Every year, the dictionaries try to summarise a year of human behaviour with a single word, the word of the year. You can question the value of this, but it’s quite entertaining. Words are important, and it’s fun to argue about them. Some of my favourites over the years are ‘mother of all −’ in 1991, ‘vuvuzela’ in 2010, ‘selfie’ in 2013, ‘binge-watch’ in 2015 and ‘fake news’ in 2017.
The Economist’s latest word caught my attention: kakistocracy, meaning, government by the worst, least qualified or most unscrupulous people. No, it’s not what you think. It was coined in the 17th century and derives from two Greek words, kákistos (worst) and krátos (rule). Oxford’s 2024 word, ‘brain rot’ (mind-numbing scrolling) just missed the list.
The most recent words reflect what we spend far too much time looking at, clicking on, reacting to, or pretending not to enjoy while enjoying it a lot. Oxford’s choice of ‘rage bait’ basically means outrage as a media product, “click here to get angry about something you can’t change.” For me the definite winner was ‘AI slop’, the endless grey mush of AI-generated content pouring into our feeds as low-quality outputs produced with minimal or no human effort. The digital marketing firm, Graphite recently published a study showing that more than 50% of articles on the web are generated by AI.
We worry that using AI can cause dumbing down, or a decline in our ability to think critically. The fear is that if students start relying on AI tools, they will never properly develop basic skills in critical thinking and problem solving. According to a recent study by scientists at MIT, this might be true. Over a four-month study, they observed clear differences in how adults engaged cognitively when writing essays with and without AI. Participants who relied on AI tools showed consistently lower brain activity, weaker recall of their own writing, and a reduced sense of ownership of their writing compared with those using search engines or working without help. When the AI users were later required to write unaided, their performance and engagement failed to recover to the levels of the brain-only group. The researchers concluded that extended dependence on AI erodes independent thinking and sustained mental effort. I’m glad there was no AI when I was a student.
When calculators first came out in the 1970s, the solution was to make exams much harder. Instead of doing long calculations by hand (remember the slide rule), students were expected to use calculators and apply their mental energy to more complex problems. The bar was raised so that they had to work at least as hard as before. The problem with AI is that the bar has not been raised. Students are often given the same assignments and held to the same standards as they were five years ago, but they have more help. AI can indeed be harmful in this case.
On the other hand, just as calculators allowed us to tackle problems that were previously out of reach, AI can support new, more demanding problems. The solution is not to avoid using LLMs, they have huge advantages to offer. Instead, people need to become smarter users by understanding how they can further their goals rather than being harmed. If you need a quick answer to a question, use your favourite AI copilot; but if you want to develop deeper knowledge in an area, relying on LLMs alone is not the answer.
Our article by Gavin Halse on ‘The AI skills every engineer now needs’ illustrates this beautifully. He says that to use AI responsibly and effectively, you need at least a basic understanding of how they work and where they can fail. Generative AI then becomes a powerful assistant. Without it, you are effectively ‘prompting and praying’, exposing yourself to errors and professional embarrassment.
Some of the best advice I‘ve seen for students and knowledge workers comes from reteuro.uk.za. Basically you need to maintain your cognitive ability by thinking first. Pause and jot down ideas or questions. Start with a statement, test it through AI with counter-arguments, and explain what you’ve learned. Rewrite AI suggestions in your own words and test yourself later. Keep a log of AI suggestions you accept or reject. This reinforces the process, not just the final work. Essentially if you want to use AI effectively, run a ‘thought sandwich’. Think for three minutes, prompt for two, think again for three without the screen, then write. This will keep your brain engaged while allowing you to take advantage of the power of AI.
| Tel: | +27 11 543 5800 |
| Email: | [email protected] |
| www: | www.technews.co.za |
| Articles: | More information and articles about Technews Publishing |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version