The impact of ChatGPT on education in a corrupt system

We are perhaps in the midst of a technological revolution, one of the outputs of which has been ChatGPT. People have been more or less divided on this topic; some are quite optimistic, and others pessimistic. But what about the impact of ChatGPT in a country like Bangladesh? Is there a sufficient reason for us to be pessimistic about ChatGPT? The answer is yes. 

If one were to single out one prominent factor that makes us pessimistic about the impact of ChatGPT, that single factor would be corruption. In fact, the introduction of ChatGPT in a highly corrupt country with weak institutions is a lethal combination.

To understand the link, let us consider a research paper carried out almost 25 years ago. In it, Hall and Jones (1999) stated: poor countries have less per capita output because their social infrastructure is weak and poor, not because they don't have enough physical capital. The authors defined social infrastructure as institutions and government policies that provide incentives for firms and individuals, and these incentives ultimately determine productivity, physical capital accumulation and income. 

Two principal components of social infrastructure are government policies and institutions, and the quality of these two is determined by how much corruption there is. Weak institutions are primarily the result of corruption.  

There is indeed corruption in Bangladesh, and consequently, institutions are weak too. The incentive structure is quite distorted, thus stifling productivity and physical capital accumulation which in turn stagnates income. 

Given this, how can we establish a clear-cut link that elicits how ChatGPT can be quite destructive in a corrupt regime with weak institutions? To do this, we can rely on the shirking model pioneered by Stiglitz and Shapiro(1984). The basic insight this simple model provides is: in a corrupt economy shirking is a common phenomenon. And if the weak institutional structure is combined with corrupt practices, then the effect of shirking is devastating. 

To make this more concrete, we can look at the impact of ChatGPT on our education with a shirking perspective. It is common to see either teachers or students shirk their responsibilities: teachers don't read exam papers well or don't read assignments; students copy and paste materials and so on. 

Due to a weak institutional structure and enforcing mechanisms, this practice often kills the quality of education. Now given this, the introduction of ChatGPT will incentivise students to produce class and exam materials using ChatGPT. On the other hand, a weak institutional enforcement mechanism would let these students pass on courses being solely taught by ChatGPT, eliminating creative and independent thinking. 

So, if a trend like that were to happen, we would basically have a generation of students taught by ChatGPT, the long-term consequence of which is 'Dumb Human Capital.' Due to corruption and weak institutional structures, this shirking practice will accelerate in our country. This is why we have a reason to be very worried. 

However, ChatGPT will not be able to carry out the same level of destruction in the western nations where there are strong infrastructures. In fact, those countries will take the use of ChatGPT to the level where automation and intelligence will bring out the most productive output. The picture will be quite opposite in countries where corruption and weak institutions are prevalent: instead of taking productivity to the next level, it might encourage practices that would cripple growth and human capital. 

The implication is that we have a straightforward reminder here: ChatGPT is a signal to fix corruption and the weak institutional structure. 

Md Jamal Hossain. Illustration: TBS

Md Jamal Hossain works at the Boson Institute for Econometric Research in Germany.

Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the opinions and views of The Business Standard.