Can you program Human Intelligence?

How far is Science-Fiction from Reality?

A question humanity has been asking itself for centuries. Fiction as a matter of fact is mere imagination whereas innovation is any imagination that could be produced using technology. So now the million-dollar question —

Can we replicate Human Intelligence?

Well, the answer to that question is not a simple yes or a no. That’s right it’s not impossible for technology to stretch out as far as being able to perform complex tasks as well as humans, but that’s something we all are aware of, what’s the more astonishing story is that we can build systems that have the capacity to multi-task with the same efficiency as humans.

As an added bonus one can even introduce Emotional Quotient into the system for life-like performance.

Before we dig deeper into the HOW, let’s try and understand the WHAT?

What is Conscience?

Before I put forth my perspective.

WARNING! This is a debatable topic and everyone is open to their own version.

In layman’s language Conscience is the reason for every individual’s actions. Let’s say we were to alter the same statement in terms of science.

Conscience is a highly evolved intelligence system which has the ability to comprehend, reason, evolve, learn, analyse and respond to the stimuli in its environment.

Now let’s read the definition again and focus on the words —
Comprehend
Reason
Evolve
Learns
Analyse
Respond

We know that today’s technology of deep learning has very high accuracy and immense capabilities to perform tasks such as those mentioned above. Then how do we differentiate between AI and Pure intelligence?

That brings us to our next question —

What is Artificial Intelligence?

There are thousands of articles floating around in space answering this very question. So without wasting much time on detailing what it is. At the very basic we can say —

It is a primitive synthesised form of intelligence.

Let’s just say that this is the primordial organism upon which the natural factors act to instigate the process of Evolution to transform into Conscience or Human Intelligence.

That brings us to our next and most crucial question —

How does Evolution happen?

In general, evolution is a by-product of environmental changes leading to alterations in a particular species or organism. Now in the same manner years of research have evolved the field of AI to such an extent where imagination is the only limit.

But what if I want to synthesise Imagination? Now here’s where things start getting interesting. Based on our discussion above, we know what are the essential components for programming Conscience. Let’s take a deeper look into each one of them.

1. Comprehend

The system should be able to cater to a vast variety of inbound data types. Such as Sound, Images, Videos, Numbers. Once the system can accept, now it is upon the intelligence system to differentiate between the variety of feeds and extract relevant information and store it within its memory in the most optimised manner for future reference.

2. Learn

Now that the system has stored all the relevant information received from the environment, it’s important for the intelligence to be able to identify and detect the patterns occurring across the massive volumes of data.
The system needs to re-iterate and validate itself over the data repeatedly if the learned patterns and variations are stable across all variants, or has it made a mistake in understanding the trivial features of the data.

3. Analyse

Now that the system has learned the identifying features of the inbound data. It is time for it to be able to construct models that can distinctively differentiate as well as create a relational mapping between the highlighted features and patterns of the data.

The points mentioned up until now are easily achievable by the existing frameworks. The twist in the story begins when the next steps begin.

Credits: Aleksandr Plikhta

4. Reason

The system was trained till date to act based on a specific input pattern. This is the stage where the system needs to take the next leap.

Reasoning is a very important segment of our brains. Humans are considered the most evolved species for various reasons, one of the most crucial of them is — Intuitive reasoning and actions. Our brains have the capacity to predict future actions based on the actions perceived till the moment from our senses. Not only that our brains also compute the different possibilities that can arise from several parallel permutations of action sequences.

This action performed by our brains is also called Reasoning. Let’s now talk about how we can implement something so highly complex and evolved into our primitive intelligence system.

One of the groundbreaking algorithms in the field of AI is called “Time Series Modelling”. It has the capacity to predict actions far into the future based on sequences of actions received till date. Now if we were to expand the implications of this concept to cater to all kinds of data types and models, then at some high level we would be able to replicate the functionality of Reasoning.

Another feature that we did speak about is parallelly computing the various possibilities of actions based on intuitive reasoning of previous actions or the current course of action. Now to replicate such a highly evolved intelligence you would be wondering there would be some groundbreaking topics.

To our surprise, we can resolve this issue by introducing a small noise into the input feed and alter the idealistic or real-time information to simultaneously compute several possibilities and most importantly make the system “intuitive” to forthcoming actions.

5. Evolve

Taking a look back at our system, it has the capability to learn patterns, comprehend a variety of inputs and make intuitive predictions.

But the game has just begun

The bigger challenge now is to cater to multi-tasks and not just parallelly processing a single problem statement but also to process multiple varieties of problem statement.

We humans have the ability to classify, detect, predict and be intuitive of the proceedings and that too not just for a single task but multiple tasks at the same time. Let’s take an example to understand this better — When we see a car on the road, we perceive motion, we identify not just the car but multiple elements in our vision, as it is an action in live motion we tend to process and predict the next time frame, say for argument’s sake an accident or hit and run case, at the same time our brain performs other tasks as well of identifying the number plate, warning the surroundings, taking precautionary measures.

Now according to our above example, there were a lot of possible problem statements that arose from a single story and each one of them requires its own niche model to resolve its own edge cases. But it’s not always that we have abundant data at our disposal. After all, we humans also learn by processing if not more but at least Petabytes of information yearly and learning from the same.

In order to resolve such a problem, we can use Incremental + Reinforcement learning to teach the system with practically no data. It’s as good as teaching a child from scratch but at lightning-fast speeds. Now that you have resolved your data issue, let’s talk about handling multiple models. For different kinds of tasks, we create different types of niche models, but what if we could create relations or interdependencies between the same.

Just as our brain has dynamic action items defined based on intuitive stimuli perceived with relational mapping, we can leverage the Knowledge based Graph systems to create dependencies and relations across niche models to perform tasks dynamically based intuitive predictions and feedback. received.

This way the system would grow with time.

6. Respond

Now that our system has evolved into a highly intelligent being, all that’s left is to get the responses from the system and interact with the same.

Conclusion:

All in all based on our discussion, we have come to the conclusion that building and imitating human intelligence is not impossible anymore. The catch still remains is if one can create the perfect amalgamation of the technologies mentioned above. After all, we all know that “Sophia” is not Science-Fiction anymore.

I hope this article triggers the creativity within you to become crafty with your ideas and create amazing technologies. 😁