Why do we need to learn programming when AI will do the job for us in 42 years’ time?
Let me break this down into two separate questions:
Q. 1 Will AI make programmers redundant?
Q. 2 Why teach coding to children when AI will do the job?
These are very different questions with very different answers.
Let’s start with the first one. The answer is yes or probably.
I include probably as an option because it is not yet clear (not to me anyway) whether the full-blown AI-driven world some believe possible (and others fear) will actually materialise. But this post is not about AI per se so I’ll move on.
Programmers will be made redundant by the full-featured version of AI, yes. As will doctors, accountants, lawyers, CEOs and every single profession. Yes, including yours.
## Programmers will lose their jobs. You will too. Who will go first?
Assuming we accept the premise that this is where we will reach at some point in the future, let’s break this down into which professions will succumb to AI first. AI is about having an algorithm perform human-like decisions. The more rule-based those decisions are, the easier it is for AI to do the job.
Taxi and bus drivers are the ones who should be retraining into a new profession yesterday. The technology in this case is already there and fully functioning. Only the regulatory aspect has delayed self-driving cars and kept them in their jobs. Driving from A to B is entirely rules-based: you pick a route, follow the rules set out in the highway code, hit the brake pedal if a child steps in front of the car etc… I’m not going to discuss the ethical dilemma relating to this issue here, although that also shows that even in the most straightforward conversions from human to AI-led professions there are still significant hurdles.
Next in line are accountants, lawyers and doctors. These professions are very rules-based at heart although human experience and interpretation is key to such professions. If given a choice we’d rather see an experienced consultant rather than a junior doctor on their first day at work even though their rules-based knowledge should be identical. But experience becomes irrelevant when computers are doing these jobs since they will be able to rely on millions of case studies that can be processed in less than a second making them millions of times better doctors / lawyers / accountants than human ones.
Life’s too short to go on listing every profession I can think of so I’ll finish this mini list with company directors / CEOs. We have now moved a few steps further away from the 100% rules-based end of the spectrum we started with. The challenges here are larger, but in our utopic (or is it dystopic?) vision, not insurmountable.
So where does programming fit on this scale. Many who view programming from the outside believe it to be the rules-based activity par exellence. But this is a misconception. It’s only the computer-facing side of programming that is rules-based because our computers, currently lacking the I from AI, cannot cope with anything else. A programmer starts with a problem that needs solving and crafts a solution using the tools that a programming language provides him or her with. That aspect of programming - the creative problem-solving side - is at the same time the hardest and most rewarding part of programming. That pushes programming closer to the CEO-end of the scale, in my view.
## On to the next question: why bother teaching coding to children then?
Our smart phones and computers do amazing mathematical calculations, so let’s stop teaching Maths to children. Auto-correct in word processors make the teaching of spelling and grammar redundant, let’s scrap English too from schools then. Google has all the information we will ever need on biology and physics, say, no need to learn them at school.
You get the gist.
More seriously, programming should be taught to children not so that they can become programmers in the future (we don’t expect them all to become professional mathematicians but we still want them to learn Maths). As mentioned above programming is a problem-solving exercise, probably one of the best types of logical problem-solving exercises there is. Therefore what children are developing through this subject is how to think, how to find different solutions to solve a problem and how to decide which of those solutions to use. This is a skill that humans, whatever they choose to do with their time once AI has taken over their jobs, will always need.
## Post Scriptum
I didn’t mention teachers in my rather brutal you-will-lose-your-job section above. This is an interesting one because although many aspects of teaching can be taking over by AI (it is fairly rules-based), the teacher’s role goes beyond that. Children are learning how to become adults (human adults, to be pedantic) and schools of course have a role to play in that. I suspect we will always want a human to fill that role.
Assuming schools will still be around, that is. It seems none of the subjects we teach our children will be necessary for them, according to some. That would be a brave new world.
About the author: Stephen Gruppetta is not quite sure what his profession is. He is a scientist (physicist) by training (fairly AI-safe) and is currently a managing director — he doesn't quite like using the term CEO (reasonably AI-safe for now). As part of his job he teaches children — programming, of course, as if you hadn't realised that from the views expressed above (completely AI-proof, see post scriptum). He is not worried about AI taking over his jobs anytime soon.
>>> Convinced? You can now go ahead and book a course for your daughter or son then…