Artificial Intellect: Quantum nature of human interactions and sum of knowledge

The most difficult questions of the business do not have a singular answer. If it were then managerial remuneration would have touched its low point, large part of the managerial function would have been replaced by AI or its equivalent.

Think of the most difficult questions like new product launch and the strategy behind it, or market penetration, changing a cost position vis-a-vis industry, drawing the future cost curve of an industry, creating an outlier of an industry or take the more simple ones like selling more than the others or reducing cost faster than the others. None of this have simple answers and it involves intellect, algorithms that must be tested, trials and tribulations, but most importantly it must be driven by an urge to seek the most illogical of answers, actions that have never been attempted before with an entrepreneur's mindset to withstand failures repeatedly. To create an alternate hypothesis constantly is where the human dissatisfaction starts with a result, thus no answer would ever be complete.

AI on the other hand will seek logical answers to a puzzle, move through a maze of options and get the most accurate of solutions based on the algorithm we have designed for it to work through. AI would not seek a solution that would make a data point look like an outlier, or take an audacious attempt to genuinely fail in the first several attempts.

AI is very important in the areas where we are seeking a definitive answer to a variable set of data points and we need probabilistic output to be computed like average, median or mode or even run regressions, to help us with decision making. When this is automated and frills are added like prompting us to do something based on large sets of data points which is not humanely possible to be computed at every instant, we have a dilemma of sort created.

Think of it if I am a planner and I am prompted at the outbreak of my day what my most important task is based on a nightly run of all data points collected for the previous day? It is a great ready reckoner for me as I would not otherwise know what the data points actually cumulatively signify or what signals are apparent in the several sets of data with large complexities involved.

But to assume that these suggestions are cumulatively exhaustive would be very wrong. What if there is new information that is outside the system that is not tested and which if applied could dramatically change the result? This is where the intellect comes in which AI lacks, to be able to seek a solution where a new information is constantly being added and not to be satisfied with a logical solution.

AI would continue to be satisfied with a solution, the human mind would never be as it would try a new approach not tested and create an alternate hypothesis.

Roger Penrose and his two books, Emperor's New Mind and Shadows of the Mind, has dealt with these questions on AI and how an algorithm based approach could never be sufficient to replace the quantum capacity of the human mind which plays like a billiard board, picking up new information from every strike of a ball with the other and the total knowledge of the system changes once contact happens.

Think of the world of information today and how each interaction (billions happen every moment) between participants in the network changes the sum total of human knowledge, how could you possibly think of creating an algorithm to model the output of these interactions, which are of quantum nature?

New information being added and the processing of the same by the recipients and fresh set of interactions generated out of it is happening at exponential level. Each minute several years of data is being added and processed and millions interact with it now, which in the past was a passive mass of one way interaction with data.

AI is therefore extremely important to sieve through this maze and give us what could be a ready reckoner of sorts for us to deal with. But this should not end up like tail wagging the dog.

The human mind remains the dog, it cannot be driven by the tail of information crystallized by an AI paradigm. AI is an essential element otherwise the human mind would be only picking and choosing parts of the maze of data, but it is not sufficient recipe for us to act.

The human mind is like a an endless memory space, no matter how much we add, there is no question of a "memory full" prompt. When we try to create a system with limited attributes and characteristics and then provide a solution as the best alternative to the human mind, we must at least realize that we are not dealing with a finality of a solution but perhaps the next best.

Seeking the best solution is a matter of human intellect, it does not fall in the domain of AI, which is all about intelligence, not intellect. The intellect will always be on a journey, while intelligence is about reaching a destination.