“The future is in algorithms.”
I was first told that by Mr. Smarty Pants of the Austin Chronicle (believe it or not, Kristen and I actually worked with Mr. Smarty Pants for a while at Origin).
Over the years, it’s seemed to me that the ways the world is changing are very much driven perhaps not by the algorithm per se, but the increased awareness of how much we are swimming in a sea of mathematics.
“Algorithm detects Canadian politicians’ spin” at the New Scientist is just the latest in a long series of things like this. The news article summarizes the results of running an algorithm that detects certain patterns of speech associated with deception on political speeches in Canada.
It’s far from the only case like this. We’ve heard about Paul Ekman’s research into detecting liars using nearly invisible “micro-expressions” that flash by on people’s faces. We’ve learned that brute statistical analysis can identify the gender of someone based on how they answer a quiz (this one looks to be a more complex version of the Simon Baron-Cohen stuff I referenced in the book) or how they write a passage of text. (The preceding was written by a male, sez the algorithm. But many of my short stories, such as “Breathe”, report as female. On gender-based tests, I tend to show up as exhibiting strong traits from both genders, which supposedly makes me of interest to scientists because they want to dissect me).
There’s all sorts of interesting learning flying about from the mere ability to gather together the results of truly large amounts of people and examine the characteristics that are found.
Some of the results are a little scary: we learn just how much we live our lives in thrall to automatic reactions, to prejudices that are inborn, to patterns that are swirling in the reality around us and that we just treat as the current of our daily lives, wafting us from belief to belief and position to position.
When you examine the ways in which we vote or buy it turns out that there’s a lot less of us in it than there is of all of us. And as time marches on, we learn more about the algorithms underlying ourselves.
This stuff fascinates me. Anyone got more?
13 Responses to ““The future is in algorithms.””
Sorry, the comment form is closed at this time.
Blogroll Joel on SoftwareRaph Koster Sunny Walker Thoughts for Now Sex, Lies and Advertising
Just a technical note: your link URL to the Austin Chronicle in the first sentence is setup incorrectly. This isn’t so much of a problem on its own, but somehow the mistake has completely corrupted the LiveJournal Syndicated version of this post.
IQpierce: That is awesome.
Raph: The Road to Reality has fascinated me for the past month or so. The first chapter suggests Penrose’s personal perspective on how “mental perceptions”, “physical reality” and “mathematical truth” interplay. I’ve become a quiet advocate of mathematics in the past few years, from Chomsky to Godel.
I think I fixed that link, and maybe it went back out on the RSS feed when I tried editing it?
Saraid, I just looked up the Penrose book on Amazon, and frankly, it’s scaring me away. 🙂 All the reviews seem to emphasize how it starts with calculus and gets tougher from there…! As someone who never actually took calculus, I’m not sure I could get thru 1000 pages that use it as a starting point.
(I am saraid; I hadn’t nicknamed yet when I wrote that comment.)
First of all, the RSS feed fixed.
Second, it doesn’t. Chapter 1 has no math. Chapter 2 leads with the Pythagorean Theorem and seems to cover geometry. The first brush with calculus is Chapter 6.
Chapter 1 deals with mathematical philosophy, really, and underlines why math is an important aspect of existence. I don’t expect the book to be easy; the reviews say very pointedly that it won’t be. So I plan to finish this book sometime around December 2006. =P It’ll be slow and tough, but it’ll hopefully be worth it.
I laugh at their silly gender quiz.
Part 1, Angles: Scored better than the average male.
Part 2, Spot the Difference: Scored dead center between both genders.
Part 3, Hands: Right brained. Not so — right in the middle, really. But you can’t hold both thumbs on top.
Part 4, Emotions and Systems: Scored better than average female on empathizing, and better than average male on systemizing. Whatever.
Part 5, Eyes: 5. Weak. In my defense, I was distracted by all that appalling mascara.
Part 6, Fingers: Completely useless measure. I’m 0.94 on one hand and 1.03 on the other! And that pretty much tells the whole story.
Part 7, Faces: Why did they bother with this one?
Part 8, 3D Shapes: 12 of 12. And I do not have unusually high testosterone levels at all. Nor does my mother. I’m sorry, I just don’t buy it.
Part 9, Words: It mistabulated the number of words I typed. I should’ve scored really high on this one, but they thought I typed only one word. Tsk.
Part 10, Ultimatum: Oh, so now I have low testosterone, instead. Thanks for demonstrating how silly this all is.
Overall, I scored exactly the same as the average male. My boobs beg to differ.
The Gender Genie is down for repairs, so alas, my prose will not reveal the truth of things. Ultimately, I think that I am like you, Raph, showing strong traits of both genders.
Anyway, back on topic…
This is a good read:
http://physicsweb.org/articles/world/16/10/7/1
Though, honestly, I’d bet good money you’ve read about all of this before, somewhere (and probably even commented on it).
Money quote:
“Yet modern physical models of social phenomena are not really imposing some deterministic tyranny on human actions. Rather, they are simply acknowledging that in reality our choices are often extremely limited.”
As a person interested in games as well as math, check out Stephen Wolfram’s A New Kind of Science… while I found some of his rhetoric a bit much, the richness and complexity of cellular systems built on simple rule sets is very provocative (and very familiar to anyone who plays or makes games). Mathematica – the ultimate game builder’s kit.
One could probably move into a legitimate, deep discussion of the significance of games as both an artform and something that is “wired” at the core of us.
I’ve done a moderate amount of work with cellular automata as part of games work. There’s a TON of stuff that can be done, some of it somewhat outside the box.
Good article, Tess! That’s exactly the kind of stuff I was referencing; and yeah, I was familiar with all those examples already. Small Worlds does indeed cite some of that stuff, in particular the segregation study.
I think you forgot the commas between words on that test. I did too, at first, then caught the error and went back and inserted them.
An earlier quiz based on Baron-Cohen’s work showed me as a “B” brain, high-functioning in both systematizing and empathizing. I hypothesized that most game designers, particularly of online games, might fall into that category, and when I asked MUD-Dev folks to take the quiz, they mostly came out as B’s.
The quiz I was really looking for was one that used to be on one of those tacky dating sites, which did something clever: they asked your gender, then put you through a battery of goofy questions. Then they correlated particular answers to gender. After a few million people took the quiz, it got very accurate at predicting gender. Alas, I don’t know where to find that quiz now. 🙁
I have worked as a healthcare analyst and data miner for years now, spend most my day thinking about different algorithm’s constructs and data modeling…..so yes they are in fact not only in what you buy or how you vote, but also in how you use healthcare, and how healthy you are….
I cant predictivly model the price of a candy bar and who will buy it, but I can tell you who is likely to get diabetes from eating it 😛
I went off on a rant about the Baron-Cohen work on women_dev at one point (sure you can find it in the archives; you might have even participated in the discussion), and one of the folks who worked on the research explained that most people are the mixed type, and it’s not actually at all unusual for someone to be good at both.
Apparently, the press did them very poor justice in talking about their research, both by emphasizing the whole “male” and “female” brain thing (they were only meant to be names of types, and not necessarily so firmly associated with genders), and by making it sound like most people were meant to be systemizers OR empathizers, when really most people are relatively balanced. It doesn’t surprise me. The popular press is notorious for making a shambles of science reporting.
Still, though, my biggest problem with the B-C work is that I can’t figure out how it’s actually useful. Systemizing and empathizing are clearly not a single axis. So, most people are in this big cloud of B, and they have almost nothing to say about us.
For them, it’s useful. They’re trying to understand autism. They know there’s a large overlap between systemizers and autistics, and no overlap between empathizers and autistics. But, we knew that already. The whole thing feels a bit tautological, really. What was the experiment? One of these days, I’m going to dig up the actual study, so I can stop ranting, and figure out what hypothesis they were trying to prove. 🙂
Pretty much all gender differences seem to be of that sort — two hugely overlapping bell curves, with the extreme cases being what gives us stereotypes.
I must say though, that my understanding of the Baron-Cohen is slightly different from yours — yes, most people are a balance, but the balance tends to tip one way or another, and what’s unusual is showing high functioning in both areas (I seem to recall a chart marking the B’s in that region of the graph).
My understanding of Baron-Cohen’s research comes from his book (which was admittedly a bit more aimed at a pop-sci crowd). It was quite good — I recommend it.
Hmm, I can’t find the graph-that-was-all-over-the-press now, but my recollection of it pretty much had the Bs defined as anyone who wasn’t clearly E or S.
Anyway, it’s not a big deal. I’ll have to look the book up. I do have a solid respect for some of the other things that Baron-Cohen has worked on. His research on the theory of mind deficiency in autistics is of particular interest to me, having been raised by someone on the spectrum. My siblings and I were sometimes accused of using ESP, because we appeared to have an uncanny ability to read each other and predict one another’s actions when we played games. 😉