Page 1 of 2 next >>

A New Age: AI and Machine Learning Meet the Cloud

The author Elbert Hubbard once said, “One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man.” These are words to take to heart as we think about the future of technology. Napoleon Bonaparte took the point a few steps further, expressing the view that 10,000 men are replaceable, but a single man is not. However, his story culminated in both generational disaster and a societal redefinition that lasted for centuries. We are precipitously close to entering an age of total automation which will include robotics in every phase of life, but hopefully we will avoid a new age of Bonaparte.

As children, our notion of robotics was that of “Rosie the Robot,” the benign robot-maid on the cartoon series The Jetsons. Yet, today many of us use the Roomba robotic vacuum cleaner in our homes. The Roomba cannot rearrange living room furniture to our liking, something not recognized by the science fiction writers of previous generations, but Roomba-like machines can perform an amazing array of arithmetically programed tasks. The most dramatic example is that of the modern self-driving car. You can wander through the futuristic streets of any Silicon Valley town late in the evening (or, for that matter, Pittsburgh or Seattle), only to be stunned when you encounter a small automobile with a “driver” so inattentive as to not even be looking in the direction of the steering wheel. “Data” companies, whose modern genius manifests as a series of “Tinkertoys” barely envisioned by the retro-generational science fiction writers, are hard at work and play at the same time, collecting data. Today, every major car manufacturer is working on self-driving cars. Shortly, common vehicles will be capable of transporting us to our next appointment, before we even remember we have that appointment scheduled.

The expense to design and manufacture robots is rapidly declining. Big data, artificial intelligence (AI), machine learning, and deep learning, coupled with componentized software and near-unlimited processing power, are enabling new classes of robotics never possible before. The graphical processing unit has evolved far from its humble video game beginnings into a system of such processing power that a microsecond is considered to be unacceptably high latency. Large fortunes are being generated as these companies are developing into entities that resemble the great figures and even empires of antiquity. When considering the technology industry as a whole, it is understandable if we wonder if we are witnessing the inception of a new colossus.

The is the glue that solidifies and connects all these new systems. Through the Internet of Things, a smart home can be controlled through a smartphone or even an intelligent device such as Amazon Alexa. In the next few years, our homes will be littered with so many intelligent appliances that even ever-logical Mr. Spock would raise an eyebrow.

Modern robots learn the way a child learns, through imitation. No longer must every action undertaken by a robot be pre-programmed. This is the wonder of ubiquitous and nearly infinite amounts of data, processed at speeds only limited by Einstein’s equations, stored in tiny arrays that a few decades ago would have been the size of planets, and analyzed by inference engines only partially discovered. This is the wonder of deep learning where AI is no longer science fiction. Instead, it is the focus of every serious venture capitalist in the free world. These trends are quickly becoming reality, and they serve as the intellectual fuel for the rapid expansion of robotics inside the home and out. Tomorrow, there will be a robot in your home reading your kids bedtime stories and helping to you to plan or prepare your next dinner. Rosie is a lot closer to reality than you think.

Containers, Containers, Containers

One of the major flaws in the culture of modern technology is the lack of imagination in naming conventions. The word “container” is one such unfortunate overuse of a term. The technology world has found many uses for the word “container,” but the most recent one describes a “logical encasement of a functional system defined to both partition and abstract the underlying operating system.” Strangely, this concept is similar to machine virtualization while simultaneously encompassing the opposite idea. If your company is not already using software containers, there is a good chance it will be in the foreseeable future. More simply stated, a container is a self-contained unit of application code that contains the minimum resources needed to accomplish its task. A container “contains” the code, runtime executables (Windows or Linux), system tools, and libraries required to execute a given task. Containers are the “worker bees” of the internet, doing their jobs while the high-profile application receives all the credit and notoriety. The most useful aspect of these self-contained worker bees is their portability across environments which circumvents the common problems associated with supporting software environments that are not identical.

Page 1 of 2 next >>


Subscribe to Big Data Quarterly E-Edition