A fascinating read in the New York Times tells how Plato, the Greek philosopher more than 2,300 years ago, was concerned about the wide adoption of a form of technology called writing. Allow anyone to start writing, he lamented, and even fools and scoundrels will pretend they are someone “omniscient and wise”. Writing has, of course, since moved on and now we have ChatGPT, the generative artificial intelligence chatbot that is taking the world by storm. The startling technology uses pattern matching to simulate conversation and can write just about anything, from computer code to movie scripts and essays, and much more. All you have to do is tell it what you want. The Economist reports that by the end of January, two months after its launch, ChatGPT was being used by more than 100m people, making it the fastest-growing consumer application in history. During that time, the bot’s capabilities have been tested in every possible manner, with its possible application in education being amongst the most scrutinised. The findings have been met with surprise by experts. When prompted to sit school tests, ranging from grade four to university exams, the chatbot was shown to pass with unexpectedly good results. “It was C+ student,” proclaimed a US university after the bot passed a law exam. Other university tests ChatGPT has aced include an exam in finance and the US medical licensing exam, reputed to be one of the most difficult standardised tests there is. But there’s a downside. While the technology has demonstrated the ability to provide lengthy, thoughtful and thorough responses to questions and prompts, it is also highly susceptible to inaccuracies. Still, it is arguable that the bot’s potential for good outweighs its negatives. And, speaking about education, a debate has been raging about its potential impact in the classroom. Some American schools have banned the use of the chatbot in class, while others are embracing it at the behest of teachers arguing for its teaching and learning potential. This discussion has not gained much prominence in the region or Africa just yet. But, like many parents, I’m invested in it and would like to see our educators and policymakers join in the conversation. My pre-teen son is as precocious as any in many of our households when it comes to navigating the internet—within safety limits, of course. Should I ban the use of the ChatGPT at home in support of its probable prohibition at his school to make him “learn using his head”, as some have argued? Many are asking the same question. And like some of those following this debate, I’m persuaded by the calculator argument. When the price of a basic calculator drastically dropped in the 1980s making it affordable and widely used, many schools banned them. Today, the use of the calculator is commonplace as a pencil in the classroom, including in the exam room. And, not to make too fine a point, so did writing become widespread despite Plato’s misgivings. Yet the great philosopher probably had a point. Had he lived to see what advanced technology has wrought, he might not be surprised to witness how the scoundrels, if not the fools, are peddling disinformation and fake news on the internet and social media for nefarious ends. Experts acknowledge that ChatGPT capabilities could greatly aid the sophistication and spread of fake news and disinformation. But the technology is only now being tried out and is being tweaked to address the sticking issues. No technology is perfect and one can be fairly confident that, like the calculator, the chatbot will become commonplace, driven by both moral and economic interests. Only last week, Microsoft was announcing it is set to incorporate ChatGPT capabilities in Bing, it search engine. This is aimed at improving its market share, for which the company has heavily invested in OpenAI, the startup behind the chatbot. Not to lose out to the competition, Alphabet, the parent company of Google, also announced it would incorporate its version called Bard to its popular search engine. Alphabet will base its chatbot on LaMDA, a generative A.I. model it has been planning to use for a while now. The jostling for supremacy by the tech titans underscores the reason ChatGPT is being described in superlative terms such as the next big thing and revolutionary. It’s also a reminder of how far the innovation has come. The first ever chatbot with pattern-matching capability to simulate conversation is called ELIZA and was developed around 1966. It was followed by a succession of others of which more recent if better-known iterations include Apple’s Siri and Amazon’s Alexa, before the emergence of the Generative Pre-Trained Transformer (GPT-3) on which ChatGPT is based.
