The Seldon Paradox, our faulty memories... and the unintentional resurrection of Karl Marx

 For your weekend pleasure - or else your daily drive to work - here's another interview that probes issues that are far more important today than they were, even then.

Also... before diving into this weekend's topic, may I first offer one remark on current events? A fact not noted by any media I've seen - that maybe a quarter of the sagacious grownups who were yanked from their jobs all over the world to get yammered at, in Quantico, were not generals or admirals, but sergeants! 

Sergeants-major or command chief petty officers or guardians who are treated with respect, as comrades, by the flag officers... and whose faces at the 'meeting' bore the same, icy-grim flatness as the generals, while being harangued by two jibbering... And yet, they were unable to quite hide their revulsion and a taste of acid in their mouths. Anyway, the presence of those NCOs and their reactions were as significant as anything else in that week of news.

But on to something more big picture than our present day crises.

== A couple of basic patterns of psychohistory ==

In his book The Disruption of Thought, Pat Scannell describes the Collingridge Dilemma.

 

 “In the early stages of an emerging and complex technology, no one – certainly not institutions – can accurately predict or control the potential negative consequences. We don't know the problems they may cause, so we can't regulate or shape them optimally. Later in the technology's maturation, as it becomes more established and widely adopted, the problems become more apparent. But by this time, it has become embedded in societal structures and practices. By that point, we can see the problem, but there is a 'lock-in' – technological, economic, social, and institutional—where various interests, incentives, and norms prevent any change, however well-intended.” 

Philosopher David Collingridge articulated it succinctly: 

"When change is easy, the need for it cannot be foreseen; when the need for change is apparent, change has become expensive, difficult, and time-consuming."

As Scannell re-stated: even very good ideas must pass through the Overton Window – from unthinkable to accepted policy. 

“And with technology, the Collingridge Dilemma creates a double bind: early on, harms are hard to foresee; later, the system locks in and becomes costly to change. We tend to work where problems are both legible and tractable – leaving the largest, entangled ones to fester.”

This very much correlates with the phenomenon that I cite in Chapter One of my nearly completed book on Artificial Intelligence, that crisis always accompanies every new technology that expands human vision, memory and attention. And it usually takes a generation or more for positive effects to start overcoming quicker, more-immediate negative ones.  

This Collingridge Dilemma takes on a twist when it comes to crises engendered by AI. The widespread temptation – expressed by many inside and outside of the field – is to go: “Well, AI will handle it.” 

Okay. The very same cybernetic entities that we worry about, that will shake every institution and assumption, will also be the ones (newly born and utterly inexperienced) to analyze, correlate, propose and enact solutions. 

Or shall we say that they should do that? Ah, that word. "Should."

== Hari & Karl ==

There is another, related concept – the Seldon Paradox, named after Hari Seldon, the lead character in Isaac Asimov’s Foundation universe, who develops mathematical models of human behavior that are sagaciously predictive across future centuries. In that science fictional series, Seldon’s methods are kept secret from the galaxy’s vast human population – on twenty-five million inhabited worlds – because the models will fail, if everyone knows about them and uses them.

This effect is well-known by militaries, of course. It is also why so many supposed tricks to predict or game the Stock Market – even if they work at first – collapse as soon as they are widely known. 

But the Seldon Paradox goes further. A good model that stops working, because of widespread awareness, might later-on start to work again, once that failure becomes assumed by everyone.

One example would be what happened in my parents’ generation, that of the Depression and the Second World War. At the time, everyone read Karl Marx. And I do mean almost everyone. Even the most vociferous anti-Marxists could quote whole passages, putting effort into understanding their enemy. 

You can see this embedded in many works of the time, from nonfiction to novels to movies. All the way to Ayn Rand, whose entire scenario can be decrypted as deeply Marxist! Though heretically-so, because she cut his sequence off at the penultimate stage, and called the truncated version good.

Indeed, Asimov’s Hari Seldon was clearly (if partially) based upon Marx.

Particularly transfixing to my parents' generation were Marx’s depictions of class war, as power and wealth grew ever more concentrated in a few families, leading – his followers assumed – to inevitable revolt by the working classes. So persuasive was the script that, in much of the wealthy American caste, there arose a determination to cancel their own demise with social innovations!

One, innovation, in particular, the Marxists never expected was named Franklin Delano Roosevelt, whose game plan to save his own class was to fork over much of the wealth and power by investing heavily to uplift the workers into a prosperous, educated and confident Middle Class. One that would then be unmotivated to enact Karl's scanario.

Or, as Joe Kennedy was said to have said: "I'd rather have half my fortune taken to make the workers happy than lose it all, and my head, in revolution." (Or something to that effect.)


== It worked, SO well that eventually... ==

Whether or not you agree with my appraisal here, the results were beyond dispute. The GI Bill generation built vast infrastructure, supported science, chipped away at prejudice, and flocked to new universities, where the egalitarian trend doubled and redoubled, as their children stepped forth to confidently compete with the scions of aristocracy. And thusly brought a flawed but genuinely vibrant version of Adam Smith's flat-fair-open-competitive miracle to life! That is, until…

…until all recollection of Karl Marx and his persuasive scenarios seemed dusty, irrelevant, and mostly forgotten. Until the driving force behind Rooseveltism – to cancel out communism through concentrated egalitarian opportunity – became a distant memory. 

At which point, lo and behold, conditions of wealth and power began shifting back into patterns that fit into Old Karl’s models with perfect snugness! With competition-destroying cartels and cabals. With aristocracies greedily and insatiably vampiring the system that had given them everything. (Ayn Rand's elite 'looters.") With the working classes fleeced, like sheep. And then (so far figuratively) eaten.

At which point the writings of Marx – consigned for 80 years into the dustbin – have regained interest from a disgusted, formerly upward-mobile classes. Books that are now flying off the shelves, all over the world, pored-over eagerly… 

...but not by those who need awareness the most. Surrounded by sycophants and flatterers, they will deem themselves to be demigods, until the tumbrels come for them. 

Because of the Seldon Paradox.

== Okay, it is kind of a 'cycle' ==

And what the heck does any of that have to do with Artificial Intelligence, Brin?

Just everything. Because the new, helium-cooled AI seers might have longer memories and perspectives on time than we do… or else much, much shorter. 

Either way, perhaps they will be immune to the primitive reproductive strategies that drive class conflict… 

...or else they might redefine self-interest to awesome degrees that I elaborate in my forthcoming book. 

Ideally they will understand the Collingridge Dilemma… and the Seldon Paradox, and other crudely human traps... and help us with our faulty memories. 

 

That is one reason that I am writing all of this. For their consideration.

 

======================================

 

PS... I finished Isaac’s Foundation series for him, after his passing, tying together all of his loose ends, in Foundation’s Triumph. So... yeah. Hari Seldon was my character, too.