The Fourth Industrial Revolution is Here and Why Google is Getting it Right

Industrial revolutions are about harnessing the power of the day to drive a breakthrough in business or social conventions to produce a sea change in our everyday lives.

The First Industrial Revolution used water and steam power to drive industrial activity. We saw the rise of the locomotive that enabled rapid transport of people and goods. Suddenly commerce could span the once insurmountable mountain passes and open prairies that kept civilization in pockets. Mechanization of the farm also meant less people were needed in food production, freeing people to explore 

more academic topics. 
The Second used electricity to fuel the rise of the modern factory and around the clock productivity aided by multi-shift workers and the light bulb. This in turn drove further urbanization and accelerated moving people off an agrarian lifestyle and the rise of centralized cities and a suburban lifestyle.
The Third used microprocessors and information technology to automate production planning. This saw the rise of process efficiency. Former manual processes were now handled by computers freeing up people to again rise above these manual task e.g. ERP, Supply Chain and CRM efficiency, ATM’s, Self-checkout, ecommerce.

Now a Fourth Industrial Revolution is here.

This new revolution is happening because of a few things that have come together in recent years.

The rise of hyper scale cloud computing.  At the click of a mouse I can rent computing power that 10 years ago would have taken millions of dollars of investment, floor space, people and power to access.  The democratization of compute.

IOT – Sensors are everywhere. These sensors are continually exhausting unparalleled amounts of data. Whether it’s bio-informatics as simple as your Fitbit to genomic sequencing to predict your propensity for thyroid cancer and formulate a targeted therapeutic and improving outcomes that just a decade ago were trial and error guess work. Or IOT sensors on each stalk of corn in a farmer's field driving higher yield and eliminating crop loss from over/under watering, fertilizers and pests.

And lastly advanced algorithms in Machine Learning and Artificial Intelligence. In March of 2016, the world was stunned when Google DeepMind defeated a legendary Go champion Lee Se-Dol from South Korea.  Most AI experts thought that type of deep neural networking was a decade away. Go is a simple game and take only a few minutes to learn, but the possible combinations of moves are endless. The number of potential board positions is: (That big number over there) a number greater than the number of atoms in the universe. That's a lot of atoms.  Because there are so many different combinations, Go is the “Holy Grail” of AI.

Which brings me to why Google is getting it right.

Cloud computing takes scale.  Check

A hyper scale network free from the public internet.  With a private fiber network in 70 points of presence in 33 countries.  Check.

But AWS, Azure and even IBM and Oracle would argue “me too” on the above.

So why is Google poised to be the center of the Fourth Industrial Revolution? 

YouTube, Vision API, Video API ,Tensor Flow and the democratization of AI that's why.

For instance, every minute of every day, 300 hours of Video are uploaded to YouTube.  1.3 Billion people watch 5 Billion videos every day. (That’s democracy in action) When you search for “Dogs” on YouTube, you’re not just searching meta-data tagged manually to a video. YouTube uses the open source Tensor Flow library to find all images of a dog embedded in ANY video.  Even if the video has nothing to do with a dog or a dog just happen to stroll across a few frames of an hour long video about baseball.  But how did the algorithm know what a dog looks like? There’s lots of different dogs out there and coding a traditional algorithm to cover all dog conditions would be ridiculous.  Google Machine Learning and Google compute with its ability to process billions of references to hone the algorithm to statistically predict that image is a dog. That’s how.

So how does the ability to find a dog in a video spell a sea change for mankind? 
It’s not just about the videos or the images or the audio, it’s about combination of massive amounts of compute on demand, a super-fast network, the advanced ML algorithms and the ability to unlock immense amounts of intelligence from the unstructured and machine generated sources.  Couple this on top of the open source nature of the Tensor Flow libraries and you have the flywheel effect of Billions of people, data sources and compute engines. (More democracy in action)

Imagine next generation apps to predict fraud, money laundering, powering self-driving autonomous cities that optimize people’s needs to get somewhere with the capacity constraints of the freeways. But this time, everyone is humming along at 80 MPH with zero accidents.

Insurance claims being handled quickly, paid efficiently with minimal fraud because the corpus of accident video, image and report data has seen this exact scenarios 1000’s of times before.

Applications were once constrained by how fast a human could type or a printer could print. That’s all changed.  Say you want to learn how to replace the transmission in a delicate piece of engineering like your late model BMW. Wouldn’t you much rather learn by watching a video specifically customized for you with the aggregated relevant clips from the largest video library on the planet with each step by step procedure and pitfalls?  A video tailored perfectly to your experience level and perspective.  The alternative would be a dry manual written from one perspective and for a mass audience without regard to your needs and reference points. Life is good when the worlds data and compute resources come together.

The platforms have arrived and are getting more powerful by the minute.  The application possibilities are gated only by our creatively.