The death of the columnist?

Artificial intelligence, Africa, and whether computers will replace me in 2023

‘As the new year begins, we find ourselves filled with hope and optimism for the future, seeing it as a time of possibility and the promise of new experiences and opportunities.’

So says ChatGPT, the recently released free OpenAI natural language model, when prompted to write a sentence about why we are usually more optimistic about the future at the start of the year.

Columnists, to name just one profession likely affected by the emergence of AI, may disagree.

OpenAI is an artificial intelligence company founded by Sam Altman and Elon Musk in 2015. In November this year, it released ChatGPT, a tool that allows anyone to ask questions in natural language. It responds within seconds.

It is a remarkable thing to observe. As many others have pointed out, it is unbelievably good at synthesizing information. Ask it to compare Schumpeter and Hayek’s views on competition policy, and you’ll get an immediate and informed answer. Or ask it to juxtapose baroque music and rap in 200 words. Or to write a poem about FIFA in the style of Shakespeare. As proof, here is the third stanza of the poem it created (in an instant):

But alas, not all is fair

In this world of FIFA where

Corruption lurks, foul deeds are done

Leaving some to question what has won

It is both awe-inspiring and, well, a little scary. Because even as I can see the obvious benefits of this technology, it seems very plausible that the future of columnists (if not even poets) is doomed. Or is it?

To find out, I did the obvious thing: I asked my new friend the machine whether his creation had put my job as a columnist on the line. His (its?) answer – perhaps, although it depends on the quality of the columnist – did little to ease my fears.

Fears that new technologies, from the spinning jenny to the ATM, will disrupt our way of life are not new. While most of us have benefited from remarkable technological advances in the last few centuries, some were, inevitably, forced to make radical changes, like finding a new job. The spinning jenny made England much wealthier, but spinners and textile workshop owners struggled to survive because of the efficiency of the new machines in the large new factories. Two centuries later, the Automated Teller Machine (ATM) was expected to destroy thousands of banking jobs in England. It did not do so and instead made bankers more efficient as they could now focus on other things.

What kind of impact will technology like OpenAI have? Let’s put it to the test (again). I asked ChatGPT to write a 300-word essay about the advantages and disadvantages of OpenAI for Africa. I also asked it to refer in its answer to five specific examples based on recent research and to include a personal anecdote. Here it goes:

OpenAI is a leading research company that has made significant contributions to the field of artificial intelligence. The company has developed a number of innovative technologies, including natural language processing and machine learning algorithms, that have the potential to revolutionize various industries and sectors.

One potential advantage of OpenAI for Africa is that it could help to drive economic growth and development on the continent. For example, according to a recent study, the use of artificial intelligence and machine learning in the agriculture sector could help to improve crop yields, reduce waste, and increase efficiency. In the healthcare sector, AI could be used to improve the accuracy and effectiveness of diagnoses and treatment decisions, leading to better patient outcomes. Additionally, AI could be used to optimize transportation networks and logistics, improving efficiency and reducing costs. These examples demonstrate how AI could lead to increased economic growth and development in Africa, which could ultimately benefit the wider population.

Another advantage of OpenAI for Africa is that it could help to address a number of pressing social and environmental challenges facing the continent. For example, artificial intelligence and machine learning could be used to improve the accuracy and effectiveness of weather forecasting and disaster response, which could help to mitigate the impact of natural disasters on communities in Africa. I have personally witnessed the devastating effects of natural disasters in Africa, and I believe that the use of these technologies could significantly improve the ability of communities to prepare and respond to such events. Additionally, AI could be used to optimize resource allocation and management, helping to conserve natural resources and reduce pollution.

However, there are also some potential disadvantages of OpenAI for Africa that should be considered. One of the main concerns is the potential for job displacement as a result of the increased use of artificial intelligence and machine learning. If these technologies become widely adopted in various industries, it could lead to the automation of many tasks that are currently performed by human workers. This could have a negative impact on employment and wages in Africa, particularly if people do not have the necessary skills and training to adapt to new roles. Another disadvantage is that AI could perpetuate existing biases and inequalities if the data used to train the algorithms is biased or unrepresentative.

Another disadvantage of OpenAI for Africa is that it could lead to the further concentration of wealth and power in the hands of a small number of individuals and organizations. This is because the development and deployment of artificial intelligence and machine learning technologies often requires significant financial and technical resources, which may not be readily available to all stakeholders in Africa. As a result, there is a risk that these technologies could further exacerbate existing inequalities and power imbalances on the continent.

In conclusion, OpenAI has the potential to bring about significant benefits for Africa, including economic growth and development, and the addressing of pressing social and environmental challenges. However, it is also important to carefully consider the potential disadvantages of these technologies, including job displacement, the perpetuation of biases and inequalities, and the concentration of wealth and power, and to take steps to mitigate these risks.

It took exactly 22 seconds to write this. It added five references to academic articles, but the links to four of them did not work. (This is probably because the model was trained on the internet up until the end of 2021, and the academic publishers’ URLs changed in the last year.) And as with most students, it struggled to keep to the word limit (534 instead of 300).

But to be perfectly honest: had a student submitted this as an essay in my undergraduate course, I would have been impressed, perhaps even scoring close to a cum laude. And I’m sure most readers, had I not been explicit about it, would not have guessed that the text had been generated by a computer rather than by your columnist. I particularly loved the personal touch – ‘I have personally witnessed the devastating effects of natural disasters in Africa’! (By the way, I asked it the exact same question in Afrikaans, and the answer was equally impressive.)

ChatGPT is, of course, correct in stating that AI offers immense opportunities for Africa. Even ChatGPT as a tool can do wonders on a continent where English is still a second language for many. But here is my (humanly and therefore fallible) prediction: it is a technology that will augment rather than displace. The questions I can ask it, and the speed at which I can get a response, will increase my efficiency. I can corroborate facts quickly, and can easily produce a semi-intelligible argument. (Politicians take note!)

But the strength of ChatGPT also hinges on the creativity of those giving the commands. And that still requires human intervention. In fact, the risk of ChatGPT is that we all become conformers, converging to a mean. As the scientist and essayist Erik Hoel recently remarked: the next generation of AI is both more intelligent and more boring than its predecessors. ‘It means that there is somewhere deep within the so-far breathtaking progress of AI advancement, a tradeoff.’ The more it synthesizes, the more it will standardise. That is great for many things, but it neglects an important part of the human experience and, importantly, the capitalist system: human creativity. Novelty, outliers, craziness, pushing-the-boundaries, out-of-the-box thinking is not something that can be coded. The parameters of a machine can only take you so far.

OpenAI will challenge certain industries more than others. My own industry – higher education – has no idea the size of the challenge it will have to confront next year, as students start using ChatGPT. We will need to rethink not only how we assess students but also what we teach them: in a world where an explanation for everything is literally a click away, how do we teach economists, engineers, and English Lit majors creativity?

As a researcher and columnist, ChatGPT will make my work more efficient, more accurate and, hopefully, give it more depth and sophistication. It will allow me to do more of the things that I enjoy – think of creative new research questions and methods – and less of the things I do not – grading undergraduate essays (hopefully!). Used correctly, ChatGPT and similar AI services will be a boon rather than a bane for most of us, even columnists.

That, at least, is my hopeful and optimistic view for the new year.

An edited version of this article was first published on News24. Photo by Evan Yang on Unsplash.