Human Like Machines (HLM)

Good thing about puzzles is that all the required information to understand it completely is available. If we treat life as a puzzle, this point becomes invalid. There is information asymmetry. We don't have all the information. Not even the critical information that is required. If we evaluate life in these terms, it become a more interesting situation.

Dearth of information makes life more interesting. How? Let me explain. If we were to have have access to all the information required, then life is solved for us. This means that a finite set of rules would be enough to tackle any thing related to life. This makes the emulation of human brain in the neural networks successful and complete. This in easier terms will mean that creation of a machine that can think like humans is possible.

Because we can take decisions in the world of information scarcity, and this can't be done by a machine, it is impossible to create a human-replica machine.

Another line of thought could be that even when we don't have access to complete information, we are able to decide based on some heuristics and some rules of thumb, which are rules again. So, it is possible to create a rule-based human-like machine (HLM).

This can be countered with the fact that the depths of these rules of thumbs are unfathomable. Even if they are finite, their application is not fixed. Even if it is fixed, it is mutable, very unlike the characteristics of machines, where the rules to large extent are immutable.

Considering these limitations, machines that emulate humans could be created, but the information that needs to be made explicit is largely tacit, the representation of this information increases the complexity.

There is a trend in terms of identifying the meta rules, the rules that are the basis for the actual rules, this will make the rules more abstract and hence decrease the size of the rules. There is an increasing effort in these directions of exploration for creating HLMs.

Though there are some breakthroughs in creating HLMs, they are largely restrictive and in no measure equal to humans. But with in a given field of expertise, are very handy.

Some day, it would be possible to have a world where the Stephen Spielberg's AI would become a reality. But that is only for future generations to witness.