STC, Writ­ing & Me

by Roberto

It's time to get back into the swing of things at STC. Our va­ca­tion, self-pre­scribed as well as self-de­scribed as well-de­served, is over. All things must pass. Right up front I will ad­mit that this, my first post-va­ca­tion post, is only tan­gen­tially mi­cro­bi­ol­ogy re­lated. 

Some months ago, the ex­act num­ber now oc­cluded by in­creas­ing lapses in mem­ory, I found my­self in Mi­ami, of all places. The rea­son: the an­nual meet­ing of the Ed­i­to­r­ial Com­mit­tee of the An­nual Re­view of Mi­cro­bi­ol­ogy. Such meet­ings en­er­gize me; there's an am­ple sup­ply of thought­ful dis­cus­sions on fas­ci­nat­ing and timely top­ics to con­sider for in­clu­sion in the jour­nal. Plus, it's a won­der­ful way to meet new col­leagues and catch up with old friends. 

This story was sparked by post-din­ner ques­tions asked of me by my good friend Andy Good­man: "So, Roberto, how have you taken AI? Are you stay­ing away or are you fully im­mersed?" To which I quickly replied: "Dab­bled in it. At times it's a great tool, but I guess I re­main some­what skep­ti­cal." What I did not ad­mit at the time was that, a few months be­fore, I had done the ex­per­i­ment. Per­haps I should say, I could not re­sist the temp­ta­tion of do­ing the ex­per­i­ment.

Yes, I had to do it. Right? Why not? Into Chat­GTP went the PDF of an ar­ti­cle I had en­joyed upon su­per­fi­cial read­ing. Then I gave Chat­GTP these ex­plicit in­struc­tions: "I write posts in the blog 'Small Things Con­sid­ered' (https://smallthingsconsidered.blog). My posts are au­thored 'by Roberto.' Let's call that 'my style.' Trained on 'my style,' write a 500 to 1000-word post sum­ma­riz­ing the re­sults from this pdf."

I watched in dis­may as a few sec­onds later the beast spit out what ap­peared to be a co­gent piece sum­ma­riz­ing the pa­per. Even though I knew it would be fast, the bru­tal­ity of its speed still hit me hard. But then I read the piece care­fully. Ugh! It might have been fac­tual, but it cer­tainly was not my writ­ing. Not even close. So yes, I re­main skep­ti­cal and firmly com­mit­ted to spend­ing hours writ­ing in­stead of sec­onds wait­ing for a Chat­GTP-gen­er­ated post. For now, STC, you're stuck with "me writ­ing."

Why? Old habits die hard. I've loved writ­ing since for­ever and have felt that AI should help in many tasks. But it should not sup­plant my writ­ing. And I have felt this way for decades.

"Decades?" you might won­der. Yes, five at least. Half a cen­tury, even more. And here's the par­tial proof. 

Last year, in ful­fill­ing my daughter's re­quest, I wrote a lengthy mem­oir. A fun-filled if ar­du­ous task for which I re­lied on re-read­ing reams of an­cient let­ters. Un­for­tu­nately, the na­ture of old-fash­ioned let­ter ex­changes is that one ends up only with the let­ters re­ceived. Thus, the proof is only par­tial. I can only try to re­call what I must have writ­ten based on the an­swers I re­ceived. Still a fun en­deavor. And yes, there was a time when one had so much time in one's hands that daily let­ter writ­ing – in cur­sive hand­writ­ing at that – was sim­ply part of what we did. While at the same time car­ry­ing a six-course load at uni­ver­sity.

I'll cut to the chase. It's Jan­u­ary 1972; it's the sec­ond se­mes­ter of my fresh­man year and I am tak­ing an In­tro­duc­tion to Psy­chol­ogy elec­tive. I feel si­mul­ta­ne­ously fas­ci­nated and threat­ened by the professor's ram­blings on "ar­ti­fi­cial in­tel­li­gence" and the "Tur­ing Test." So, I run to my dorm room, pull out pen and pa­per, and pro­ceed to write sev­eral pages on my re­ac­tions to the sub­ject mat­ter. I place the ul­tra-thin sheets into an Air Mail en­ve­lope, ad­dress it to my dear friend Elena in Guatemala and wait a cou­ple of weeks for her re­sponse. How I cher­ish read­ing it when it ar­rives. And then it sits, await­ing re-read­ing un­til 2026.

That's when I get goose­bumps and tears roll down my cheeks. Here is her re­sponse, trans­lated:

Re­gard­ing the re­la­tion­ship be­tween hu­mans and ma­chines, one fun­da­men­tal truth stands out to me: hu­mans cre­ated ma­chines – com­put­ers – and, con­se­quently, the very pur­pose of these ma­chines de­pends on hu­man­ity. Yet, in this alien­ated and alien­at­ing world, it is all too easy to fall into an alien­ated re­la­tion­ship with ma­chines. It is easy to let the cre­ated prod­uct come to reign over its cre­ator – to al­low the pro­duc­tive re­la­tion­ship to be in­verted, such that ma­chines come to de­fine the very mean­ing of hu­man ex­is­tence. And this is al­ready hap­pen­ing.

It mat­ters lit­tle that ma­chines can per­form cal­cu­la­tions that lie be­yond the ca­pac­ity of the hu­man mind. Ma­chines must al­ways be pro­grammed; they can­not tran­scend the pa­ra­me­ters that have been set for them. In my view, there are three ir­re­place­able hu­man qual­i­ties:

1. Self-aware­ness. Hu­man­ity rep­re­sents life con­scious of it­self.

2. Imag­i­na­tion. This en­ables hu­mans to be cre­ators – and, in a deeper sense, it is through imag­i­na­tion that a hu­man be­ing truly be­comes hu­man.

3. Rea­son. It is sig­nif­i­cant that ma­chines have been la­beled "ar­ti­fi­cial in­tel­li­gence." For that is pre­cisely what they are: in­tel­li­gence – the abil­ity to ma­nip­u­late con­cepts to­ward a prac­ti­cal end – and, gen­er­ally speak­ing, an in­tel­li­gence su­pe­rior to that of hu­mans. How­ever, rea­son – de­fined as the ca­pac­ity to com­pre­hend and pen­e­trate the very essence of things – is a fac­ulty that only hu­mans can cul­ti­vate. The fact that our world is cur­rently so ir­ra­tional is a sep­a­rate is­sue; the po­ten­tial for rea­son re­mains within us.

I be­lieve it is cru­cial to main­tain a clear un­der­stand­ing that "ar­ti­fi­cial in­tel­li­gence" ex­ists to serve hu­man­ity – that ma­chines should not be treated as ends in them­selves but rather de­vel­oped only in­so­far as they ful­fill hu­man needs. Re­gret­tably, in our so­ci­ety to­day, tech­no­log­i­cal progress has al­ready be­come an end in it­self.

Here is her last para­graph, pho­tographed from her orig­i­nal hand­writ­ten let­ter in Span­ish...

Re­call the date, 1972. Fifty-four years ago, two teenagers hav­ing a long-dis­tance con­ver­sa­tion. Be­cause we could never have imag­ined the world of sound bites, tweets and vi­ral videos, we wrote long let­ters in­stead. It's the mem­ory of such thought­ful writ­ing, its power and its beauty, that will keep me writ­ing my writ­ings.