Ooh Aah Optimisation

Before I started my blog, I thought about many names for it. One of those was World Optimisation. But that sounded a bit shit, so what I’ve done instead is to write about optimisation around the world, but named my blog, Mark’s Blog. It’s a blog written by someone called Mark, see? My creative genius is scary.

Last week I was partying on a yacht in Ibiza. Actually, that’s completely untrue - I was reading about artificial intelligence, robotics and algorithms. Who says I don’t know how to have fun? I was reading about them because, as you’ve probably noticed, a lot of capital has been pouring into AI.

Now, any successful business is in a continuing process of optimisation, and what we’re seeing with AI is the magic of computational rationality. Yes, the outcomes from AI are a step change, but it's a trend continuation. Operations research has been growing as a discipline ever since it was founded in World War II. The 1970s gave us the birth of the personal computer age, with software such as spreadsheets (hey, don’t knock them!) enabling more optimisation. The age of the internet in the 1990s led to the age of technology titans and the next wave of optimisation. So it’s not that new. Even artificial intelligence has been around since 1956 and has gone through many bleak winters.

Artificial intelligence, machine learning, and deep learning have been edging into the business world for years and years. You can go back to the 1990s and see the seeds of it. Renaissance Technologies, the ridiculously successful hedge fund, used quantitative and statistical analysis that we’d probably consider some form of ML/AI nowadays to analyse, determine and execute trades.

In 1999, in the last century, in the last millennium, Amazon hired a guy called Jeff Wilke. His remit was to improve Amazon’s internal processes - the endless business drive for efficiency, efficiency, efficiency. In 1999, Amazon started using the holy trinity of data, algorithms and technology to improve delivery speed, reduce defects, improve demand forecasting accuracy and improve customer recommendations. There’s no doubt, in my mind, that a large part of the success of Amazon is down to its commitment to science, experiments and optimisation to improve every little detail endlessly.

If the only tool you have is a hammer, you tend to see every problem as a nail. Well, yes. But if you have data, every problem is an optimisation problem. Unsurprisingly, Amazon has created a competitive moat where every little bit is optimised to the fullest. (I’m hoping now that I've said nice things about Amazon, Bezos will invite me onto his new $500 million superyacht).

You want another example? Here’s a Facebook job advert from November 2005! You’ll see the requirements include “experience with … machine learning.” 2005! What proportion of people could have explained machine learning in 2005? Talk about being ahead of the curve!

Oh, you want another one! Fine. In 1996 Larry Page created the somewhat eponymous PageRank algorithm, and in 1998 Google was founded. That changed the course of history, undoubtedly. Now Larry is enjoying his life on desert islands; he deserves to. But, back to my point, PageRank was effectively a Markov Chain. Now, sure, when I put my data scientist glasses on, I will tell you that this is absolutely not machine learning; it’s applied statistics. But whatever, it is optimisation.

To show it’s not all about Silicon Valley, take Toyota. Kanban, just-in-time manufacturing and Jidoka are all crucial components of the Toyota Production System and all are optimisations.

And there are a million more examples. You use websites like Skyscanner to book a flight, Apple Pay to pay for the tube, Expedia to book a holiday, Airbnb for somewhere to stay, Tesco for your groceries, Chiquita for your bananas and on and on…

There’s a link between these examples: the holy trinity of optimisation has created products, services and processes far beyond human precision and capability. Suppose the earth were to be transported back to the Dark Ages. We’d eventually return to where we are now via the industrial revolution, the internet economy, and endless optimisation (hopefully quicker and with fewer wars and pandemics).

The world of business and the world is being endlessly and relentlessly optimised, pushed and prodded into an ever improving place. Data, algorithms and technology. This is the holy trinity of optimisation. Though if you take a step into the business world, you’ll see there’s a whole lot more optimisation to be done.

And don’t forget, optimisation leads to productivity gains. It is no coincidence the more developed and industrialised, and therefore optimised, the country, the higher the output!

Thanks for reading, I’ll see you next time!