Autonomous Intelligence; the machines are coming for your job – Part 2

In the last article, we were talking about how AI is an inevitability for almost every industry. There is no getting around the fact that this disruptive technology is going to alter the way we operate our businesses. While there are many jobs that will remain untouched longer than others, there are a few that are just about out of time.

We could be like the companies that ignored technological evolution… Ahem, Blockbuster, or we can learn from them. It comes down to planning, preparation, and being fluid when the tide changes.

I’ve laid out a few things to think about as you decide if and how you want to be ready for whichever subtle, or not so subtle, change is coming your way.

Ahead of the curve

There is no real telling when the robot overlords are taking over your business, but I think I have time before SkyNet sends the job terminator. We start by accepting that technology is altering every facet of our lives, and unless a giant solar flare fries our laptops, we’re not going backwards. AI and automation will invariably shift the way your business operates. Whether it be by moving employees from one skillset to another, by insourcing something, or by employing a bot to love on your Facebook Messenger.

We can look to the major companies and how they are reacting to the shifting tech. They have much more capital to invest in ideas which they believe will pay off. For instance, Mercedes is in the autonomous truck market and Volvo is supplying cars for autonomous Uber. Of course, they can also afford to write off more loss. Caveat Emptor.

Leaders within an industry serve to act as markers for coming changes which will trickle down and, ultimately, reorganize your business. It is only in our best interests to stay abreast of these advancements and investments.

If you’re feeling particularly innovative, you can use your expertise to become that leader. You could design an automation process which would be worth sharing (at a profit?). Being first to the market is excellent, but being second often means being able to do it better.

Watch for gimmicks

A healthy amount of skepticism and due diligence goes a long way for any sales claim. This is no different.

It’s easy to get sucked into the idea that AI and automation will just solve every problem your business has. Just plug this program into your network and you are going to be 35x more efficient.

The reality is without data, an AI is not going to be able to make better decisions than your people. Possibly worse. What we are looking for is machine-learning that can learn and adapt faster and more efficiently than our puny human brains can.

It may take some time to research and even learn a whole new lingo; machine learning, API, cognitive. What do all those mean? Heck if I know, but they sound smart, don’t they?

I did find this list from to help work through the Snake Oil Sales for AI systems.

Let data be your guide

There is only one driving force in deciding if and how you can use an AI system; how good is your data. It’s where you start and where you end.

Without a good set of data to work with, there is no metric by which you are able to determine if an AI system is yet capable of improving a process. Once you start recorded information and tracking your processes, you can start to analyse that data.

Start with your bottlenecks. A bottleneck is just like it sounds, the point in your process that is restricted for some reason or another. The trick now is to gather as much data about this bottleneck as possible. Importantly, why is it bottlenecked and could it be done better if a machine were to do, if a machine can do it?

Which part of your business flow is the slowest? This is good advice, regardless of how you want to fix it. The point here, though, is to find the places where you’re going to be outpaced by the early adopters or innovators.

One of the easiest examples showing data-derived benefits is how UPS used algorithms to determine the best delivery routes. According to a 2013 Wired article, an average of 120 deliveries a day, per driver, means the possible delivery routes “far exceeds the age of the Earth in nanoseconds.” I don’t even know what that means, but it sounds like a lot.

What I do understand, is that UPS saves $30 million per year if each driver has one less mile to drive each day. That’s a number I can get my head around.

Set hard targets with your data. What do you have, can you make it better, and by how much? There’s a good chance that an AI system may be better at predicting seasonal inventory management than a person, especially if that program is connected to other data which could tip it off to external factors. For instance, a fire in a chemical factory in Germany could mean a slowdown in automotive production in Canada. Connecting the dots requires a lot of data and machines can process it faster than humans.

What happens to the people?

Until our computers learn to program themselves, we are going to need people grind through the day-to-day application coding. There is this mythology of the basement-dwelling, loner coders. The reality is it’s turned into what some have described as another blue-collar job.

As part as your on-going commitment to growth and training, you may find value in converting your labourers into coders. Train people to code – most applications don’t need the creative algorithms and deep university knowledge, they just need to drive like the basics of an engine. There is two-fold value to this.

The first is maintaining the investment in the employee and their retention. You have already spent the resources to hire and keep that person. Offering further training and opportunity allows them to grow with the company.

Secondly, your people already understand your processes. While you may have to invest in the infrastructure to allow for coding, that money will be offset by the cost of training and explaining the processes to outsourced companies. Or not. You will have to do the math

Final Thoughts

We live in interesting times, do we not?

The tumultuous nature of change can be difficult. It is stressful and uncertain. It is possible that there are few times in history when humanity is facing a complete shuffle of our daily lives as this. I will leave that to the historians to argue.

For all the possibilities, we have more information than anytime before. This gives us such an advantage. We have the opportunity to benefit and profit if we take the time to plan and hedge our bets. And definitely hedge your bets. Being fluid and able to deviate from your plans when the next technology ripple comes out of nowhere and derails everything.

What are your plans to capitalise on the coming changes? Are you ready to make waves or do you think being a stalwart of old-school is your best game plan?

Let me know what you think!

Thank you,
Paul Bergsma