The four steps of the Robot Revolution — Infographic
“YES — Robots and algorithms will also take over YOUR job!” How did we arrive at this point? And why is (some of) it your fault?
I recently spend a lot of time reading about the fourth industrial revolution, browsing through countless headlines like the one above. Most of the articles focus on the negative unemployment impacts, but don’t really tell the full story. I asked myself: “How is this robot revolution actually happening?”
This article and infographic is what I came up with. If you are curious to know more about the technological developments leading up to the robot revolution keep on reading below the graphic.
How is the robot revolution happening?
Today, robots are taking over some of the last industrial jobs left for humans like — no joke — stacking boxes of varying sizes. And yes, we all carry some responsibility for this.
Remember playing Nintendo Wii Tennis in 2007?
Do not deny it!
Indeed, one could argue that the success of the Nintendo Wii led to the developent of advanced sensors, giving road to the next generation of industrial robots and many more technological developments.
The (R)evolution of sensors
Playing virtual tennis with a Nintendo Wii controllers can be much like playing real tennis. Forehand, backhand, volley. You swing your arm and your virtual self also swings an arm, ideally hitting the virtual tennis ball. To make this happen, many sensors needed to be included in the Wii and — for the first time — mass-produced. Among them was the accelerometer, a sensor which now enables your smartphone to switch between landscape and portrait mode (and some other stuff).
Under huge competitive pressure from Nintendos success, Microsoft decided to take a risk. Instead of copying the Nintendo Design (like the failed Sony Playstation Move) they developed the revolutionary Microsoft Kinect.
The Kinect is able to track people in real 3D space. A sensor-technology worth tens of thousands of dollars pre-2010. Microsoft priced it at ~$150. Researchers and the science community quickly made use of the new technology. A friend of mine actually abandoned his theoretical master thesis in robotics and rebuilt it around “virtual assistance in the hospital operating room”, using a modified Microsoft Kinect.
Today, robots use this technology to “see in real 3D space” an ability previously only available to humans. This enables them to take over jobs previously performed by human workers.
New sensors gathering new data and enabling new technologies have been developed relentlessly. Sensor companies gathered $4.3 billion dollars of investment from 2006–2016.
This progress is what gives your smarthphone, the size of a pack of cards, the power to command an array of advanced sensors.
Internet of Things
This (r)evolution of sensors has enabled the production of many more “smart” machines and appliances → the internet of things.
Fitness trackers, NEST thermostats, WiFi-speakers, Elevators which can automatically request a maintenance check-up if needed, automated factory lines…
They are already here: Millions of internet-connected machines with sensor-technologies to collect information. And billions more are coming.
Digital Lives
Our lives have gone digital. You might say “”But I have my hotmail account since 1996 — how can you say this is happening only now?”. I know it seems like ages ago, but the first smartphone only came out in 2007. (Apple iPhone, and no, blackberry doesn’t count).
In less than ten years we went from zero smartphone users to 2.1 billion in 2016. That is 2.1 billion miniature computers we suddenly carry with us through the world.
Digitization is also happening in the service sector. Huge corporates are completing massive upgrades of their IT infrastructure and business processes. It took them some time, but maybe even your insurance company has a useful app today.
Why is this important for the robot revolution? I argue this: True digitization is just happening now. Whether thing or living person — digital data is now being created everywhere.
Big Data — Digital World
In fact, our digital world creates never before seen amounts of data. WE, humans are creating never before seen amounts of diverse data. In one minute (that is 60 seconds) 6.944,444 snapchat videos are watched. Instagram users like 2,430,555 posts. The weather channel receives 13,888,889 forecast requests and Siri answers 99,206 questions. ONE Minute!
The collection of unstructured and diverse data is called Big Data.
Until recently, we didn’t have the power to analyze such unstructured sets of Big Data. But that has also changed.
More processing power
Computers have been getting exponentially faster and faster at processing information since the 1970's. Today everybody is able to access thousands of high-tech computers across the globe using services from Amazon, Google or other providers. Humanity can combine processing power of never before seen scale.
With this raw processing power we are able to analyze Big Data in new ways. For example we are now able to use methods for which we were lacking the power before. They are called machine learning.
Machine Learning
In Machine Learning a software algorithm first works through training data with known solutions. After training the machines are then able to solve unknown similar problems.
Due to Big Data more and more training data is available. This is why potential applications of this technology are increasing every day.
One example: language.
Traditional text processors can correct your grammar and writing. For this purpose smart developers wrote grammar rules and dictionaries into the software.
But no child learns language through grammar rules. First most of us just look at the world, our parents and other people talking (training data). After ~18 months children just start speaking. Children listen and learn the rules automatically from the context.
Machine learning works the same way. The algorithms learn by analyzing the “hidden” information and rules in their training data. Then they figure out the rules by themselves. This is how Google create their translation engine.
When datasets are large enough, the knowledge in all that data will often trump the efforts of even the best programmers. — Martin Ford
One step further: Automation
So here we are. We create Big Data through our digital world. This data can be analyzed by smart algorithms through machine learning. Through these processes machines can learn how to do specific tasks on their own.
One machine learning one task is just a small step. Like the Facebook-algorithms which can tell whether it is your face on a picture. Our brains can do countless tasks like that.
But machines are able to learn a lot of tasks. And combine these tasks.
The Nintendo Wii triggered developments which eventually led to machine vision for robots. In a similar way we are now teaching machines more and more complex tasks, like driving.
Researchers agree that “repetitive and routine” tasks will be replaced first by machines. This process is not restricted to blue collar industry jobs like fast food workers. Cozy office desk jobs are under pressure as well. Like accounting.
Accounting is clearly repetitive and routine. There are millions of balance sheets and trillions of entries from accounting systems available as training sets. Large sets of accounting tasks can and will be automated.
But there are also examples for algorithm journalists, military, delivery services and even computer musicians. And these are just from the last weeks.
Given these developments it seems clear that machines will take over more and more responsibilites and jobs in the future. Will a full-fledged “artificial intelligence” take over whole job categories in the next years? Maybe for some, not likely for others. However, with Big Data availability increasing for all domains of life and industry, machines will get better and better.
What next?
How should we deal with this robot revolution? Will humanity come up with new jobs to replace the lost ones? Will we live a life of abundance or of unemployed unhappiness? What are options countries and governments can take to mitigate the effects of these developments?
I want to tackle some of these questions in my next article. If you have any thoughts, ideas or specific interests let me know in the comments.
Thanks for reading!
Philipp
Let me know some of your thoughts in the comments, I am curious to hear what you are thinking. Also click the little heart ♥ if you liked the article/graphic. Thank you :)