Saturday, January 29, 2005

porn and sensationalism in science

From Volokh and Instapundit, this story about monkeys "paying" to see other monkeys. I have serious doubts.

There isn't enough information in the story to know for sure, but I don't see how they are going to get the idea of money across to the monkeys. More likely they give the monkeys an immediate choice: keep looking at this picture, or get some juice. If so, then there isn't any concept of building up wealth or of extended trade-offs (meaning, I could save money now and spend it on something else later). I'll be surprised if there is any economic analogy to be drawn at all beyond choices of one immediate benefit or another. Now if they had shown that monkeys could be given spendable tokens and make the choice to save the tokens for later instead of spending them right away, now that would be amazing.

Not that this clarification renders the research worthless. It is interesting that monkeys prefer to see certain pictures than to eat. And it's interesting to know what pictures they prefer. But --assuming I'm right-- this talk of "paying" is a sensationalistic misuse of language. It's intended to get publicity and funding, not to give accurate scientific information. The accurate report of the experiment, I conjecture, is this: "monkeys sometimes prefer to look at sexually-arousing pictures of other monkeys than to have a bit of juice" But that's a lot less surprising and a lot less headline-worthy than "monkeys pay for porn" isn't it? Scientists ought to be chastised for sensationalism because it interferes with impartial and objective discussion.

Researchers in artificial intelligence are great offenders in this regard. One constantly sees articles claiming a demonstration of some human trait on a computer, but a close reading always reveals that it is nothing more than a simulation.

Imagine that your friend Tom breathlessly announces to you that he has programmed his computer to have emotions. You go to see this remarkable machine and he shows you proudly how when he logs in, he gets a message on the screen: "Why hello, Tom! It's so good to see you today!"

"See!" Tom cries, "It's happy to see me! It's demonstrating emotion!"

But all Tom did was have the computer print a message. The computer wasn't happy to see him, it was simulating a person who was happy to see him.

AI claims make the same mistake. Whenever an AI researcher claims he has a computer than can think or learn or introspect, he is either confused or he is speaking metaphorically (and if he is speaking metaphorically, he knows that he will be taken literally). Computers don't think or learn or introspect. They execute programs. They change state. They branch on state that might be viewed as more integral to the execution than normal data is.

This inflated language serves two purposes for them: it impresses the naive and it gets funding. I imagine that researchers in other fields have similar motivations.

don't they believe in America?

Rich Lowrey describes the arcane process of filling the government in Iraq. The most striking thing to me is how distinctly European it is. I'd like to know how much influence the Bush administration officials had in this system. If all the parliamentary fluff was their idea, or if they didn't object to it, I'd like to know why. The US is the longest-living democracy in the world. We have a system carefully designed with checks and balances to keep one group from taking over. It has worked remarkably well at preserving liberty and promoting success. Wouldn't it have been good for Iraq too?

We have a good track record. European-style parliamentary systems don't. Few of them are really more than a century old and they all have, at one time or another been taken over by a dedicated coalition that seriously harmed the country. I don't know how much of this difference in track records is actually due to the style of government, but you can't discount it as a factor. It's appalling that Bush administration officials seemed to take the European model as the default in spite of these facts.

I wonder how much was influenced by State Department officials with socialist leanings --people who preferred the European system because another difference in track records is that European-style parliamentary systems tend to be more prone to socialism than the American system. Were American officials, working for a reputedly strong capitalist president, actively working to promote socialism? And if so, why wasn't it stopped?

the teddy bear tragedy

I once kidnapped a teddy bear. My roommate and I were visiting at a friend's house --actually, it was probably for a Bible study, which I guess makes our behavior even more inexcusable-- and our friend --let's call her Terry because that's her name-- had this teddy bear she was inordinately fond of. She was also a really great cook.

As so often happens, these seemingly innocent and unrelated factors --the Bible study, the love of an innocent bear, the talent for baking, and possibly the imbibing of sugar-laden soft drinks-- let to a tragedy that lingers to this day.

In short, we kidnapped the bear for cookies (the kidnapping isn't the tragedy that I was talking about, that comes later). Not just any cookies, not just chocolate chip cookies, but chocolate chunk cookies. With extra chocolate. Before I go on, I'd like to observe that the practice of making laxatives in chocolate flavor may have seemed like a good idea at the time, but in point of fact, it has led to untold and unnecessary suffering.

So where was I? Oh yes. Well, we sent an appropriately constructed ransom note. We started to clip letters out of magazines but that was really tedious so we decided to just write the ransom note by hand. The incident is many years in the past, but I believe the ransom note was something like this:
We have the bear. Don't call the police or any church leaders or the bear gets it.

If you ever want to see your bear again, you must deliver two dozen freshly-baked chocolate chunk cookies to Mark and Dave at the next Bible study meeting. They aren't the kidnappers, they're just innocent go betweens who will deliver the ransom and then return the bear to you after inspecting the package for sufficient chocolate-chip-cookie content. Lots of chocolate chunks. And don't forget, Mark and Dave are completely innocent.

best wishes,
your friendly neighborhood bearnappers.
I think Mark drew a picture of the bear with a gun pointed at its little head.

Well, the exchange went off like clockwork and the cookies were delicious. I suppose this is an opportune moment to point out that I have a firm policy of not keeping sweets in the house because I have absolutely no self-control. I can go through a dozen cookies in one sitting, no problem.

You know, I should have realized that something was up with the way Terry handed us the box and the sweet way she said "Enjoy the cookies". Of course, we protested that we couldn't "enjoy" the "cookies" because we had to deliver them to the bearnappers.

I ate six or seven of them that night and Mark ate maybe half as many. The next day --well, starting pretty early-- we were suffering some odd flu-like symptoms involving dehydration and a need to stay close to home. I think we both missed classes (I'm still not at the real tragedy here).

We joked about the possibility that Terry had put laxatives in the cookies, but neither of us took that seriously. That was the real tragedy. We actually thought about the freaking laxatives, and neither one of us had the two brain cells to rub together that might have made us think: "Hey, if we thought about laxatives, maybe someone else did too!" The humiliation was worse than the, um, down time. We finished off the cookies that day and the next, suffering from those "flu-like" symptoms the whole time.

Oh, in case you missed the point, there actually was chocolate-flavored laxative in the cookies.

Friday, January 28, 2005

UN, US, and elections

Christopher Cross at Paterico's Pontifications has a story on the UN trying to get the US military in Iraq to stop promoting the elections:
Perelli and other U.N. officials are concerned that such U.S. military involvement is compromising efforts to convince the Iraqi public that Iraqis are directing the elections.
Talk about a lame excuse. The UN opposed the Iraq war and campaigned against Bush as much as they thought they could get away with. I don't think it takes any great imagination to suppose that they might also want the Iraq elections to be a failure --just to prove that they were right.

Alternatively, maybe we should just take them at their word. They are worried that Iraqis might begin to associate America with their democratic elections and maybe even be grateful for it. That would be a nightmare all right.

storyblogging and stock options

Darn it all to heck.

Sorry about the bad language there, but I'm really bummed that I forgot to announce the storyblogging carnival again. It's at Tales by Sheya this time, and entries are due by midnight tonight.

I'm thinking I won't make it this time. I've been working even longer hours than usual this last couple of weeks. On the bright side, my company's stock is up and people are starting to notice our product.

It's about freaking time.

I have to tell you though, the skepticism is understandable. We are a tiny company that competes directly with some of the biggest names in high tech, and against some of the most established products, and we are trying to tell people that we are ten to fifty times faster than the competition. Who is going to believe that? I didn't believe it when they said it in my job interview.

In fact, it led me to really worry that the company was some kind of scam: no product yet, a shady history with the Feds, and claiming ridiculous speedups over some of the most popular software products in the world. I felt I was taking a real risk in coming to work here.

Of course, now we have a real product on the market, the FCC has completely exonerated the company of any wrong doing, and I've seen the benchmarks myself. It's absolutely remarkable. I wasn't kidding when I said a year ago that I expect to do really well with my stock options.

It's funny. I can see the internals of our own product and see all this room for improvement and I think, "Gee, it's not that we are so good, it's that those other guys really, really suck."

Anyway, I guess room for improvement is always good. We already have performance that is too good to believe, and we know that when we get tired of implementing new features, we can always go back and double or quadruple the speed. It would hardly make sense to do that now because we have a hard enough time convincing people we aren't lying about the speed already.

And, aaaaah... I don't usually bother to say this but in case any investors read this blog and know what company I'm talking about I want to stress that 1. I don't speak for the company at any level, 2. this blog is entirely my personal creation and has nothing to do with the company, and 3. that thing about improving performance is a complete guess and neither the company nor I can be held to it.

Thursday, January 27, 2005

sorry about that

It looks like Donald Crankshaw stumbled into my NDA-Enforcement Device (NED). I feel bad about it, but hey, he knew the risks when he asked me a question that might have a secret answer...

Monday, January 24, 2005

common-law polygamy

So what do you call a man that lives part of the week in one house with a woman and his two children by her, and part of the week in another house with another woman and the three children he had with her? And the one he is divorced from is pregnant again? Since most (all?) states now recognize common-law marriage and his circumstances with both women certainly seem to be common-law marriage, doesn't that make him a common-law bigamist? And isn't bigamy illegal in Tennessee?

(link from Roger L. Simmon)

yes, my knuckles are white

At lunch today, the guy next to me was drinking a can of Diet Pepsi. Ahh, that sweet, sweet elixir of caffeinated bliss, that nostrum of analeptic virtue. Would that this cerulean flask of joy were sitting at my own elbow. Would that I could take into my own hand this aluminum chalice of all that is stimulating, thence to raise it with trembling anticipation to my parched lips, to taste thereon the bitter sweetness that is caffeinated beverage. Would that I could imbibe this blessed effusion, could feel the salubrious pharmacon suffuse through my viscera, bringing my bodily humors to that ecstatic state of saturation which I have known to be so profoundly therapeutic in days past. Would that I could indulge one last time...

The guy with the DP asked me why I was staring at the table and drooling so I had to explain that I was quitting caffeine but that there were some minor residual cravings. After lunch I went over to the fridge and actually had one of those blessed cerulean canisters in my hand, it was cold and inviting. No, I mean really inviting. I could hear it. It had a sexy woman's voice, "You know you want me. Here I am. All you have to do is pop my top and pull me to your lips...".

My self control prevailed at the last and I put it back. Sigh.

Sunday, January 23, 2005

exclamations marks and errors

I've tried to get the other programmers in my office to stop using exclamation marks in the messages sent to users. It's simply wrong to have the computer shout error messages at you:
Variable not defined!
String too large!
Just what are they trying to convey with those marks? Excitement? Anger? Happiness? I can't think of anything that would be conveyed by an exclamation mark that is appropriate for an error message in a professional business product. Yet I just can't get them to stop doing it. Why not?

Here is a clue: none of the programmers that do this is a native English speaker. Yet this can't be the influence of just one foreign culture because the worst offenders are from both India and Germany. And I think the guys from Russia and Taiwan do it too.

No doubt part of the problem is poor (or non-existent) instruction in English composition. But I think another problem is just that they don't have enough experience in reading English. If they did, they would recognize how seldom the mark is used and how it is always in either loud or emotionally-charged situations. As it is, all they know is what they were told in class, probably that it's some sort of emphasis.

As a slightly snarky aside: if I was writing in a foreign language and an experienced writer in that language told me I was doing something wrong, I'd listen to him.