Robots are coming, but not for your job

machine_learningA week or so ago a very good friend of mine and I were talking on Skype.  One thing and the other and the conversation shifted to Artificial Intelligence (AI), machine learning and robots.  This morning I received from him an email message with a link to the article “Robots Are Coming, But Not For Your Job” published by Bloomberg by Carolynn Look from Bench Mark.

The photo in the article shows several kids toys in the shape of robots.  Over the photo there is a caption that reads “Industry association says machine labor will augment, not replace, human labor”.

As usual, please take a moment and read the original article to get the gist of it (

The article starts by making a reference to science-fiction.  That tends to set the tone for the entire article.  It follows by the statement that a study by the VDMA Robotics and Automation Association has concluded that robots are not taking away jobs from humans.  The purpose of the VDMA is to support companies that produce robots / automation.  Talk about personal interest and conflict.automation_robots

As always there are lies, damned lies, and statistics (from wiki: A phrase describing the persuasive power of numbers, particularly the use of statistics to bolster weak arguments).  The amount of jobs in the world has increased based on the increase in population.  The key is to find out if the new jobs are of similar or better quality than the original jobs replaced by automation.  Most important to society is if the original laborers were able / helped to make a transition in education and skills to work next to robots and / or provide support for the new technology.  The answer to such concerns does not seem to be so bright.

It is true that a person working with some type of automation will be able to bolster their productivity.  As a simple example take a look at word processing.  The robots in this case did do not take a human shape as is typically depicted.  A person in need of a typed document would dictate it to a person (i.e., secretary).  Depending on the length and complexity of the document it could be typed by the person who took the dictation or sent to a group of people that worked in the Word Processing department.  They would use word processors typically using custom hardware.  The document would be returned typed in double space for review by the author.  After all changes were made, the final document was printed in the proper stock, signed and shipped to the distribution list.  Today, almost the majority of documents are typed and delivered directly by the authors.  Secretaries and word processing centers are a thing of the past.  The people that where and could be employed is such positions were forced to learn other skills and not longer could consider such jobs.  The same happened to the buggy whips.  Not too many of those are currently being manufactured or used.  People that worked pumping gas and checking cars for oil and windshield fluid (at the time just plain water) had to look for different jobs.

machine_learningThe article contains a forecast by the WEF (World Economic Forum) regarding the estimated loss of about five million jobs by 2020 (less than four years from today).

Finally the article touches on China.  The Chinese government under lead of Xi Jinping decided a few years ago to gain control of the world in most (never say all) fronts.  China has the largest populations in the world.  One might expect that given such amount of laborers, China would be one of the last countries to embrace robotics.  That clearly is not the case.  They are using robots and are planning on increase their use to become more effective in manufacturing goods.

Please note that automation and machine learning represents the current state in AI.  Besides the social issues of retraining and educating workers, most (not all) people has been able to make the switch.  The less prepared have not.  Robotics has replaced what we call blue collar jobs.  Machine learning is moving as a threat to white collar workers.

There are many TED talks and YouTube videos about the state of the art in AI and neurology.  Sooner or later (discussing such topic could take several thousand pages if that) humans will be able to understand how the brain works.  At that point we need to make sure that the computer programs and robots that follow will always work on behalf of humans.

To get a taste of the current state and possible future of AI read “The Master Algorithm” by Pedro Domingos.  The book provides a good overview of what is going on and the future of AI and society.  I agree with most of the ideas in the latter part of the book.  History has shown us time and time that there are always people that believe rules do not apply to them and do something out of what is expected.  You can figure it out on your own.neurology_human_brain

I am very interested in machine learning and extremely eager expecting the discovery of the Master Algorithm (or whatever it may be called).  I have spent a considerable amount of my free time in the past few years learning and experimenting with machine learning.  I am a firm believer that the answers will come from neurology and then computer science will pick up on such discoveries.  Algorithms and programs need a model which we do not currently have for the human brain.

The Naïve American

thanks_emojiPS:  GGGGGGG (for anonymity reasons) thanks for forwarding me the Bloomberg article.

Be Sociable, Share!

Leave a comment

Your comment