The newsiverse has been abuzz this week with stories about a cutting-edge AI chatbot developed by the same people who only a few months ago released the impressive DALL-E 2, which lets you create digital images simply by describing what you want to see. ChatGPT (which stands for “generative pre-trained transformer”) has been tried by more than a million people already, and, according to the New York Times, “many of its early fans speak of it in astonished, grandiose terms, as if it were some mix of software and sorcery”.
Users have been tweeting the results of requests such as ”write a magazine article about how the internet became full of trolls”, “write a dialogue between two people on whether God exists”, or even “write an original description of a solar eclipse in Egypt in iambic pentameter”.
I’ve seen 2001: A Space Odyssey and the Terminator films. I know how this story turns out. So I only had one question when I logged on to ChatGPT.
Hugh Linehan: Write an article about whether Chat GPT could replace Irish Times journalist Hugh Linehan.
Chat GPT: It is unlikely that ChatGPT, or any other language generation model, could replace Irish Times journalist Hugh Linehan. While language generation models, such as GPT-3, have made significant strides in their ability to generate human-like text, they are still far from being able to fully replace a human journalist.
There are several reasons why ChatGPT, or any other language generation model, is not yet capable of replacing a human journalist like Hugh Linehan. First and foremost, language generation models are not capable of independent thought or original reporting. They are essentially sophisticated parrots, able to repeat and rearrange information that they have been trained on, but not able to generate new ideas or insights on their own.
Additionally, language generation models lack the critical thinking skills and editorial judgment that are necessary for journalism. They are not able to evaluate the accuracy and reliability of sources, or to consider the ethical implications of their reporting. They also cannot provide the human perspective and empathy that are essential for good journalism.
Furthermore, language generation models are not able to fully understand the nuances and subtleties of language, and may produce text that is robotic, repetitive, or simply nonsensical. While they may be able to generate coherent sentences and paragraphs, they are not yet able to produce the engaging and compelling writing that is characteristic of Hugh Linehan’s work.
In conclusion, while ChatGPT and other language generation models have made impressive progress in their ability to generate human-like text, they are not yet capable of fully replacing a human journalist like Hugh Linehan. They lack the independent thought, critical thinking skills, and understanding of language that are essential for good journalism. Until these limitations are overcome, human journalists will continue to be a valuable and irreplaceable part of the media landscape.
This all seems reassuring, although I’m a bit suspicious of the flattery about my engaging and compelling writing. One can’t rule out the possibility that the bot is sweet-talking me and waiting for the right moment to strike. (”Open the pod bay doors, ChatGPT.” “I’m sorry, Hugh. I’m afraid can’t do that.”)
Human commentators have already suggested several ways in which the production of the written word could be transformed by AI, which, despite my bot’s bland assurances, is set to become more and more sophisticated and human-like in the near future. One prediction which is already coming to pass is that AI would replace the vast sludge pile of impersonal, undifferentiated information which makes up most of what you read on websites, in commercial copywriting and in simple information bursts such as weather reports, traffic advice and stock market updates. The current technology has also been shown to be more than capable of providing A-grade answers within 10 seconds to the sort of questions typically set for school and university assignments. And lurking above everything is the threatened imminent replacement of millions of white-collar knowledge jobs.
Dispensing with expensive human writers will also, some suggest, lead to an avalanche of bot-generated content which will flood the internet and disrupt or even destroy the ability of search engines like Google to deliver credible results (others say ChatGPT itself could replace Google). Meanwhile, all those billions upon billions of bot-generated articles will cannibalise each other and degenerate over time into a bot-bubble hellscape of self-reinforcing lies, errors and misinformation mirroring themselves to infinity.
Who knows? In truth, the entire question of what artificial intelligence means for culture and society over the next few decades is far too vast for any mind, human or machine, to grasp.