Three years it’s sat

Posted on by

My copy of A New Kind of Sci­ence, that is. As a per­son “who might per­haps rather not be named” — but as it hap­pens is per­fectly happy to be so — I point you with pride and great respect to Cosma Shalizi’s apro­pos and sur­pris­ingly timely review of Stephen Wolfram’s mag­num what­sis.

Timely, because (per­haps by quan­tum cel­lu­lar automata shit hap­pen­ing in the very fab­ric of the uni­verse) the damned thing came up today in a con­ver­sa­tion.

This entry was posted in Uncategorized by Tozier. Bookmark the permalink.

5 thoughts on “Three years it’s sat”

  1. This is kinda nit-​​picky, I know, but one ‘graf in this review struck me as wrong:

    “Another egre­gious weak­ness is biol­ogy. Wol­fram dis­plays absolutely no under­stand­ing of evo­lu­tion, or what would be nec­es­sary to explain the adap­ta­tion of organ­isms to their envi­ron­ments. This is related to his pecu­liar views on method­ol­ogy. If you want to get a rough grasp of how the leop­ard might get its spots, then build­ing a CA model (or some­thing sim­i­lar) can be very illu­mi­nat­ing. It will not tell you whether that’s actu­ally how it works. This is an impor­tant exam­ple, because there is a clas­sic the­ory of bio­log­i­cal pat­tern for­ma­tion, or mor­pho­gen­e­sis, first for­mu­lated by Tur­ing in the 1950s, which lends itself very eas­ily to mod­el­ing in CAs, and with a lit­tle fine-​​tuning pro­duces things which look like ani­mal coats, but­ter­fly wings, etc., etc. The prob­lem is that there is absolutely no rea­son to think that’s how those pat­terns actu­ally form; no one has iden­ti­fied even a sin­gle pair of Tur­ing mor­phogens, despite decades of search­ing. Indeed, the more the biol­o­gists unrav­el­ing the actual mech­a­nisms of mor­pho­gen­e­sis, the more com­pli­cated and inel­e­gant (but reli­able) it looks. If, how­ever, you think you have explained why leop­ards are spot­ted after com­ing up with a toy model that pro­duces spots, it will not occur to you to ask why leop­ards have spots but polar bears do not, which is to say that you will sim­ply be blind to the whole prob­lem of bio­log­i­cal adaptation.”

    I’m not say­ing that Wol­fram *does* have a good grasp on biol­ogy — I don’t want to defend on ANKoS at all. But my under­stand­ing is that lat­eral inhi­bi­tion, orig­i­nat­ing with Tur­ing, elab­o­rated by Gierer and Mein­hardt and con­tin­u­ing later with oth­ers, *is* a model which is used to describe (say) pat­terns which arise dur­ing devel­op­ment. I’m think­ing here of spa­tial pat­terns in the embryo dur­ing drosophila devel­op­ment… but there are a ton of other exam­ples. I got back from a retreat about two months ago where lat­eral inhi­bi­tion mod­els were even men­tioned as a model for the way that yeast cells extend their “shmoos” dur­ing mating…

    Maybe I should write about this at QoW sometime…

  3. Oh, and I know — Nijhout’s work (which as I recall made the cover of the big green Springer book) on but­ter­fly wings is pretty clearly Tur­ing pat­terns. And seashell pat­terns, ditto. And maybe some­thing involved in api­cal meris­tem place­ment, but that escapes me at the moment. Too long not a botanist.

  4. Yeah… I mean, like I said, I’m not a biol­o­gist. But I was lis­ten­ing to some biol­o­gists who (basi­cally) were say­ing “the model of ‘lat­eral inhi­bi­tion,’ inspired by Tur­ing, is a good model for this phe­nom­e­non X with the fol­low­ing mol­e­c­u­lar com­po­nents…”, so I feel com­fort­able in chal­leng­ing at least *that* por­tion of his essay.

    But over­all, it was a fun read. I spent part of last night and this morn­ing track­ing down and read­ing the papers and reviews he links to. Read­ing Shal­izi is always.… I feel like it boosts my cre­ativ­ity. His writ­ing opens my mind — I usu­ally come up with good ideas within 6 hours of read­ing some­thing he’s writ­ten, even if it’s on a com­pletely dif­fer­ent sub­ject :-) .

  5. Turn­ing on com­ments: never. If I did imple­ment them, I’d feel obliged to keep up with them, and keep them spam-​​free, and this I really don’t have the time or the moti­va­tion to do.

    Lat­eral inhi­bi­tion: This might call for some clar­i­fi­ca­tion. Tur­ing pro­posed a very spe­cific, and very ele­gant, mech­a­nism for lat­eral inhi­bi­tion using only two chem­i­cals, which mutu­ally inhib­ited each other, one dif­fus­ing much more effec­tively than the other. The Tur­ing mech­a­nism has been demon­strated in arti­fi­cial chem­i­cal sys­tems. There are also bio­log­i­cal sys­tems which seem to use forms of lat­eral inhi­bi­tion, but with a far more com­pli­cated chem­i­cal basis than paired Tur­ing mor­phogens, which mat­ters a lot for evo­lu­tion­ary issues (as Bill could explain bet­ter than I). In all hon­esty, how­ever, it’s been a cou­ple of years since I was read­ing up on this very actively. Ref­er­ences would be appreciated.

    Praise: For­tu­nately, you can­not see me blush.