--- title: "Teaching kids to think matters more than ever in the age of AI" description: "Learners are reaching for instant answers, raising concerns about the skills they may lose or never learn." type: "NewsArticle" publisher: "Daily Maverick" site: "https://www.dailymaverick.co.za" section: "PAUSE THE PROMPT" author: "Jean Greyling" author_url: "https://www.dailymaverick.co.za/author/jean-greyling-2/" canonical_url: "https://www.dailymaverick.co.za/article/2026-11-15-teaching-kids-to-think-matters-more-than-ever-in-the-age-of-ai/" published: "2026-11-15T08:38:31" lang: "en-ZA" word_count: 1059 --- # Teaching kids to think matters more than ever in the age of AI > Learners are reaching for instant answers, raising concerns about the skills they may lose or never learn. By Jean Greyling · Published 15 November 2026, 10:38 SAST ## Key points - Generative AI is changing education and software development, but human judgment and strong fundamentals remain crucial. - Employers warn that young developers are becoming too dependent on AI and losing sight of core coding principles. - The answer, argues the author, is teaching learners to “think, plan, solve” through computational thinking — not simply prompting chatbots. - Research and classroom experience suggest AI can support critical thinking when used intentionally, while coding tools can build confidence from the earliest grades. ## Content In January 2017, I drove onto campus for the first time since September 2016, after the extensive Fees Must Fall protests. I remember feeling a sense of hope after a very dark four months during which I often questioned the future of academia in South Africa. Little did I know what was waiting for me. Three weeks later, my BSc Honours in computer science student, Byron Batteson, walked into my office with an idea for his project: “Prof, I want to create an app that will teach learners coding concepts without the need for computers.” For the previous 20 years, I had been promoting our department and discipline in schools across the Eastern Cape. Many times I drove away wondering whether I had achieved anything by talking about software development in a school with no computers, and sometimes with limited electricity. So the TANKS coding app was born, later evolving into the Tangible movement, which now has a footprint across five continents. Back to 2017. In those days, I basically had two arguments for why corporates should invest in our coding project. The “Fourth Industrial Revolution” had resulted in huge software developer vacancies being reported in South Africa and many other countries. And then there was the planned coding and robotics curriculum in its infancy. Our mantra became: “Teachers need to prepare themselves and their learners for this curriculum by implementing our games.” The three curricula for Grades R to 9 were gazetted in 2024, but still need to be rolled out to all schools. ### **The revolution is now** “AI will be like the lava of Pompeii. It will roll over us and we will be buried in it.” – *The Devil Wears Prada 2*, 2026 This quote summarises the revolution we are experiencing right now. I do not need to expand on all that is being said about generative AI and its potential positive and negative impact on the world. I choose to write from personal experience and observation. The obvious question we, as Tangible, need to reflect on is whether what we do in schools is still relevant. If one listens to Elon Musk, Sam Altman and all the other AI gurus, the end of human software development is near. That the software development world is being disrupted cannot be disputed, but this is probably true for all “knowledge careers”, including medical doctors, engineers and accountants. However, feedback from employers of our students paints a reality far removed from the gurus’ predictions. Although AI tools are creating huge amounts of code, the human in the loop remains ­pivotal. One aspect mentioned is the need for good software engineers to quality-check AI-generated code for bugs. Furthermore, the market for software will never be saturated. As software becomes cheaper, more small companies and individuals become potential users. We are hearing from employers that young developers have become too dependent on AI tools and many lack the ability to understand core coding principles. Consequently, in our department, we have actively started to introduce our students to AI tools for software implementation from third year onwards, while the focus remains on the fundamentals of coding in the first two years. But a bigger need for what Tangible does now lies elsewhere. The “lava of Pompeii” hit me in August. Tangible called on teachers to submit recipes and win R500. Not because we wanted to compile a cookbook, but because a recipe is a great analogy for what a program is – “a list of instructions that tell a computer what to do”. The idea was that they should come up with creative recipes that we could use. More than 50% of the submissions were clearly AI-generated. It dawned on me that if even teachers were replacing “think, plan, solve” with “prompt”, what were they expecting from their learners? ### Practical steps neglected After noticing similar behaviour with his second-year students, a colleague remarked: “It is as if they think this is the only way to find a solution – it’s normal to them.” By “think, plan, solve”, I am referring to the basic components of computational thinking: decomposition (breaking a problem into smaller parts), abstraction (identifying what is important and what is not), pattern recognition (recognising similar problems and solutions), and algorithm (the steps to solve the problem). These are very practical steps for solving problems in all spheres of life. Given what I have described, it is my strong opinion that any tool that enhances computational thinking should be fully embraced by schools and teachers to lay the foundations of problem-solving. Various tools, organisations and materials, including the coding and robotics curricula, can assist with this. Simply prompting without a computational thinking skill set will set learners up for failure in the new world we have entered. I recently attended the World Conference on Computers in Education, where many papers grappled with this challenge. Dr Corna Olivier reported on her completed PhD at the University of Pretoria. She found that when chatbots only offer answers, they diminish opportunities for critical thinking and lower student engagement, which she refers to as cognitive surrender. She showed that when generative AI chatbots are designed intentionally, they can move students back from cognitive surrender to cognitive presence and critical thinking. As a movement, Tangible offers various coding games, national and global tangible coding tournaments and free unplugged lessons from Grade R upwards, aligned to the gazetted coding and robotics curricula. All enhance computational thinking. Dr Embeth van der Wal, a PhD graduate from the University of Pretoria, investigated the implementation of the foundation phase Tangible lessons. At the conference, she reported that implementing these lessons improved teachers’ self-efficacy in presenting coding and robotics in the classroom. The teachers’ feedback supports the idea that computational thinking can be taught well in the early grades. And this leaves me with a sense of 2017 déjà vu. Although I have serious concerns about the future of education, I have hope – if all educators manage to get the fundamentals of “think, plan, solve” in place. Only then will we have no need to fear “prompt”. **DM** *This story first appeared in our weekly DM168 newspaper, available countrywide for R39. The e-edition of Daily Maverick’s weekly newspaper, DM168, is now free for readers who are signed into their Daily Maverick account. Click*[*here*](https://www.dailymaverick.co.za/168/)*to access the latest e-edition.* ![Image](https://cdn.dailymaverick.co.za/i/4ur437KtO3lx5RpQJwXsYKawNxw=/200x100/smart/filters:strip_exif\(\)/file/attachments/3000/DM-13112026001CPTJHB_831834.jpg)