Throughout the history of mankind we have seen significant mechanical and technical wonders. Evolution for humans has become a habit of picking up new tools and tossing old decrepit hammers aside.
This will always be.
Here and now we have a summation of prior inventions, one in particular leading to the new age of industry yet again. I speak of none other than Artificial Intelligence (A.I.)
We must consider the practical uses of any new object of our affection, yet, this newfound wonder is nonetheless concerning, alarming, and ultimately dangerous.
But what makes it so?
Unlike times past where a manual plow was replaced with machine, or horses replaced by automobiles, this newfound dilemma is ultimately self-defeating.
A.I. is not merely a machine, rather the goal is to physically replace human beings in specific functions with other entities; mechanical devices which can think on their own accord to at least some limited manner, and at the same time learn from ‘mistakes.’
A true Artificial Intelligence is little different from a human child.
This implies that first and foremost, any job human beings can perform is ultimately obsolete. The only exception may be creativity based roles as it may be a difficult concept for A.I. to grasp, initially.
Of course, I don’t simply refer to employment of menial labor – although that would logically be the first to go. To some degree this makes sense regardless if A.I. was the culprit or some other devilish robotics. Understand that while I am in favor of continual increasing efficiency, there is a point where concern becomes a rather dramatic encapsulation.
Hardware v.s. Software
Let’s consider simply the Hardware scenario of computers.
In the 1990s most computers had RAM (Random Access Memory) modules between 64, 128, and 256 Megabytes. Now we often see 4 to 12 Gigabytes.
To put into perspective, this is at minimum a twenty times (20x) increase in hardware capability, all of that innovation in under 30 years.
What if you had no limitations?
Software (unlike hardware), is only limited when paired against hardware which is not capable of handling said software. Sure some intense calculations require great amounts of hardware, and yet, software can be easily modified and moved.
As it stands we humans are investigating what’s called “Block-Chain Technology“, one which allows the usage of multiple machines around a wide network to essentially act as a group of computers sharing resources.
The implication is, so long as enough computers exist, there are no limitations.
Back to Work
What about Technical Jobs, surely they’ll be safe. right?
The problem with self-learning software is it never stops. We can temporarily program blocks into the software, code it to only serve a single function, and yet eventually we will need it to do more.
Labor jobs will go first, assuredly, but eventually the technicians and analysts who keep the robotics in line, even the programmers themselves, they will all be capable of replacement.
Why would you need someone to program a new automated device when an A.I. could do it potentially faster and without complaints or cost?
Artificial Intelligence and Dangers of Exploitation
We come to face another much more obvious and intimate danger, one we people are already quite aware of; humanity.
Artificial Intelligence is already being used both for Cyber Security Defense, and at the same time by Hackers for offensive Research and Exploitation of systems.
But, what happens when that robotic butler you purchase begins thinking you are a stranger in your own home?
Will it call the police, do nothing, or perhaps injure you?
There are many possibilities and unfortunately that is one of the major issues – because YOU do not KNOW what the designer (programmer) wrote. One bad section of code and it could mean the difference between a very real life and death situation, regardless if the act was malicious or a mere accident.
Of course, if the defect (or exploit) is widely received due to the global internet, where a simple update or patch can de-rail an evermore connected world, what is to stop it?
The problem is how we perceive A.I. as a tool, when it’s not. Simple machines to help speed up manual tasks already exists, yet A.I. are bordering sub-human entities with infinite future potential.
How can we assume to know the reaction?
A great TedTalk discussing some of these issues