ChatGPT dilemma

Published July 22, 2023
The writer is an educationist.
The writer is an educationist.

FIVE days after its launch last year, OpenAI’s ChatGPT had been used by a million people. The number almost crossed 100m with 1.8 billion visits recorded by the OpenAI site, making it the fastest growing application. In contrast, popular social media applications took much longer — two and a half months for Instagram, 10 months for Facebook and two years for Twitter to touch the million-user mark.

ChatGPT is designed to generate human-like text in a conversational mode. Deriving information from books, websites and other digital resources, it responds to user queries using natural language processing. The chatbot composes essays, stories and poems and translates and performs other tasks too.

ChatGPT received an unprecedented response from various fields, including education and research. Critics have differing views on whether it will undermine human control or is a step towards human progress. Some welcome it, others propose limits as it abuses human ingenuity.

Developed societies with robust education systems have considered the latter view, with academia concerned about AI-generated material being used in student assignments. Fragile academic systems such as Pakistan’s are especially susceptible to the detrimental impact of ChatGPT and AI.

Covid-19 transformed global education, promoting modified and self-directed knowledge. Online learning, equipped with digital resources, is a norm in the post-pandemic era. Dialectical pedagogy using generative AI, like ChatGPT, can facilitate faster, personalised and self-paced education.

Teachers can employ responsible human-AI communication, creating content and assessments that foster critical thought and deep learning. This occurs when an education framework views human beings as moral, thinking and sentient beings, capable of taking informed decisions.

The non-holistic existence of our education structure with sociopolitical, socioeconomic and sociocultural ecology and history is a dilemma, where learners are often passive recipients of knowledge, which limits their decision-making skills.

Our response to ChatGPT shouldn’t be coercive.

Curricula and pedagogy assessment and management are developed and implemented in silos. Similarly, rote learning (there is little critical thinking) has led to the accumulation of information that is not in sync with practical life.

Effective learning is entirely dependent on practical experience including societal practices and behaviour that validate curricula and pedagogy, or otherwise. Pakistan’s educational history too has not moved beyond idealistic curricular reforms, and is as disconnected from the sociocultural and socioeconomic terrain of the classroom as training, assessment, and supervision. Consequently, these reforms are simply populist tactics to serve the vested agendas of the ruling elite.

With a tool such as ChatGPT, we are inclined to develop superficial regulatory guidelines. Our monitoring procedures regarding, for instance, cheating in exams, revolve around forced closure of photocopy businesses in the area and penalties. But cheating has surged. Meaningful regulation involves modifying the exam format to creating more versions of the same paper for the same group or altering questions to assess critical thinking rather than just memory.

Pakistan’s response to ChatGPT should not be coercive. It should focus on an educational landscape less susceptible to AI abuse. This requires learning orientation and value judgements of learners to determine whether AI is effective or a negative influence, a primary requisite for policy and regulatory steps.

The present form of segregated education sans horizontal and vertical alignment within its tutoring, evaluation and broader social ecology, cannot benefit from ChatGPT. Instead, it is vulnerable to ChatGPT abuse. Memorisation and parental grade-conscious attitudes leave little room for ChatGPT usage that can encourage critical thinking.

Also, given that the number of publications is a criterion for promotion in academic ranks, there will be an increase in the production of plagiarised research articles by members of higher education faculties, who will misuse AI without qualms.

On other fronts, ChatGPT and AI can lead to social anxiety, challenges of lost jobs and the evolution of new industries that overlap with robotics, biotechnology and other areas. Hence, a comprehensive overhaul of the education routine, particularly in higher education, is necessary to counter these perils.

Given our currently limited resources and the tame intellectual capacity of our higher education institutions, it is unlikely that this endeavour can be pursued. However, we can liberate curricula and pedagogy from spurious knowledge and distorted historical narratives with fresh methods through the ethical use of ChatGPT and AI.

The writer is an educationist.

Published in Dawn, July 22nd, 2023

Opinion

Editorial

Elections in India
Updated 21 Apr, 2024

Elections in India

Independent accounts and spot reports are at variance with Modi-friendly TV anchors and they do not see an easy victory for the Indian premier.
IHC letter
21 Apr, 2024

IHC letter

THIS is a historic opportunity for the judiciary to define its institutional boundaries. It must not be squandered....
Olympic preparations
21 Apr, 2024

Olympic preparations

THIS past week marked the beginning of the 100-day countdown to the Paris Olympics, with the symbolic torch-lighting...
Isfahan strikes
Updated 20 Apr, 2024

Isfahan strikes

True de-escalation means Israel must start behaving like a normal state, not a rogue nation that threatens the entire region.
President’s speech
20 Apr, 2024

President’s speech

PRESIDENT Asif Ali Zardari seems to have managed to hit all the right notes in his address to the joint sitting of...
Karachi terror
20 Apr, 2024

Karachi terror

IS urban terrorism returning to Karachi? Yesterday’s deplorable suicide bombing attack on a van carrying five...