Artificial Intelligence Evolution – An Interpretative Perspective!

While we may look at the evolution of machines and computing innocently enough; the fact is that the age of Artificial Intelligence is upon us. And whether we want it or not, it will eventually happen that machine intelligence would someday become better at everything. So let’s look at where we are headed.

Tim Urban, in his web series on AI, classifies AI into three broad categories – ANI (Artificial Narrow Intelligence), AGI (Artificial General Intelligence) and ASI (Artificial Super Intelligence).

Almost all of the AI we see around us today in our machines and devices – the so called smart devices – is based on a collected (and ongoing) data set and some kind of a machine learning algorithm. Imagine any natural language processing service like Google Assistant, Alexa, Cortona and Siri. They start with a base data set and keep learning over time to become smart enough to respond and perform tasks intelligently for you. These services have access to a huge number of data points that they can run deductive ‘reasoning’ decisions from. This is what constitutes ANI. It is the very basic Artificial Intelligence and has taken decades for us to get to. It is dependent almost entirely on humans to provide computing resources (processing speeds, storage, interconnects) for it to evolve as it does.

Over time however, as computing resources surpass the speeds of human brain processing speeds, a set of instructions would take a fraction of time to execute by machines. Artificial Intelligence would reach at least the intellectual capabilities of humans (sans some important factors though). This is the stage of Artificial General Intelligence. AI would be able to accept inputs like humans do, but be able to process them faster and more accurately. The evolution from ANI to AGI is expected to be slow and would take at least one more decade according to experts.

And then comes ASI. The transition from AGI to ASI is expected to be exponentially fast – maybe a couple of hours! This is where AI is not dependent on humans for the resources that it requires to function, evolve and grow. It would theoretically be able to command and create its own resources to quickly reach biblical proportions of intelligence. It would be able to think in ways that would take decades for a human brain to think in! It could assemble its own resources at a nano-level and produce them en-mass!

Scared, are you?

Some experts predict this age to be upon us as early as 2025. However, the general consensus is around the year 2050 for the first signs of ASI to be there. Doomsday theorists are already scared of the machines taking over the world and making humans obsolete.

Here is a thought though – Humans as a species, have evolved over millions of years to have cognitive brain abilities, intelligence, intellect, ethics, morality and emotions. As of now, there doesn’t seem to be a way to impart these into machines. AI can be given a task, a set of instructions and a definitive outcome to achieve. Whatever it deduces as tasks to execute to achieve its goal, may be detrimental to humans – but AI may not even comprehend it. Its interpretation of instructions is completely different from how the human brain interprets.

As far as I am concerned – I’m convinced we will have superior levels of Artificial Intelligence in my life time. AI would help us discover completely new materials to build things, repair human tissues and body, help us move to unforeseen frontiers in exploration, energy etc. AI would bring about impacts on nano as well as mega scales!

What is your view on how the future is going to be? Let us know in comments.

Follow us and our updates on: Twitter, Facebook & LinkedIn