn't clouded by political calculations of his own future in the party.

There is an unbridgable gap in the Demcoratic party, between Dean and the DLC. But it is unbridgable because the DLC is not allowing to be bridged. They tear down any attempt at bridging the divide- they have drawn their line of death in the sand, and promised to fight to the last against Dean.

Hey, it's that time of the year. Mainly want I want to do is set my prognostications into stone, so that 12 months from now I can look back and see how well I do. There is a certain temptation to load this list with sure-bet predictions- there will be unrest in the mideast, the presidential election will be hotly contested, the sun will rise in the east and set in the west, things like that. No, I want to skate closer to the edge. Nine things I think are likely, but are by no means sure. Those are the more fun predictions. Why nine? Because that's how many I could think of.

1) Dean will win in a huge landslide, with Bush only getting a handful of states, maybe none at all. I base this on the candidate's positions right now. Bush is already walking away from seriously big states, like California, New York, and Illinois. Bush looks at those states and doesn't see any way to win. Dean looks as "solid read" states like Texas and Georgia and goes "we might be able to win those states- it's a long shot, but don't write them off just yet". This will force Bush more and more to defend what should be "safe" states- and give the momentum over to Dean. Plus, I don't see either the economy or Iraq turning around, and those will be two huge loadstones around Bush's neck.

2) Intel will announce a 64-bit x86 chip. That isn't that radical a proposition- they are almost forced to. But this is- it will be instruction set compatible with the Opteron. It'll come out later this was because Microsoft demanded it, in return for supporting the Itanic at all. Not that Microsoft support of the Itanic will help diddly squat, that chip is already dead, the corpse just hasn't stopped twitching yet. Intel will, of course, try to paint this as some big victory, but everyone will know Intel just got gutshot by AMD. The chip won't actually be available until '05 or later, but it'll be announced in '04.

3) 2004 will be the year of Linux on the desktop. Yeah, I know I've been predicting this since 1998. The delay has served Linux, however- allowing it to acquire more polish, and sneak onto yet more desktops. For example, I haven't used a windows desktop, at home or at work, regularly in 5 years. And I know I'm not alone. Linux is sneaking onto the desktop the same way it snuck into the server room- without management noticing. 2004 will be the year management notices. The driving causes will be new Microsoft licensing restrictions, the further delay of Longhorn, and yet more virii problems.

4) The Blair Goverment falls over Iraq. Bush & co. don't really care much about the Blair goverment. If by doing something- like pre-emptively pulling out of Iraq and leave Blair holding the bag, or blaming Blair for faulty information that lead to the invasion- Bush thinks he might help his (increasingly desperate) re-election bid, he won't think twice. Tough luck, Tony.

5) The US pulls out of Iraq late-june or july. The timing will be entirely decided by political calculations- early enough for the declaration of victory to have an effect on the elections, but not so early that the inevitable civil war, Turkish invasion, etc. will have an effect. Nah, this isn't aggressive enough. So I'll add- and the US initiates military actions against one of the following: Syria, North Korea, or Iran, in September or October. If it's NK, it'll be a military blockade, if Syria or Iran, an invasion. By that point, Bush should be seeing the writting on the wall from prediction #1, and possibly looking for a way to distract people from the civil war in Iraq. But the ball will bounce the wrong way- everyone will denounce the military action as blatantly political, so it won't help and will probably hurt Bush.

6) A new law will be introduced into congress mandating DRM in all PCs at least. This law will have the backing of at least three interest groups- the media company conglomerates MPAA and RIAA will be the instigators, Microsoft who will want to shut down Linux (especially on the desktop), the insider politicians threatened by Dean's internet campaign (including all the presidential wannabes from the Senate and House who will be holding grudges in a couple of weeks). All three groups will realize that the Dean insurgency will be opposed to such a bill, and if they want it passed they need to do so now. It'll either fail by narrow margins or be declared unconstitutional however.

7) There will be more financial scandals, including one that will wipe out billions of dollars of shareholder value and display Sarbanes-Oxley as the fig leaf it is. The Dean campaign will try to make an issue of this, but the press will go with Bush and downplay the entire incident.

8) More bad news for Intel. The Athlon-64 line of chips will top 3GHz real clock rate. Meanwhile, the P4 will be stuck with hot spots and excessive heat and won't make it to 4GHz. This will position AMD to retake the clock rate crown (last held by AMD with the 1GHz Athlon) in 2005. Worse yet, AMD's better per-clock performance will be that even with a 1/3rd clock rate disadvantage, the Athlon-64s will be clearly superior to anything Intel has (the P4 needs to be 50% faster clock-per-clock than the Athlon-64 to be competitive). This will become even more pronounced as 64-bit apps become more common and people discover that they're about 10% faster than their 32-bit cousins (not because of 64-bits, but because of the 8 extra registers AMD added in 64-bit mode). Come about mid-year, Intel will be forced to make an emergency process change to SOI to address these problems. Unfortunately, the process shift will explemplify "more haste, less speed" (sort of the tag line to this entire charade) and be dogged with problems and delays. Meanwhile, AMD's market share will simply grow, especially as whitebox manufacturers start churning out Linux-based 4-way and 8-way Opteron servers for 1/10th the cost of equivelent servers from Sun, HP, and IBM.
So they aren't quite as reliable- buy two.

9) Dean will be shown to have incredibly long coat tails- the Deaniacs will get out and donate money to and hit the ground for Democratic politicians across the board. The Democrats will sweep back into power in both houses of congress- it'll be 2002 or 1994 but in the other direction. Basically any "vulnerable" seat will go to the Democrats. Not veto-proof majorities, but solid majorities. The Republicans will vow to mimic the Democrats "obstructionist" tactics of the last two years. OK, that last part was a "the sun will rise in the east" prediction.

OK, so I was slumming a bit today. But I found this article, on how Dean could win in '04. On Free Republic.

Personally, I think the article underestimates the importance the economy, the debt, and Iraq will have in the election (to the detriment of Bush), and overestimates the importance of guns in the election. I think the NRA will simply sit out the election. They'd have a hard time explaining an endorsement of Bush when they gave Dean an 'A' rating, especially to the IRS wondering why they should still have tax-excempt status. Bush would like to keep the campaign on God, Guns, and Gays, but I think most people simply won't care.

But what this does show is that even the Freepers are starting to consider the possibility of a Dean presidency.

If Trippi is a Jedi, Dean is a Ninja.

As proof, I offer Dean's IMDB entry. Right there, in black and white- Dean was in "Ninja III- the Domination". As Red Shirt #5 or something, but still.

Damn. Now I have to find and watch what is obviously a piece of dreck.

I don't want to turn this blog into a bash-Josh 24/7 blog. Frankly, I don't think the guy is that important. But reading this latest entry from him made me start thinking about the change in tone we're seeing.

I've been expecting for a while that the field will collapse into a Dean and an anti-Dean, and the only question is who will be the anti-Dean. What I find humorous is that Josh doesn't seem to feel Dean's lead to be worthy of comment, that what's interesting is the fight for second place. I don't think this is a slam on Dean (it might be, but I don't think so), I think Josh has a point. Recent polls show Dean with a 19-point lead over Gephardt in Iowa, and a 25-point lead over Kerry in New Hampshire.

Gephardt could still win in Iowa, due to the fact that it's a caucus, not a primary. If you caucus for a candidate in Iowa who gets less than 15% of the vote, you can change your ballot. Currently, that includes Clark, Kerry, Lieberman, and Edwards supporters. It's a more or less open secret that the Gephardt campaign plans on heavy use of cell phones to try and woo caucus goers for these candidates to switch over to Gephardt at the caucus. This gives Gephardt a strong argument in favor of him being the anti-Dean. It also explains why Kerry is suddenly abandoning New Hampshire and campaigning hard in Iowa (although I wonder if Clark's presence in New Hampshire and absence in Iowa has anything to do with that)- if Gephardt has an unexpectedly poor showing, Kerry might be able to pick up the "cell phone vote". The argument against Gephardt is his lackluster showing elsewhere- can he beat Dean anywhere except Iowa and Missouri?

Which answers, I think, Josh's question. The Clark campaign doesn't want Gephardt (or Kerry) to win in Iowa. Anyone who defeated Dean in Iowa would immediately position themselves as the anti-Dean. A surprise upset by Kerry in Iowa is even worse (I wouldn't stay awake nights worrying about it, were I Clark). A Dean win in Iowa can be written off by saying "of course he won in Iowa- he hasn't run against me yet!" A win in New Hampshire is just local-boy advantage as well. I suppose they could write off a Gephardt win in Iowa as local-boy advantage, but the Clark campaign needs to be carefull about how often they use that line, lest people notice how close Arkansas and South Carolina are. Of course, SC is just one of six states having primaries that day- by 3 Feb we're into multi-state airwars. Which is exactly the campaign that Clark wants to run.

Also note- 3 Feb is the date for Missouri's primary. This is Gephardt's home state, a win here is expected, and an insufficiently big win for Gephardt qualifies as a loss. Of course, even a close loss here is a diaster. But if Gephardt has any anti-Dean cred come 3 Feb, a MO win is pretty much in the bag. So a Gephardt win in IA basically means a Gephardt win on 3 Feb- taking thunder away from Clark.

No, Clark does not want a Gephardt win in Iowa.