Artificial Ignorance – The Downside of Machine Learning

The man ahead of me at the airport boarding line looked nervous. The airline agent was holding up his passport, face high, as her glance switched from his face to the passport, and back to his face. She appeared to be doing a side-by-side comparison to ensure that he was who he said he was.

The screen of the newly installed face recognition kiosk had flashed red as he stepped on to the designated spot on the floor and stared into the machine’s camera. I wondered if the machine not recognizing him was a good thing or a bad thing. Either he was an infrequent flyer, or he had been flying under the radar — not part of any profiles already known to the system.

But, that point was moot. He was in the system now, forever.

As the agent, who seemed satisfied with his credentials, waved him through, she quipped half-jokingly, “The machine is still learning.”

She may not have realized it, but she had just hit the nail on its head. The machines are learning. And, learning fast. They are developing artificial intelligence, relegating humans to less-demanding roles.  

The downside of machine learning, human artificial ignorance!

There is no doubt that human intelligence is slowly giving way to artificial intelligence. Advancements in Facial recognition, Computer Vision, Natural Language Processing (NLP), and other such technologies all depend on machine learning. Simple identification of images and voice is no longer enough. Computers are being trained to emulate the human brain and learn how to interpret images and audio content based on context, similar to how children learn to differentiate a cat from a dog.

Until Artificial Neural Networks (ANN) that mimic the human brain and its nervous system came along, humans held the upper hand. What was simple for humans, was not so simple for computers. While a computer could use facial-recognition software to identify a person in a security line at an airport, emotion detection and human action detection were beyond its traditional capabilities. If a passenger was nervous and fidgety, it took a human to detect that.

Machine-learning-driven artificial intelligence is changing that.

Cashier-less stores, drones delivering pizzas, edge computing in manufacturing, computer vision in healthcare, robots that pick and pack, autonomous cars, digital assistants and other such transformative technologies are all about displacing human intelligence with artificial intelligence.

Unless you have a glass-is-half-full kind of attitude, you can easily slip into artificial ignorance. Why learn things that you don’t need to know? Learning to drive would be a waste of time; the future is all about autonomous cars, right?  

Humans are there to feed the machines with data so that they can learn. Big databases such as ImageNet that hold over fourteen million images have been created purely to train the machines to recognizes patterns based on the context and state of an object.

If your job is all about speed, volume, and repetition, you may want to start exploring options.

Heck! Even simple bloggers like me are slowly going the route of artificial ignorance.

Applications like Otter and devices such as Google Pixel 4 use NLP to provide free speech to text transcription capabilities that allow me to record and edit posts at my convenience. Artificial intelligence-powered Grammarly ensures that my posts are grammatically correct and include unique words and expressions while minimizing passive sounding sentences.

The joys of artificial ignorance!

Having said all that, there is a flip side of opportunities that machine learning and artificial intelligence bring.

At the high-end, there are the coders, the programmers, the engineers, and the scientists who are vital to developing and perfecting the technology. Then there are the businesses that leverage these technologies to transform and disrupt established models. And finally, there are the operators, the technicians, and other doers who learn to work with the machines and manage them on a day-to-day basis. There is no saying that you couldn’t be one of them.

As I stepped up to the face recognition kiosk, I held my breath. Have I been learned?

Sure enough, the screen flashed green. I was a known quantity.

As I settled into my seat on the plane, I tried to think of a job that machines can’t take away in the near future.

Enterprise Sales!

I felt smug.

I closed my eyes and prepared for lift-off. Pink Floyd came to mind:

Welcome, my son, welcome to the machine.

Where have you been?

It’s alright, we know where you’ve been…