Why has (fake) AI/ML ripped through our academic system?

A few months ago, I had written a post on how the AI/ML phenomenon is destroying our engineering colleges.

Let us now come to the reasons as to why this is a worldwide phenomenon.

Consider the old days. There used to be one university and tens of affiliated colleges. Research was done in the university and there too it was not a strict requirement. In India, we still managed to produce a C V Raman, S N Bose, Homi Bhabha and Vikram Sarabhai. The affiliated colleges were pretty much like schools where the teachers only taught and then prepared the students for university exams. The teachers had some respect because good teachers could help their students get better grades. Grades had some value and at a coarser level, first division or second division made a big difference. This was pretty much the shape of higher education till the late eighties and nobody really protested. Many students failed and depression/suicides/etc. were far too uncommon. Recall the "maa mein pass ho gaya" (Oh! mother, I have passed) lines in old Hindi movies.

Down went the Soviet Union and then rose two strong forces: hard capitalism and hard distributive socialism.

Hard capitalism: A huge amount of capital flowed in and private colleges mushroomed. Good and bad money flowed in equal measure and characters of all hues and colors started to open engineering colleges. As a result, the number of engineers being produced in developing countries increased by 10-20X. In itself, this is not a bad phenomenon given the simultaneous IT boom and percolation of knowledge. But that killed the university and took away academic rigor with it. Imagine you are the President's nephew in Krakozia (a hypothetical country), would you tolerate the university failing most of your students? What will you do? You will declare your college as a university and simply give everybody 90%. Even a blank paper will fetch 50%. In a developing country with lax laws or their implementation, this is bound to happen. Hence, arose the idea of a "deemed university".

Hard distributive socialism: This operates on a single principle. All persons are created equal, they have the same abilities, and thus they, in ideal situations, should get the same grade. If they are not getting it, then blame religion, society, caste, creed, gender, race, sexuality, and the whole world other than the student herself. The fact that the student was partying throughout the semester or was caught cheating, is not a consideration. The only metric for evaluating a program is student stress. Stress reduction is the primary goal of any teaching ecosystem, the system is a tyrant, and every student is somehow a victim. The faculty and administration are presumed to be guilty before being proven "less guilty".

Q> How did the system react? Well-intentioned governments found that teaching, grades, assessment, evaluation, and sadly excellence and academic integrity had all gone down the drain. There was a need to reinject excellence into the system (subject to the constraints of the President of Krakozia's nephew owning an engineering college and his niece being a student union leader (stress reduction warrior)).

Most countries started creating hard metrics for evaluating colleges and the chief among them was the number of journal publications and citations. This was linked to all aspects of faculty growth primarily increments, perks, and promotions. In fact, papers are required for faculty to even maintain their current jobs, otherwise they can be fired with a day's notice. How do you mass produce papers and keep the nephew and niece happy? Can millions of faculty take state of the art work and overnight start beating them? Look at what will happen to our processors, software, IC engines, rockets, and steel plants, if this were indeed to be true? No worries: AI is the savior with its Python libraries. Just running a piece of Python code on some data regardless of the engineering discipline is much much easier than proving a new theorem or designing a new piece of hardware or increasing the efficiency of an engine or designing a new structure. Such sloppy work has now become the acceptable norm in most of the developing world and there are a lot of journals that also guarantee an astronomical citation count/impact factor. Who thinks about real world skills that the industry cares about?

Just collect some data (belly circumference, hours of Netflix - hours of Arnab Goswami, number of spam calls by property dealers) and correlate with your cholesterol levels and annual salary. Python libraries do the rest of the magic (you don't have to understand a thing). CNNs, GANs, LSTMs, RNNs, Transfer learning: 1 paper each. Life is set !!!

Where are we left now? Teachers and teaching have been devalued. Teachers have to become researchers regardless of their abilities and interests. Given that a gradeless utopia where only interests matter does not exist, and students have now become paying customers, the hard left + hard right have pretty much taken away the incentive to study and teach. The hard right has made degrees and grades free. Hapless governments with oversized welfare programs and raging unemployment are throwing objective criteria such as papers and citations at their colleges (deemed universities now) to somehow discipline the system. And in let's say 150 out of 197 countries, universities are happily gaming the system by producing piles of AI/ML papers to ensure that they satisfy such criteria.

Where has this left the developing countries? There is a serious skill shortage. Starting from factory operators to website designers to system admins, the market is empty. A few who have these skills charge 10 times more than what they should be paid. The gap between rich and poor countries in terms of technical knowhow is increasing. This is what the hard left and hard right have reduced us to.

Developing countries have to make a choice. Do they want to continue this scheme of things or teach their students hard skills, which will contribute to their industrial base and ultimate economic and military might. It is high time that natural intelligence, natural sincerity and natural hard work get their due.