Last week at a supermarket check-out, the young operator floundered when the woman in front of me wanted to pay for her groceries partly in cash from the shower of coins in her bag, and the balance from a card. The simple maths necessary to calculate the resulting amount on the card was beyond her.
She was a seemingly bright, and very friendly young woman, making a few dollars of spending money with a part-time job after school. However, without the cash register telling her the change to be given, she was drowning, embarrassed, and floundering.
It led to me thinking about the value of the computer between our ears which does need to be trained to do the simple stuff. In times past this was a core of what we absorbed at school, but now is the standard turf of data processing from a simple computation to advanced AI tasks.
If we do not train for the simple stuff, how can we critically analyse any proposition we face from the simplest calculation to the life changing choices we all face at some point?
Any data processing tool from the simplest calculator introduced in the sixties to todays advanced AI models does not have that uniquely human capability to apply intentional, cognitive power to a choice or activity.
We are able to reason without the necessity of a large database, we can apply experience and context to a situation and find patterns and connections that no machine can (yet) identify, so long as the foundations of critical thinking are in place.
Those foundations are eroding rapidly as we increasingly rely on machines to do our ‘thinking’ for us.
As I use AI tools, the opportunity to ask ‘what if’ questions is a fundamental part of the ‘thinking’ process much touted, rightly, as an enormous benefit. However, the answers to that ‘what if’ question are entirely dependent on someone having asked the question before, and the ability of the machine to surface the question, and the context in which it was asked, to deliver an answer.
Often the answer is incomplete, misleading, and occasionally fictitious. Partly this is because the question was not asked in a sufficiently clear and unambiguous manner for the machine to understand the intent. Secondarily, it is because the machine simply cannot think the way we humans do.
Delegating the repetitive and mundane parts of the work we do to a machine will increase our productivity, so long as we spend the time released to do stuff the machine cannot. If we stick our feet up and spend the time down a digital entertainment rabbit hole, the cognitive erosion displayed by the young lass on the checkout will eventually rot our cognitive capability beyond the point where it can be repaired.
