# StevenClark.com.au

## Algorithms are not Evil Incarnate

One of my favourite units in the Bachelor of Computing was a core second year unit called KXA201 Algorithms; I got a 76 (Distinction). One thing that really stoked me about it was the hardcore geeky Maths/Physics computer guys struggled. Someone like me, a strength in good writing, found it a breeze. Anyway, listening to this weeks chaotic journalism hype about algorithms being some sort of evil incarnate, it’s no wonder I’m compelled to explain why they are certainly not evil, of themselves.

To be clear, I’ve studied undergraduate Computing with an average score of 80 (High Distinction) and a Master of Business Administration with a Specialisation in Journalism and Media Studies with an average score of 80 (High Distinction). You can doubt me, I don’t care. But here’s my feedback to journalists… because journalists don’t study anything like algorithms, or even statistics, economics, or fucking anything really. Except journalism. Most journalists, in a pinch, are talking/writing through their naive arses.

So here goes. An algorithm is a recipe. Nothing more, nothing less. Your cookbook is full of algorithms to make food. It’s a simple enough idea that I won’t belabour the point. Thus, algorithms are not evil incarnate. We use algorithms in our everyday lives to solve these types of problems – how to make a beef casserole; how to build an IKEA cupboard from flatpack. When you enter an elevator there’s a detailed algorithm prescribing the order of operations – the floor the elevator stops at next and when to open and close those doors; what happens when buttons are pressed. Algorithms are part of our lives.

In fact, you need algorithms in any space where you want to explicitly program how to achieve the result in a problem domain – it’s called imperative programming. There are also declarative programming languages where the programs know what to accomplish, but have to figure out how to accomplish the task in that domain space. So, anything where the program is designed in steps (uses an algorithm to achieve the result) is considered imperative programming.

Here’s the problem. Big corporate like Google and Facebook and a gazillion others have manipulated us over the years using algorithms. Because algorithms, recipes, allow them to figure us out; they sell our information or use it in-house to customise and increase revenue. They get caught over and over again and for that reason we don’t trust them (or do we?). Anyway, that’s how we got this maligned idea that the mere suggestion we could solve a problem with an (God forbid) algorithm brings down the entire ceiling with outrage.

But the traffic lights run on algorithms and your communication systems incorporate algorithms and you wouldn’t have GPS without algorithms. So your outrage at algorithms being evil incarnate is based on a misunderstanding.

Algorithms are a part of our lives and will only get more powerful as we need to solve more complex problems. It’s a complex world. Yes, maybe it was a badly designed algorithm that people are complaining about. Alternatively, maybe the previous academic performance of their children before COVID-19 would be a great indicator of their potential achievement during this pandemic. All the time keeping in mind that many students would have slacked off at home, rather than study. And therefore achieved lesser grades because they simply haven’t done the work.

Given my pessimistic view on human nature, I’d suggest that for students who didn’t do all the work in their home space in recent months the idea of being marked on past performance would be a win for the little guy.

So there it is… an algorithm is the recipe that makes complicated stuff work. The alternative to those types of problems would be to embrace chaos. I’m not sure what the parents in the UK are saying about the grades of their progeny, but if computer algorithms can’t solve the scoring problem then let’s hear the answer… I’m waiting. Tick, tock, tick, tock. No, because it’s all a beat-up in journalism to sell you papers. Go hug an algorithm.