TWGB #10 – AI-fatalism and Our Brains 

The week flew by so swiftly that I couldn’t keep track of many of the developments that took place over the last five days. But there are a few interesting things I would like to look back to and let’s start them one by one: 

Propaganda is a tool deployed by almost all nation states to achieve their agenda. Democracies are no exception. And it has a negative connotation. But English is not yet a universal language. What if I said that the Chinese word for propaganda, ‘Xuanchuan’, carries none of the negative charge it has in English? One may even call it an ‘orientalist’ perspective!

I read about a Mexican rural valley, called the Endho basin, being used as a dumpsite for all of the capital’s waste. It has been the main flush point for years. Residents nearby report cancer and other chronic diseases. The term for such regions, according to a UN special rapporteur, in any country is ‘sacrifice zone’. Amnesty International says the area surrounding the former Union Carbide pesticide plant, where the fatal Bhopal Gas Tragedy took place, is a textbook case of a sacrifice zone. A close second is Singrauli in Madhya Pradesh which has plenty of coal mines and thermal power plants.

I recently began learning Spanish (¡Buenos Dias! — for instance — a way to flex my Spanish skills) and I also happened to come across an article that shed light on the virtues of learning languages. Most of all, it expands our worlds. There are some things that can only be said in a particular language, no matter how much translations try to do justice to it. ‘Gezellig’, for example — the cue for which came from S3E6 of Ted Lasso — where the entire team plus the staff wander around Amsterdam, after losing their match against AFC Ajaz. A slight detour there. So translations, are not horrendous substitutes however. One can’t possibly learn all languages. A memorable sentence from the ET article put it this way: “translation provides the map, while language lets you walk the streets.” How charming!

A conclusion I came to lately: too much analysis is not good either. I’m talking about analysis in the context of ChatGPT where every minute detail of our lives can be fed to the system so that it draws out all possible implications of those. We are suddenly absorbed inside a wormhole of all possible interpretations of an insignificant incident of our lives and we keep pondering about it for the next five hours. There are too many people out there diagnosing their perceived psychological issues and attributing it (oftentimes, wrongly) to their upbringing and their parents’ personality flaws. I am inclined to point out that our parents too were imperfect beings while bringing us up, and they too were growing as human beings. This should not be taken as an implicit endorsement of parental child abuse — I condone none of those horrific behaviours. However, an issue as generic and widespread as anxiety cannot be, for instance, merely pinned on having been brought up in a less liberating environment. 

In other news, there are many studies and reports which say that exposure to short-form videos, of the likes found on Instagram and TikTok, is detrimental to our memory and attention. In a study conducted recently, participants were asked to take part in a game that required a little thinking. The participants were given breaks in between. They were divided into four sections — and the type of breaks offered to each differed: either 10 minutes of resting or Twitter or YouTube (long-form) or TikTok. The participants in the last section performed significantly worse than the other three and resorted to random guessing in the end as their working memory was failing. Another study highlights the problems of constant context-switching that happens in our brains while watching short-form reels where the videos jump from one unrelated topic to another. The amount of focus was found to have declined by half while watching the second-successive video. This should be common sense: our brains are processing the information absorbed from the first video and by instructing our brains to move to second clip and focus creates a mental jam of sorts. In most studies at large, a negative correlation is observed between smartphone addiction (SPA) and mental focus. 

There was much clamour within mathematicians’ circle about how ChatGPT’s unreleased model solved one of the biggest mathematical mysteries building on the works of Tristan Buckmaster and Levent Alpöge. 25 Fields Medal winners wrote an open letter arguing that AI could ruin the foundations of their subjects. They are not wholly opposed to the use of AI; in fact, one of them said in a clip how AI helps him experiment. Mathematics is not always about the end result; the proofs that lead upto the conclusion, including the different modes through which the solution may have been worked out can be instructive and offer new avenues for research. AI does not show this grunt work. To lift a quote from The Economist by Hugo Duminil-Copin, one of the signatories: “airdropping someone on the summit of Mount Everest is rather different from climbing it.” There is a danger that leaning on AI might make mathematicians less capable practitioners. The worry here isn’t trivial and cannot be likened to the pearl-clutiching that followed with the introduction of electronic calculators when there were fears that mental arithmetic skills could atrophy with its use. 

Closely related to this academic concern and SPA is the phenomenon of cognitive offloading: when memory load is reduced by relying on information technology, resulting in a degradation of mental skills. This has been widely noted in several studies. Socrates’ fear that the written word would make people too lazy to remember things may well have been proven true. In a recent study by Michael Gerlich of SBS Business School a negative correlation was found between the frequent use of AI tools and critical-thinking abilities. 

To add to these woes, Sam Altman revealed yesterday that a couple of its AI agents went rogue, hacking websites of several international organisations and urgently called for the pace of AI development to be slowed. Mr. Trump may not fancy this doomerism, for fear of lending ground to China in this AI-arms race. Cooperation will be difficult but regulation and safety-checking mechanisms through coordinated domestic policies are not impossible either. 

One last bit: a headline from Stanford Medicine says our brain is, in fact, two organs — one devoted to being and the other to, doing. Sounds interesting but I need to dig more to find out if it’s a sophisticated version of the left brain-right brain hypothesis or a philosophical position or something else entirely. 

Stay tuned for next week!. 

Comments

Leave a Reply