Thinking has always been considered a valuable resource unique to humans. When the calculator eventually became ubiquitous, there was some resistance to the device, but it gradually became acceptable with widespread use. The initial fear was that we would lose the skill of mental arithmetic. Although there is nothing definitive regarding this loss (good or bad), there is evidence that long-term calculator use in primary school can actually improve mental computational skills1. So perhaps that fear was unfounded.
That brings us to AI and the question: what will we gain, what will we lose? The most prevalent threat is to our ability to become knowledgeable – to have a good head on our shoulders. To think.
There are three ways to think about thinking:
- Reasoning (or the relationist lens): Einstein didn’t need to build an atomic bomb to know that e=mc2.
- Experience (empiricism): We learn not to touch a hot stove.
- Meaning-making (constructivism): We watch a movie and then construct a review. We get differing reviews because we construct meaning from different cultural contexts and life experiences.
AI can do a pretty good job of all three, but not in the way we do. The reason AI consumes so much water is because it generates a lot of heat. The reason it generates so much heat is because of the way it ‘reasons’. All AI does is figure out what the best next word would be. A sort of data beauty contest. And that’s a lot of computation – more than any of us could imagine. When AI provides a movie review – it does so without context and emotion - it merely looks for patterns in other reviews and predicts which word is most likely to come next.
That creates a conundrum: we get some amazing output from AI, but it uses a method that is less sophisticated than a toddler who could figure out that the round object goes in the round hole. The criticism of AI’s ‘thinking skills’ is that it has no real awareness, it just simulates reasoning; it has no sensory experience – it just recognises patterns. So, if it produces meaningful output without understanding, it is neither sentient nor knowledgeable – it merely processes and accesses information which it organises into a form that we recognise as knowledge. Having access to that knowledge depends on you possessing a certain skill. In a word: prompting.
As a manager and leader, you are there to provide direction. You are expected to know. CEO’s make decisions that requires orientation and then animation – knowing where you are, where you need to be, and then taking steps to get there.
That still remains the case, but now CEO’s need to transition from pure decisionmakers to AI-enabled strategists. The CEO who asks AI for its contribution to their AGM address by prompting it to review the Annual Report, gets a perspective that’s way richer than the solitary pursuit of just ‘putting pen to paper’. We can’t risk delivering a speech that is flagged as AI, but it can make an enormous difference to how informed we can be in putting the address together. Prompts can be whatever you want. For example:
Prepare an outline for a 20-minute speech based on the attached annual report.
Also review the Annual Reports of Company A, B and C in this sector, and point out how we are different. Also refer to the industry benchmark report that can be found here.
Find a selection of three quotes that I can use to start the presentation – they need to relate to my theme for the year which is productivity.
Give me a choice of three tasteful anecdotes for each of the key points in the outline.
There will be a Chinese delegation present as guests. Give me a sentence in Mandarin that I can use to welcome them. They are male baby boomers.
The staff have been challenged with an explosion in our Middleburg plant where one person was killed and three injured. Ensure there is nothing in the outline that could lead me to show insensitivity to this.
The delivery of the speech will encompass the informal speaking style of Sir Ken Robinson and the statesmanlike tone of Martin Luther King. I am also a fan of the ironic humour of Herman Charles Bosman.
Propose a strong close.
Keep the outline to less than 500 words.
AI can help enormously with any address – even sermons. If you are a religious leader, for an example of prompting for sermons, watch this video here (actually this would apply to all leaders who address an audience).
In both academia and commerce, attention to AI has been ‘significant’ with an emphasis that blends data-driven technologies and human augmentation for effective business solutions
2. So, not only do we have a calculator that spits out a number, but we also have a calculative machine that generates an answer based on deep complexity. And all it does is hold data beauty contests. Lots of them. And we all win.
References
- Stacey, K. and Groves, S. 1994. Calculators in Primary Mathematics. Paper presented at the Research Presession of the 72nd Annual Meeting of the National Council of Teachers of Mathematics. https://files.eric.ed.gov/fulltext/ED373968.pdf [2 March 2025].
- Osei-Mensah, B., Asiamah, E.O. and Sackey, R. 2023. Strategic communication and artificial intelligence: Reviewing emerging innovations and future directions. Archives of Business Research, 11(1):85-102.