Remem­ber being taught about the six sim­ple machines? Like the wedge, the inclined plane, the screw? We were taught about them in ele­men­tary school, prob­a­bly around age 10.

So per­haps my mem­ory fails in the red­den­ing light of early mid­dle age: Where’s my ham­mer? On the list, I mean? The one you use to hit things with.

My wife points out she thinks it’s a “tool” not a “machine”. Shyeah, right, and a ramp is a machine how? Because you do so much with it?

If a ramp is a machine for mov­ing mass against an energy gra­di­ent, then dammit a ham­mer is a machine for increas­ing the impulse applied to a fixed object. Or some­thing like that.

So now four peo­ple, all I think rather smarter than I am, have asked me what sta­tis­tics books to read.

I think these folks are pro­ject­ing some­thing more than is war­ranted onto my recent point touch­ing on sta­tis­tics: They seem to think I know bet­ter! Haha! Alas, I am as dumb as a sack of wet ham­mers when it comes to mod­ern sta­tis­tics, hav­ing been crip­pled in my youth with an inci­dent involv­ing DEC PDP tele­types and Minitab [yes, I am that old.]. I am still wrestling with the effects of second-​​hand two-​​tailed p-​​value abuse, and though I retain sub­stan­tial use of my ele­men­tary prob­a­bil­ity the­ory, I have a seri­ously dimin­ished capac­ity to fathom any sta­tis­ti­cal process involv­ing greek sym­bols or ver­ti­cal lines (|).

My one scant advan­tage is that I rec­og­nize that I have a prob­lem. And that I like pretty graphs with lots of squig­gly lines and col­ors, which appar­ently abound in machine learn­ing texts and papers, and always seem to me to be fraught with pow­er­ful and con­vinc­ing impli­ca­tions about both the model and the explana­tory acu­men of the authors. [Wait; is that last one an advantage?]

Ah, but lack­ing a blue hand­i­capped hang-​​tag for my desk. I have, in fact, paid good money to the esteemed Dr. Shal­izi to tell me impor­tant sta­tis­ti­cal stuff [he advised me on a con­sult­ing gig]. Per­haps if we all ask him point­edly, he can sug­gest a rea­son­able course of action?

Why don’t more peo­ple know about (and use) genetic pro­gram­ming, espe­cially for sym­bolic regres­sion? GP is an approach that can be use­ful in all sorts of domains, for prob­lems rang­ing from exploratory data analy­sis to design automa­tion. SR can be a sub­tly infor­ma­tive com­ple­ment to sta­tis­ti­cal mod­el­ing projects, or it can be used as a mon­strously pow­er­ful open-​​ended exploratory machine learn­ing engine. It rocks.

So. Do you know any­thing about it? [Cheat­ing has been dis­cour­aged by elim­i­nat­ing out­bound links from this post.]

This has become a prob­lem for me. In seven con­ver­sa­tions in two weeks with col­leagues about work, includ­ing bosses and peers, I’ve men­tioned or advised or absolutely insisted they con­sider GP/​SR. In one case my oppo­site knew about GP but hadn’t con­sid­ered it because he only knew about pole-​​balancing and stuff; in four cases they thought I was talk­ing about genetic algo­rithms for para­me­ter opti­miza­tion (not that there’s any­thing wrong with that, but… no); in two cases I sup­pose they thought “sym­bolic regres­sion” meant some­thing ick­ily sta­ti­sticky, and didn’t want to go down that road, so they played like it was some fancy new­fan­gled numer­i­cal regres­sion tech­nique fad-​​of-​​the-​​day. Then, yes­ter­day, in a room full of peo­ple using fast but utterly opaque SVMs to do machine learn­ing, where the goal is to under­stand the sys­tem, they had thought about nei­ther Bayesian net­works nor GP/​SR, both of which could tell them impor­tant things about how the sys­tem works. And in this lat­ter case they hadn’t ever heard of SR.

I sup­pose now I have to do some­thing about it.

Sigh. More in a while.

OK. So let’s sus­pend our judge­ment and accept for a moment that the pace of inno­va­tion is, in fact, increas­ing expo­nen­tially. So it’s not that we have a lim­ited under­stand­ing of the real scope of inno­va­tion in the actual world, espe­cially where it stretches beyond our imme­di­ate expe­ri­ence. And not that we all sim­ply hear more these days about what has actu­ally been hap­pen­ing all along, since we have a few mod­ern con­trivances like news and Boing­Bo­ing and stuff. No, let’s assume the world is pro­duc­ing more inno­v­a­tive thin­gies. Faster. OK?

So. Is the adop­tion of those inno­v­a­tive thin­gies keep­ing pace? Is the rate of adop­tion of inno­va­tion speed­ing up expo­nen­tially? Because we’re assum­ing here that for every Really Good New Idea that appears this month, ten new Even Bet­ter ideas will appear next month. So I need to be a pretty perky adopter of new ideas, right?

Oth­er­wise… well, where do ideas nobody hears about go?

I just today sat in two lec­tures on Queu­ing The­ory, so maybe I’m hope­lessly mired in the dregs of defunct indus­trial civ­i­liza­tion. But, um, doesn’t some­body still have to make this stuff? Oth­er­wise, won’t it, like, back up in piles until some­body pays atten­tion to it?

Until, that is, we have self-​​making stuff. Besides, well… you know… the self-​​making stuff that lives on the planet already, I mean.

Or does the Sin­gu­lar­ity really just rep­re­sent a deeply ram­i­fied cri­sis for mail-​​order cat­a­log pub­lish­ers and marketing?

No, seri­ously. If there are more ideas all the time (maybe, but I doubt it), and more infor­ma­tion is wash­ing over all of us all the time (I doubt that even more), then are we adopt­ing and exe­cut­ing those ideas? Are we chang­ing our fun­da­men­tal behav­ior to cope with all the new infor­ma­tion? Are we all becom­ing dif­fer­ent from one another?

What, we aren’t already?

Caveat: Charles Stross’s Accelerando is a really, really good book that I rec­om­mend wholeheartedly.

One rea­son Kurzweil’s charted pre­dic­tion of the Sin­gu­lar­ity is sus­pi­ciously extra-​​super-​​singular::

I’ll tell you why. Because not only is the chart an arti­fi­cial and per­haps even con­scious attempt to fit the data to a pre­de­ter­mined con­clu­sion, but what it actu­ally rep­re­sents is the prox­im­ity of the famil­iar. We are much more aware of inno­va­tions in our cur­rent time and envi­ron­ment, and the far­ther back we look, the blur­rier the dis­tinc­tions get. We may think it’s a grand step for­ward to have these fancy cell phones that don’t tie you to a cord com­ing from the wall, but there was also a time when peo­ple thought it was rad­i­cal to be using this new bow & arrow thingie, instead of the good ol’ atlatl. We just lump that prior event into a “fling­ing pointy things” cat­e­gory and don’t think much of it. When Kurzweil rei­fies biases that way, he gets garbage, like this graph, out.

There’s a worse flaw yet, actu­ally, in the whole backward-​​facing extrav­a­ganza. These “par­a­digm shifts” that keep crop­ping up? So these are sup­pos­edly the equiv­a­lent of phase tran­si­tions in poten­tial, yes? Schum­peter­ian gales of cre­ative destruc­tion and so forth? In with the wire­less, out with the shirt­waist fac­to­ries; in with the lin­guis­tic finesse and effec­tive­ness of mod­erne fuckin’ Eng­lish, and out with illit­er­ate bab­ble of child­ish… well, every­body else in the world. Par­a­digm shift — um, that’s when every­body gives up what they was doin’, and starts on the good stuff. Right?

There are always Late Adopters. Alas, the Late Adopters in these cases can be very late indeed. Check it out: “Eukary­otic cells”! says the chart early on (except there’s way more bac­te­ria, still); “Class Mam­malia”! (except there’s way more bee­tles still); “Human ances­tors walk upright”! (just like a lot of dinosaurs, and all birds. Hey, every­body else! Get with the times! Falling for­ward off-​​balance is the Way to Rule the World!).

How’s that deep argu­ment go? The one often used by those other clear-​​headed thinkers? “If I’m descended from a mon­key, then how come there are there still monkeys?”

Maybe the folks who invoke that sparkling bit of wit, and the Sin­gu­lar­ity Now folks, they should talk. I detect intel­lec­tual lacu­nae they seem to share. What’s the word? Anthro… some­thing. Anthropo– Shoot; tip of my tongue.

What is that word? Some­thing about mid­dles. About the feel­ing that you — or your demo­graphic group, or your race, or your species — are the vital gate­way through which the past will real­ize the future.

It’s that feel­ing every­thing being about you. Anthropo… some­thing. Blast.

 
 
 
 
 
 
 
 
 
 
 
Oh yeah! Hubris.

(This, by the way, is also about you. But in a dif­fer­ent way.)