Dear Friends,
I’m sharing the ‘Commentator’ posts for the week. God, sorry these get so long! For me, much of the value of these is to really educate myself - and to read widely across the news spectrum. Whenever I do that I find it very revelatory how distorted the media landscape really is - the extent to which publications only focus in on stories that fit their brand and perspective. But that is not to say that there isn’t all kinds of great reporting out there. There is. And the hope is that this helps to spotlight some of it.
Best wishes,
Sam
RESISTING (AT LEAST IN THEORY) ChatGPT
Everybody is tired of the ChatGPT conversations - the articles that turn out, like a very unpleasant jack-in-the-box to actually be written by machine!; the point-counterpoint conversations on whether ChatGPT is a game-changing technology or just a tool. (And, incidentally, ChatGPT has been the most vociferous champion of its own unimportance. In The American Scholar Robert Zaretsky quotes an, ok, pretty funny one-liner by ChatGPT in which two unhappy academics wander into a bar and when prompted to give the bartender a punchline, ChatGPT suggests: “Why so glum my friends? ChatGPT is only a tool. Don’t let it define you.”)
But it is an important question - and ChatGPT is not just a tool. AI really is an epochal technology and now is the moment when it’s sweeping across the society and it’s still - just - possible to treat it as something other than an inevitability.
My main unhappiness this week is about the apologists for AI - people like Zaretsky and New York Times columnist Zeynip Tufekci - who are startlingly blasé about what’s going on. “‘Don’t worry about it’ is part of the appeal to Tufekci’s position,” writes Zaretsky. And both Zaretsky and Tufekci make some eye-poppingly jaunty claims about why ChatGPT isn’t so disruptive after all - namely, that writing was already dead to begin with, and so we can happily bury writing and move on to some new machine-enabled form of knowledge-transmission without skipping a beat. Tufekci does this by analogy to the development of literacy - contending that ancient peoples (Plato is her example) were just fuddy-duddies when they worried about what writing would do to the oral tradition. “Plato erred by thinking that memory itself is a goal, rather than a means for people to have facts at their call so they can make better analyses and arguments,” Tufekci writes. “The Greeks developed many techniques to memorize poems like The Odyssey with its more than 12,000 lines. Why bother to force this if you can have it all written down in books?” Zaretsky is even more explicit. “The current generation of students has moved on from writing. Literally,” he writes. “Most students fail to see the relevance of writing in a world - their world - that is largely post-literate.”
But I would tend to be a little less sanguine about my own obsolescence. First of all, Tufekci’s cheery vision of progress implies a one-way street that has never existed. Plato and the ancients understood full well the incalculable loss of the oral tradition - which always accompanies the introduction of literacy - and, with it, the loss of stunning powers of human memory and human recitation that literate cultures have never succeeded in regaining. As William Dalrymple writes in Nine Lives in a discussion of Indian orality, “Illiteracy was the one essential condition for preserving an oral epic. It seems that the illiterate have a capacity to remember in a way that the literate simply do not. It was not lack of interest, but literacy itself, that was killing the oral epic.”
Which is not an argument, of course, against literacy. Literacy brings its own beauty, and its own form of consciousness, and much of the cultural life of the last three or so millennia has been dedicated to exploring its glories. But it is not an unalloyed good. Ancient - and modern - cultures have hotly debated how they may keep aspects of their own oral traditions intact, and the effective compromise has tended to be to set aside certain members of the society (the bhopas in India, the manaschi in Central Asia, for that matter actors in the West) as keepers of a predominantly oral sensibility.
But the analogy is imperfect. The turn towards literacy occurred gradually over time and, ultimately, human beings were in control of it. Zaretsky’s acceptance of the accelerated demise of a form - the kids don’t want to read, so why should we make them? - is very difficult to take seriously; and sounds, to be honest, like the gripings of a teacher who’s lost control of his classroom.
The musician Nick Cave has risen to the occasion on this and has a much more red-blooded reaction to the potential extinction of his form. Sent by a well-meaning fan a ChatGPT version of a song ‘in the style of Nick Cave,’ Cave responded, as politely as he could manage, by writing:
What ChatGPT is, in this instance, is replication as travesty. ChatGPT may be able to write a speech or an essay or a sermon or an obituary but it cannot create a genuine song….Songs arise out of suffering, by which I mean they are predicated upon the complex, internal human struggle of creation and, well, as far as I know, algorithms don’t feel. Mark, thanks for the song, but with all the love and respect in the world, this song is bullshit.
Cave’s fulmination - which has hit a nerve - articulates, very eloquently, a distinction that I am constantly trying to make in this Substack. That we tend to have a very weak idea of what constitutes quality or worth. We tend to believe that ‘quality’ is being skilled at some activity or other; and the market reifies that skill by assigning some metric of ‘value’ to it, which we take to be reality. If AI has a use - and all breakthroughs like this have a use one way or another - it’s to allow us to let the scales fall from our eyes about this. Essentially - and this is the Western Project, Bacon and Faust - we’ve let ourselves be tricked into a very particular way of conceptualizing who we are. Derek Parfit, writing in 1984 - long past the point when the flaws in this way of thinking should have been obvious - defined human beings as “the creatures that can give and respond to reasons.” Well. Whoops. That definition applies nicely to ChatGPT, but human beings - whatever we are - would appear to be something very different.
The beauty of a breakthrough like AI is that it frees us from certain conceptions about ourselves - our fetish of reason; our belief that, as Cave puts it, “the direction is always forward, always faster.” It helps us to see that, whatever that is, it’s is not really our path; that what we enjoy is creating our own songs, being true to our own experience. Of course, we’re a little bit late in the day to have that realization. The genie is out of the bottle. And I do think the consequences will be extraordinary - it’s not just college papers, it’s that a whole way of being, the society of the white-collar knowledge worker, loses its reason for existence; and, unlike the machines, we find ourselves constrained by our own biology in our ability to adapt to the transformation. Farmers, craftsmen, blue-collar workers never really adapted to 19th century industrialization - on the whole, I would strongly argue, it made their lives worse, reduced self-sufficiency, reduced pride-in-one’s-work. White-collar workers never took the complaints of the Luddites and of John Henry all that seriously, because industrialization opened up avenues for white-collar work, but now it’s our turn, and I can’t imagine that it will be fun.
Thinking about this, I realize I may have had a sneak peak of this phenomenon through my totally overpowering obsession with chess. The machines came for chess first. Many of the developers of early computers - Alan Turing included - were avid chess players and, for a half-century, the goal of the programmers was to defeat humans. That finally happened dramatically with Deep Blue’s 1998 defeat of Garry Kasparov in a short match, and since then it’s become apparent that humans simply can’t contend with machines. The general belief is that this is fine - that computers are a tool, that humans and chess computers can co-exist - but the truth of it is that I have watched chess become less enjoyable over the course of my lifetime. Much of the fun of the game was analysis - arguments with your opponent after you’d finished the game; a concerted attempt to access the truth of the position. That activity - which preoccupied many people for around two thousand years - now essentially doesn’t exist. You ‘turn on the engine’; flip to the back of the book. That situation is sort of tolerable because computers are strictly banned from any serious competitive games - you can study and analyze all you want with the machine, but then at some point the barrier goes up and you are strictly prohibited from having any access to technology. In an only-mildly hysterical headline on the recent Hans Niemann/Magnus Carlsen cheating controversy, The New York Times wrote, “The Chess World Isn’t Ready for a Cheating Scandal.” I was annoyed by that headline when I saw it - when exactly would the chess world ‘be ready’ for a cheating scandal? - but there was a certain societal truth in it. Nobody is ready for the computers to slip across the barrier where they are fully integrated in human activities. It doesn’t really matter, contra Zaretsky, what the underlying ‘value’ of the activity in question is. The encroachment of computers on college essays does whatever it does - diminish the authority of college courses. The crossing of computers into competitive chess takes away the fun of the game - simple as that (which is also, as it happens, the point of the activity). And the incursion of AI into cultural transmission is similarly an existential threat to that activity. If you allow the machines to take your voice, your identity, then what is left of you?
RUSSIA’S DESERTION PROBLEM; UKRAINE’S COLLABORATORS
And on goes the war in Ukraine. The two stories at the moment are about escalation - the West giving the Ukrainians tanks - and desertion among Russian soldeirs, particularly the convicts.
I’m sure I’ll write about escalation next time. In terms of desertion there is this staggering stat from the Russian advocacy organization, Russia Behind Bars, that “of the 50,000 conscripts recruited by the Wagner Group, 40,000 are either dead or missing and only 10,000 are still fighting in Ukraine.” And then this complaint, posted to Telegram and reported on by Radio Free Europe, in which Russian soldiers beg their relatives back home to help them. "We are the soldiers of the 1st, 3rd, and 4th Platoons, 254th Regiment, 7th Company, 3rd Battalion. Please help us sort out the situation," the soldiers say. "Our commander gave us an order not to retreat from our positions. But the commander gave us no cover and no support. We had only machine guns, and all the rest of the weapons were damaged. Now they're accusing us of desertion, since the company commander says he didn't give the order. In sum, command doesn't care about us."
These stories have to be taken with a grain of salt, but they are consistent with overall patterns of the war. Ukrainian front-line soldiers report that the Russian convict soldiers are used for human-wave assaults. Satellite imagery finds newly-dug and rapidly-expanding graves in a Russian cemetery designated for Wanger fighters. And the discovery that many of the zinc coffins shipped home are empty is comprehensible, as the head of Russia Behind Bars puts it, if the Wagner Group “doesn’t keep tabs on soldiers who go MIA for different reasons but writes them off indiscriminately as dead.”
The larger issue is logistics. In The New York Review of Books, there’s an intelligent review on a new book about what’s gone so wrong with the Russian war effort - and why intelligence agencies universally missed just how weak the Russian striking force really was. “Moscow overloads its army with weapons but allots too little money and attention to the mundane stuff of logistics - spare parts, food, water, and the trucks to transport them - thus leaving supply lines vulnerable and making offensive operations unsustainable,” writes Fred Kaplan in The New York Review of Books. “Junior officers receive rote training, so they’re unprepared to take the initiative - a deliberate policy to keep them from rebelling against senior officers, though as a consequence, campaigns can plunge into chaos if they don’t go as planned.”
I am not all optimistic that this sort of mass logistical breakdown - which was the case a year ago and still seems to be the underlying problem from Russia’s perspective - means that Russia is somehow closer to working towards peace or that Putin would be toppled from within. It seems more to mean that the war just has enormous collateral damages - that Russia turns to a string of jury-rigged, brutal solutions, like the convict soldiers; that Russian commanders simply disregard the sufferings of their men.
Joshua Yaffa has a really extraordinary story in The New Yorker on collaboration in territory captured by the Russians. There’s no ‘point’ to it. It’s a long, bracing, very valuable read about an insoluble problem. The point is that, in a city under prolonged occupation, there are no good ethical choices. In Izyum, where Yaffa reports from, the city’s residents are angry above all at their mayor, who fled from the Russian advance to avoid the possibility of being forced to collaborate. “He abandoned us, ran away like a rat, and judges us because we didn’t do the same,” said a resident. “We lived through something that not many people would be able to bear, and now they call us collaborators?” And the stain of collaboration falls on people who did excruciatingly difficult, and in many ways necessary jobs, like burying the combat dead - Yaffa’s main character for the piece is Pavel Golub, a mobile-phone-accessory shop owner, who took on ever greater, and more unpleasant, responsibilities during the Russian occupation and was then beaten and detained when Ukrainian authorities returned to Izyum. “The city is clean, there are no bodies in the streets, there’s no stench or disease, and do you know who is responsible for this? Pavel,” a resident had been bold enough to shout at Izyum’s returning vice-mayor. She added, “And now you’re back in office and he’s in jail.”
There’s no particular indictment of anything on the Ukrainian end - it’s just a fact of war that a ‘liberating’ army has to deal with the problem of collaboration. And, certainly, as per Yaffa’s reporting, there’s no comparison between the roughing-up practiced by the Ukrainians and the systematic, routine torture that the Russian occupiers inflicted in Izyum. The point is that, in war, people simply face unendurable, unconscionable decisions. “Yes, we ate Russian canned beef,” a woman in Izyum told Yaffa. “What were we supposed to do, die of hunger?” Or, even more unbearably, as an older volunteer said to Pavel Golub when he was beginning his grave-digging work, “Let’s do this job and then forget about it like a terrible dream.”
LAB LEAK REVELATIONS
It’s well worth reading this jointly-reported piece by The Nation and The Intercept on internal deliberations within the NIH on the lab leak theory in February, 2020. These are NIH e-mails, unredacted for the first time through a Freedom of Information Act request, and they completely obliterate the idea that the ruling-out of the lab leak theory had anything to do with science.
The conversations, between scientists working assiduously to understand the mechanisms of Covid-19, sound exactly like the conversations that everybody else was having around this time, with a great deal of uncertainty about whether the furin cleavage could have had a natural origin or was more likely engineered. And the scientists, in the phase when they were doing science…..were split down the middle. Jeremy Farrar, of the Wellcome Institute, and a participant in the high-level NIH conversations wrote, “On a spectrum if 0 is nature and 100 is release—I am honestly at 50!” And, two days later, Farrar wrote of Eddie Holmes, who would go on to be a co-author of the consensus-determining ‘Proximal Origin’ paper, “Eddie would be 60:40 lab side. I remain 50:50.”
Those concerns were clearly shared by the very top brass, Anthony Fauci included. Fauci wrote on January 31:
I told [Kristian Andersen] that as soon as possible he and Eddie Holmes should get a group of evolutionary biologists together to examine carefully the data to determine if his concerns are validated. He should do this very quickly and if everyone agrees with this concern, they should report it to the appropriate authorities. I would imagine that in the USA this would be the FBI and in the UK it would be MI5.
And the documents follow, as well, the slide among the NIH brass towards summarily ruling out lab leak based on messaging rather than any science breakthrough. Ron Fouchier, a Dutch virologist, was one of the first to insist that lab-leak was an untenable hypothesis given the need to present a unified front. Further debate about the “accusation” of lab-leak would, Fouchier wrote, “unnecessarily distract top researchers from their active duties and do unnecessary harm to science in general and science in China in particular.” This perspective was very much echoed by other scientists in the discussions. “Like all of us, I do not know how this evolved, but given the concerns of so many people and the threat of further distortions on social media, it is essential that we move quickly,” Fauci wrote of the need for political expediency. And even the scientists who privately leaned towards lab-leak were willing to quash their reservations in the lead-up to the publication of the ‘Proximal Origin’ paper. “It’s fundamental science and completely neutral as written,” Ed Holmes (of the ‘60/40 lean’) wrote in an e-mail. “Did not mention other anomalies as this will make us look like loons.”
By the time of the publication of the consensus-shaping ‘Proximal Origin’ paper in Nature Magazine in March, the scientists were completely in lockstep - and the paper, which would be cited by 2,000 media outlets, treated as a ‘debunking’ of lab-leak. “Sorry, Conspiracy Theorists. Study Concludes Covid-19 Is Not a ‘Laboratory Construct,’” wrote, for instance, ABC News.
The unredacted e-mails put to bed exactly that illusion of consensus. To be clear, the evidence from the e-mails is not that scientific authorities (at least on the U.S. side) had unequivocal evidence of lab-leak. It’s that they were having a healthy, scientific debate about how the virus likely originated, and within a very short period of time, with no hard evidence, and through explicitly messaging-oriented discussion, reached the conclusion that was most expedient for them in shaping the narrative.
“These documents are important, and they should have been available earlier. The public has a right to know,” says Lawrence Gostin of Georgetown University, much too late in the day. “All the world has suffered from Covid-19, and we deserve to have all information open and transparent, with a rigorous evaluation of what the cause was.”
McGONIGAL’S 2016 REVELATIONS
And another unsettling revelation about the last few years of American history - the indictment of Charles McGonigal, who had been the special agent in charge of the FBI’s counter-intelligence division in New York in 2016, on charges of money laundering connected to the oligarch Oleg Deripaska.
The indictment has mainly to do with the accusation that McGonigal was working with Deripaska on an attempt to investigate assets of an oligarchic rival, who appears to have been Vladimir Potanin.
But this story is the kind of thing that is well worth digging under. Timothy Snyder, professor of Eastern European history at Yale, claims on his Substack that he pretty much has it figured out. Snyder is careful to frame his narrative as speculation, but, if he’s right, McGonigal is placed in a completely key position in the 2016 election. (A claim that’s echoed by Craig Unger in The New Republic.)
“The reporting on this so far seems to miss the larger implications,” writes Snyder. “The FBI special agent (McGonigal) who was charged with investigating the Trump campaign's Russian connections then went to work (according to the indictment) for that very same Russian oligarch (Deripaska). This is obviously very bad for Trump personally. But it is also very bad for FBI New York, for the FBI generally, and for the United States of America.”
Snyder had been convinced early in 2016 that Russia was following the same playbook it had used accused Eastern Europe to attempt to influence the American election. And in the midst of his reporting he “heard intimations that something was odd about the FBI office in New York. This was no secret at the time. One did not need to be close to such matters to get that drift. And given that FBI New York was the office dealing with cyber counterintelligence, this was worrying.”
In 2016, that would of course just been hearsay, but in 2018, in testimony before the House Judiciary Committee, James Comey confirmed the gist of Snyder’s hypothesis:
I was concerned that there appeared to be in the media a number of stories that might have been based on communications reporters or nonreporters like Rudy Giuliani were having with people in the New York field office. In particular, in I want to say mid-October, maybe a little bit later, Mr. Giuliani was making statements that appeared to be based on his knowledge of workings inside the FBI New York. And then my recollection is there were other stories that were in the same ballpark that gave me a general concern that we may have a leak problem -- unauthorized disclosure problem out of New York, and so I asked that it be investigated.
So in other words there were leaks coming from the New York office, which was overly close with Rudy Giuliani, and Comey believed that his hand was forced in ordering an investigation into Hillary Clinton’s e-mails.
Snyder writes:
It now appears that Comey made the public announcement because of an illicit kind of pressure from special agents in the FBI New York office. Comey believed that they would leak the investigation if he did not announce it.
If that holds up, and it’s pretty logical and has Comey’s congressional testimony backing it, then the chain of events runs: Putin looking to influence the election; both Deripaska and Giuliani with an overly close relationship with the FBI’s New York field office; the New York field office revealing the ‘October surprise’ of an investigation into Hillary’s e-mails; and Comey boneheadedly trying to get out ahead of it by making the announcement himself. And, as is pretty clear, it was that announcement that likely tipped the balance in a close race.
The Twitter Files recently have been pouring cold water on the idea that the Russian ‘troll army’ was the critical determinant in 2016, but that it is not to say that Russia was not involved in, as Snyder puts it, “a multidimensional Russian influence campaign on behalf of Trump.” And not so surprising that the critical piece of it was nothing so sophisticated as Twitter bots - that it was old-fashioned greasy money and a clever manipulation of James Comey.
Again, all this is speculation - and no one is pretending otherwise. But it does raise the very cogent possibility that McGonigal - as well as the New York field office - was the missing link in terms of tipping the election.
Excellent piece on ChatGPT. I agree with Nick Cave. I think there is something pretty unconvincing about people who have a good writing job saying, in effect, that young people don't need to learn how to write.
A heads up from you abt professor Tufeki . I thnky, she was a v r y good advocate for privacy and now throws up her hands. Abt GPT (gpt TV I wrote,)you have a good crosswind tack here. Both J ohn Fowles and Nick Cave have this dip in writing at all times into being off-putting. Maybe in both it is abt maintaining their own persons behind All Fiction, but in any case Fowles books have this treacle quality that takes getting used to and Cave is frenetic like the "drama" the folks complain abt. I mean really to say that they both are upholds of culture in their writing. Specially fowles writing the Aristos at the time he felt he had the agility and vinegar piss to land every where he had the image of a cohesive society. In other words they translate into maybe how they understand other people to think and that has a gracelessness abt it. But Fowles and Cave are political for me, fowles saying Artists do impassioned proclivity what everybody should be enabled to do w training."