Evolution was, and is, a great notion

Evolution was, and is, a great notion

Statue honoring Charles Darwin at the Natural History Museum, London • Photo by Hulki Okan Tabak on Unsplash

Originally published 17 June 2003

What was the great­est sci­en­tif­ic idea of all time?

The answer, I think, is clear: Evo­lu­tion by nat­ur­al selec­tion, con­ceived more or less simul­ta­ne­ous­ly by Charles Dar­win and Alfred Rus­sel Wal­lace in the mid-19th cen­tu­ry. It was their genius to imag­ine a way diverse organ­isms could arise from sim­ple ances­tors by pure­ly nat­ur­al process.

As Dar­win and Wal­lace clear­ly under­stood, if three con­di­tions apply — repli­ca­tion, vari­a­tion (muta­tion), and com­pe­ti­tion for lim­it­ed resources — evo­lu­tion is not just a pos­si­bil­i­ty, it is a log­i­cal necessity.

What was not under­stood in their time was the genet­ic basis for repli­ca­tion and muta­tion, so their premis­es were based on a cer­tain amount of spec­u­la­tion. Today, all three con­di­tions for evo­lu­tion are well under­stood and amply confirmed.

Nev­er­the­less, a con­cep­tu­al dif­fi­cul­ty remains, as it did for Dar­win. How can cer­tain com­plex struc­tures with mul­ti­ple inter­act­ing com­po­nents, such as the human eye, arise from any con­ceiv­able com­bi­na­tion of ran­dom, sin­gle-func­tion muta­tions of an organ­is­m’s genes?

Oppo­nents of Dar­win­ian evo­lu­tion call this the prob­lem of “irre­ducible com­plex­i­ty.” The chance that a com­bi­na­tion of ran­dom muta­tions might pro­duce some­thing as com­plex as the human eye is van­ish­ing­ly small, they say. Com­plex­i­ty requires a design­er. Evo­lu­tion may occur, but it will go nowhere unless an intel­li­gent out­side agency inter­venes at key points in the process.

The argu­ment is not new. As long ago as 1800, the Eng­lish the­olo­gian William Paley used the exam­ple of a man walk­ing on a heath who finds a pock­et watch among the peb­bles. The man might con­sid­er the peb­bles to be a prod­uct of nat­ur­al caus­es, Paley stat­ed, but he would nev­er believe the watch to be oth­er than the prod­uct of intel­li­gent design.

Until recent­ly, there was no way for evo­lu­tion­ists to per­sua­sive­ly answer this objec­tion. It was clear that organ­isms might find ways to com­bine sim­pler, inde­pen­dent­ly-evolved func­tions to per­form com­plex oper­a­tions, but the fos­sil record is too spot­ty to demon­strate the sup­posed inter­me­di­ate steps toward complexity.

Enter the high-speed dig­i­tal computer.

It is now pos­si­ble to cre­ate arti­fi­cial organ­isms — self-repli­cat­ing com­put­er pro­grams — that meet the three require­ments for evo­lu­tion: repli­ca­tion, ran­dom muta­tion, and com­pe­ti­tion. These arti­fi­cial organ­isms go through thou­sands of gen­er­a­tions in min­utes or hours of com­put­er time, and every step of the process can be observed. They evolve sur­pris­ing­ly com­plex structures.

In the May 8 [2003] issue of Nature, an inter­dis­ci­pli­nary group of researchers at Michi­gan State Uni­ver­si­ty and the Cal­i­for­nia Insti­tute of Tech­nol­o­gy describe one such experiment.

They cre­at­ed a sim­ple, self-repli­cat­ing, com­put­er virus-like ances­tral “organ­ism,” sub­ject to ran­dom muta­tions of its pro­gram. The organ­is­m’s descen­dants com­pete in cyber­space for the “ener­gy” nec­es­sary to exe­cute instruc­tions. They are reward­ed accord­ing to their evolved abil­i­ty to per­form log­i­cal oper­a­tions that were not part of the orig­i­nal program.

As in nature, some muta­tions were ben­e­fi­cial, some were neu­tral, and many were dele­te­ri­ous. Nev­er­the­less, remark­able log­i­cal com­plex­i­ty emerged. An ances­tor that could only make copies evolved descen­dants able to per­form mul­ti­ple log­ic func­tions requir­ing the coor­di­nat­ed exe­cu­tion of many pro­gram instructions.

As these com­put­er exper­i­ments con­tin­ue — and this is not the first — the insights of Dar­win and Wal­lace become ever more compelling.

For exam­ple, Stan­ford Uni­ver­si­ty’s John Kosa and col­leagues have pro­grammed a pow­er­ful com­put­er to use repli­ca­tion, muta­tion, and com­pe­ti­tion to dig­i­tal­ly evolve use­ful elec­tron­ic cir­cuits, includ­ing many devices for which patents had been pre­vi­ous­ly grant­ed to human inven­tors. They antic­i­pate that a patent might soon be grant­ed for an inven­tion that is entire­ly their com­put­er’s own.

Oppo­nents of Dar­win­ian evo­lu­tion might object that these evo­lu­tion­ary com­put­er pro­grams are them­selves prod­ucts of human intel­li­gent design. Yes, but so what? They still demon­strate that com­plex organ­isms can arise from sim­ple ances­tors by pure­ly nat­ur­al process.

Repli­ca­tion, vari­a­tion, and com­pe­ti­tion can indeed pro­duce unan­tic­i­pat­ed com­plex­i­ty. In fact, one rather fears the time when mali­cious hack­ers let loose into cyber­space destruc­tive virus-like arti­fi­cial organ­isms that employ Dar­win­ian prin­ci­ples to evolve ever more resource­ful ways of elud­ing our efforts to con­trol them.

Share this Musing: