AI and Automation, What's Next? A Take Over Or a Symbiosis?

by Prince CanumaAugust 3rd, 2020

Too Long; Didn't Read

Rethinking the future we want not the one that will befall us. We are in charge of our destiny.

featured image - AI and Automation, What's Next? A Take Over Or a Symbiosis?

Rethinking the future we want not the one that will befall us. We are in charge of our destiny.

Nowadays there are a lot of headlines saying things like: “AI to run a factory X or AI is going to replace human cooks in this country”.

I strongly believe that we should stop and think before automating away everything there is. In the wise words of Richard Koch:

“The road to hell is paved with the pursuit of volume. Business is wasteful, and because complexity and waste feed on each other, a simple business will always be better than a complex.”

‪Do not follow the masses by implementing the latest hot tech, because you can be shooting yourselves in the foot without knowing out of the fear of missing out.‬ Unless you are a giant in your industry like Google, Tesla or Facebook.

It’s enticing for businessman and managers to maximise profits — AI and Automation are words highly associated with those 2 words.

Instead of looking at ways to squeeze every penny from business, the biggest questions we should be asking are:

Are AI and Automation really the only best solution?How can we put humans-in-the-loop?

As we have seen not only current AI is not where many speculate but it also cannot decide on everything even though it can surpass humans in a few tasks and in narrow.

Furthermore, the worlds human population is only tending to grow larger(thank you India and China), this increased growth will imply that more jobs will be needed so it would be both wise and beneficial to find a way to create symbiotic systems where humans and AI work together and not competing against each other.

Although it’s out of the scope of this article I would like to mention that one possible way around is to find ways to implement ideas such as universal basic income.

Focusing on the wrong KPI (Key Performance Indicator)

Image credits

Businesses can endanger both their existence and the customers by focusing on the wrong KPI for their business and believing they can automate away every section of their business in order to squeeze out every drop of performance.

AI is still a fairly young technology although very famous it’s because of this spotlight on it that many companies are blindly adopting the technology that can be dangerous and deadly if not implemented in the right way. I’m not saying there is a risk-free technology, the risks are there and it’s understandable especially coming from industry-leading companies trying to push the boundaries of what’s possible instead of the many simply implementing the technology and hoping it improves your business overnight.

The goal should not be to replace humans with machines but to create a symbiotic system where both work in synergy — this is something big companies figured out a long time ago, while most companies are implementing AI to get rid of people and increasing their slice of the cake.

Let’s see some examples of how some companies are creating symbiotic AI systems using a framework called Humans-in-the-loop or Human-centered AI and thus tackling problems that troubled their industries for years.

Aviation industry

Airbus’ Autonomous Taxi, Take-Off and Landing(ATTOL) Project

Image credits to Airbus

On the 16th of January of 2020 Airbus published an article entitled “Is Autonomy the future of aerial mobility?”, where they describe the new system that they testing dubbed ATTOL(Autonomous Taxi, Take-Off and Landing) although still under development and testing, it is a quite interesting system, it does what the name suggests, it can take-off and land on its own, using only minimal input from the pilot.

Without going into the details, this is a system is powered by powerful on-board hardware that processes data from cameras and sensors around the plane and leverages developments in Computer Vision to do tasks such landing strip detection, lane detection, plane alignment and etc.

Compared to the previous approach ILS(Instrument Landing System) without a doubt this will allow pilots greater support and freedom to focus on the bigger mission and solving bigger problems, instead of being caught up in the technicalities of taking-off and landing. The pilot becomes the centre of operations.

But I believe that this system is far from ready to be deployed and used regularly because Computer Vision still faces great problems such as adversarial examples which cause them to make mistakes, this is not counting the case where we have suboptimal weather and visibility conditions. Because usually these systems are trained on data that is biased with optimal conditions, such as great lighting, weather and etc.

Instrument Landing System (ILS)

From my research, I found out that it is a ground-based landing system that requires a lot of hardware both on the ground and in the plane to allow both systems to interact and exchange data. It uses 2 antennas in landing strip(namely localizer and glideslope) to emit radio signals that are captured by the planes to give vertical and lateral position guidance of where the runway is and how to identify the centre. It is prone to have signal interference which can be caused by the signal from other airports nearby are emitting.

In my opinion, ILS is a fairly complex system, besides being very technically challenging to pilots it also needs too much hardware and software to function.

Well compared to ILS, ATTOL seems much better although it needs more rigorous research and testing under various conditions to avoid disasters when it finally debuts in a commercial flight.

Airbus’ Long-Term Vision

In their article, Airbus states that their goal is not to move ahead with autonomy as a target itself, which sounds and is great to hear if true, and I believe most companies can learn from them in this regard. Their main goal like many industry-leading companies is to address the challenges of tomorrow. But what are some of the challenges of tomorrow?

Air traffic management
Pilot shortage
Enhancing Future Operations

I was curious about why the pilot shortage was listed as one of the issues. So I looked into it, and found out that some of the reasons for the shortage are:

Increased flying hours of commercial pilots
Ageing pilot workforce
Fewer new pilots coming out of the military
And a general decline of interest in the career.

Life is full possibilities and there is no single outcome or solution to a problem. My question to you is: “Can we find other creative solutions to these issues besides the creation of completely autonomous vehicles?”

Auto industry

The auto industry is also moving towards automation.

According to this article, the auto industry plans to make 44% factories smart in the next 5 years and investments are set to increase by 60% in the next 3 years which will result in productivity gains estimated at $167 billion.

If we only focus only on the gains we might be enticed to even bet on this kind of investment blindly. This kind of investment must be accompanied by equal or greater investment in re-skilling, skill development and creation of talent pool.

There are some really great developments revolutionising many industries and the most successful implementations of cutting edge technologies such as Artificial Intelligence, robotics, automation and etc, are done in a symbiotic manner.

Conclusion

There 3 key takeaways:

The goal of business is not to replace humans with machines and vice-versa.
AI and Automation is not what makes your business successful, but it is the right implementation of these technologies done in a symbiotic manner that will contribute to it.
Investments in AI and automation must be accompanied by equal or greater investment in re-skilling, skill development and creation of talent pool.

Thank you very much for reading, you are really amazing. I do this for you.