May 11, 2002

5:22 PM

"The smartest scientist on the planet." [NYT reg req] "Conducting experiments on a computer, where he says he has logged 100 million keystrokes in the last 10 years, Mr. Wolfram wrote simple programs that generated odd and intricate patterns to test his ideas about complexity. He argues that natural phenomena can be explored as if they were, in fact, computer programs." Stephen Wolfram's own company (Wolfram Media Inc.) is now publishing his 1,197-page book - "A New Kind Science " - which was kept secret until now. They claim "..he is proposing a paradigm shift. A new twist on everything.." in explaining how the universe operates. Sounds big. Is it really?
posted by Voyageman (33 comments total)

A New Kind of Science. Sorry.
posted by Voyageman at 5:24 PM on May 11, 2002

I almost just posted about this...weird.
posted by jeb at 5:30 PM on May 11, 2002

"Forty-Two"

Sorry Couldn't resist.

It is pretty mindboggling stuff though. I'm not quite certain it'll work out as Wolfram says, though. Even hard science requires the presence of "maybe" and as I understand it*, even the most complex computational devices are still just "yes" and "no" when you boil it all down. The only thing we can do now is add a lot of "if x=y then z" type statements to qualify it's conclusions. Quite frankly, I don't know if the natural world acts quite that neatly. Though the behavior of my desktop PC may sometimes convince me otherwise, the computer is still not up to processing the idea of the "fluke" which is seemingly what makes the world go round.

*admittedly not as well as some here, and if I'm wrong feel free to correct.
posted by jonmc at 5:36 PM on May 11, 2002

Of interest?
posted by Opus Dark at 5:37 PM on May 11, 2002

The big ego, the story of an academic outsider, the simple leads to the complex, a paradigm shift ... it all sounds so familar ... ah ... yes
posted by vacapinta at 5:48 PM on May 11, 2002

I guess this makes it a double post.
posted by Voyageman at 5:51 PM on May 11, 2002

I'm sorry, can someone explain how this is earth-shattering? If I'm understanding this correctly, he's basically saying that complex things emerge from simple ones. Well, duh.
posted by solistrato at 5:58 PM on May 11, 2002

Wolfram must be a very intelligent man, so I think his theories deserve serious consideration. I'll wait until the experts in the field (whatever that is) evaluate his writings. Though, for some reason, when I read about him I am reminded of this guy.
posted by epimorph at 6:00 PM on May 11, 2002

"I'm sorry, can someone explain how this is earth-shattering? If I'm understanding this correctly, he's basically saying that complex things emerge from simple ones. Well, duh."

it sounds more like he's saying that he thinks he knows something deep about *how* complex things emerge from simple ones and how we can investigate the territory better.
posted by muppetboy at 6:03 PM on May 11, 2002

Oh we're way beyond double post territory.
posted by techgnollogic at 6:29 PM on May 11, 2002

So far MeFi has bitten on this in November 2000, January 2001, October 2001, and now May 2002... the Wolfram hype machine rolls on and on. How does he do it? Note that he hasn't published anything since the 80s, when he went the commercial route, and this book is not published by a traditional publisher, much less in any kind of peer-reviewed publication--it is self-published.

Don't misunderstand me: Wolfram is undoubtedly a smart guy, and did some important work on cellular automata a long way back; Mathematica is a wonderful program, and Wolfram's book might be great--I am looking forward to checking it out. But it's astonishing to me that Wolfram keeps laying down the hype, and the press (and MeFi) keeps on taking the bait. For days to go, and then we'll find out if the years of hype for A New Kind of Science has been justified. I wouldn't bet the farm on it.

By the way, how many keystrokes have you "logged"? Any idea? Is a hundred million a lot? Does it matter?
posted by rodii at 6:41 PM on May 11, 2002

Oh my gosh, I completely missed January 2002! Thanks techngnollogic.
posted by rodii at 6:45 PM on May 11, 2002

NYT: “[Wolfram] says he has logged 100 million keystrokes in the last 10 years, Mr. Wolfram wrote simple programs that generated odd and intricate patterns to test his ideas about complexity. He then tried to imitate designs found in nature. He argues that natural phenomena can be explored as if they were, in fact, computer programs, their evolution and behavior the products of intricate calculations. ”

SCIAM: "If Darwin had had a computer on his desk," he exclaims, "who knows what he could have discovered!" What indeed: Charles Darwin might have discovered a great deal about computers and very little about nature.

From Complexity to Perplexity
posted by raaka at 6:51 PM on May 11, 2002

I guess this makes it a double post.

Heh...unrelenting...triggered by some curious fundamental oscillation of the MeFi universe itself...predictable, no doubt, using the right algorithm...

Raaka - great link!
posted by Opus Dark at 7:08 PM on May 11, 2002

Conducting experiments on a computer, where he says he has logged 100 million keystrokes in the last 10 year

If he was really smart he would have discovered macros and libraries by now.
posted by skallas at 7:10 PM on May 11, 2002

says he has logged 100 million keystrokes in the last 10 years

Ok, thats a lot of keystrokes. I'm wondering how many of those were the backspace key.
posted by vacapinta at 7:13 PM on May 11, 2002

Hype or not, there's some cool stuff on his website. Cellular automata that figure out primes and stuff...
posted by jeb at 7:21 PM on May 11, 2002

This book is the current Segway of science. Prepare to be underwhelmed.
posted by Faze at 7:30 PM on May 11, 2002

There is a very balanced ( at least I found it balanced) story on Wolfram and his book 'A new kind of Science' by Steven Levy on the current print issue of Wired. Levy has been following Wolfram since the early eighties and has read the book. The article in not yet available online. If you are interested on the subject, you should try to check it out. It definitely has more substance than the NYT story. (Incidentally, I now think Kottke was right when he said that Wired has been getting better over the last few months).
posted by justlooking at 7:42 PM on May 11, 2002

100 million keystrokes in 10 years is about two hours of typing at 35 wpm per day, which I reckon most of us log just keeping up with email. No wonder RSI is getting more common!

But this is a funny way to quantify scientific productivity. Someone in the unix research department at Bell Labs (Brian Kernighan?) once suggested that programmer productivity should be counted not in lines of code produced, but rather in lines of code removed.
posted by myl at 9:50 PM on May 11, 2002

"..he is proposing a paradigm shift. A new twist on everything.."

Capital! Make this man my new executive vice-president!
posted by crunchburger at 9:53 PM on May 11, 2002

If I'm understanding this correctly, he's basically saying that complex things emerge from simple ones. Well, duh.

I think that here, something is "complex" primarily if it is "unpredictable." But the traditional meaning of complex - a lot of features - is also used.

When scientists try to come up with an explanation for a complex (unpredictable) object, the explanation is invariably complex (lots of features). For example, take a drip from a leaky faucet. Attempting to model just the surface of a drip could consume a physicist for a year or more. The results of all of the simple rules of science - relativity, newton's laws, electricity - are always clean and mostly non-complex.

In everyday life, complex things arise from simple ones usually because those simple ones are a lot more complex than we think. A leaky faucet is actually very complex, it has a lot of tricky little surfaces, there are a gazillion water molecules, etc.

It sounds like he's trying to do to science what Godel did to mathematics with the Incompleteness Theorem. That is, he's trying to establish firm limits on the usefullness of science, using the tools of science. Perhaps he will actually make complexity theory worthy of its hype. I doubt it, but who knows.

Chaitin, the mathematician/computer scientist quoted in the article, is very brilliant, and his support of Wolfram is a credit to this new book. However, Chaitin's work basically says that the deep foundations of math are totally random facts, and not some deep truth about the universe (that's a terrible bastardization of Chaitin, but it's the best I can do). So it's possible that Chaitin might be a bit over-eager to support a similar claim for science...
posted by Llama-Lime at 11:39 PM on May 11, 2002

Chaitin, the mathematician/computer scientist quoted in the article, is very brilliant, and his support of Wolfram is a credit to this new book

I respect Chaitin too but his endorsement struck me as lukewarm at best.
posted by vacapinta at 11:59 PM on May 11, 2002

Wolfram brings to mind a real life Max Cohen.
posted by Modem Ovary at 12:25 AM on May 12, 2002

It sounds like he's trying to do to science what Godel did to mathematics with the Incompleteness Theorem. That is, he's trying to establish firm limits on the usefullness of science, using the tools of science.

Llama-Lime, please excuse this bit of nitpicking, and my going a bit off topic, but I don't think that your characterization of Godel's theorem is fair. The main consequence it has for mathematics is that it shows that there are mathematical statements that can neither be proved nor disproved by mathematics, as we know it. With minor exceptions (such as in the case of Whitehead's Conjecture, which is one such statement) this is of interest primarily to logicians. The other mathematicians, for the most part, just go on with their lives as if Godel never existed. Godel is especially rarely mentioned in applied math, the part of mathematics that tries to be "useful" by modeling things in the world, which is also about the only part that is relevant to what Wolfram is talking about. (Most of the rest of math is just manipulation of abstract statements that have nothing to do with anything in the physical world or, at least, are not intended to have anything to do with the physical world.)
posted by epimorph at 3:13 AM on May 12, 2002

Wolfram is an excellent name for a scientist.
posted by selton at 4:47 AM on May 12, 2002

Actually, it's an excellent name for an element. Oh wait, that's tungsten.

As for Godel's theorem, sure it doesn't come up often in day-to-day math just because most mathematicians are looking for something interesting or useful, not for the complete set of mathematical truths (whose horizons keep getting expanded as we define new objects and relationships). However, the method of proving the theorem has come in handy, as people like Turing used it to great effect to show the completeness/incompleteness of certain systems.

In any case, publishing a 1000+ page tome sounds like proof by intimidation to me. Has he published in any refereed journals of late?
posted by meep at 6:33 AM on May 12, 2002

No he has not. Since he started developing Mathematica, he has been deep into the proprietary model of research, which I think is hostile to scientific progress, if not antithetical. There's been some discussion of this in some of the earlier threads cited above.

As for Llama-Lime's comments above: I don't think he's realy trying to establish "limits on the usefulness of science" per se, but on the Newtonian (or Galilean) style, in which everything physical is modeled by equations, and the relations expressed by the equations are between continuous quantities--in other words, the primary tool of physical science as calculus/analysis. There's been some discussion of this in the Wolfram threads too.

I think there's a good chance Wolfram might go down as the D'Arcy Thompson of the early 21st century (or, as much as it pains me to say this, the Gregory Bateson)--an insightful noticer of patterns whose work turns out to have little or no scientific relevance in the long run and is ultimately remembered for its beauty.
posted by rodii at 9:18 AM on May 12, 2002

I think there's a good chance Wolfram might go down as the D'Arcy Thompson of the early 21st century

I was also thinking of Goethe, who believed his Theory of Colors was his most important work - luckily he was not only a great scientist but also wrote stuff like Faust.
posted by vacapinta at 10:08 AM on May 12, 2002

I thought of Goethe too, but figured I had already exceeded my pretention allowance for one thread. The thing about Goethe, Thompson, Bateson is that they were conceptually brilliant but empirical dead ends--their theories just didn't lead anywhere and, as in the case of Goethe's theory of color, were just wrong. There's no real causal mechanism at work, so they have to posit fairly outlandish accounts of reality--like "the universe is a computer" to be able to tell a real story.
posted by rodii at 5:15 PM on May 12, 2002

Oh we're way beyond double post territory.

And yet I hadn't known of Wolfram before now. But I'm glad I do. Strange, that...
posted by holycola at 10:39 PM on May 12, 2002

hey that wired article justlooking talked about is available online now :) and so is a new kind of science!
posted by kliuless at 1:36 PM on May 14, 2002

That Wired article is pretty good, and makes me (me!) want to read the book.
posted by rodii at 3:38 PM on May 14, 2002