Systems thinking in artificial intelligence

Dr. Rahul Remanan,
CEO, Moad Computer

 One of the common mistakes made by decision makers during the process of incorporating artificial intelligence into their enterprise, is attempting to view artificial intelligence as either good or bad. I want to explain here why this approach could be an indication of fundamental ignorance about artificial intelligence systems.

 The key premise of the field of artificial intelligence is to develop smarter computers that can better understand the world that surrounds us. Artificial intelligence is therefore an important technology component that is helping us develop better tools and systems. Like every other piece of technology, the field of artificial intelligence has its benefits and pit-falls.

 Asking the question of whether artificial intelligence is good or bad is akin to asking the question of whether chemical engineering is good or bad; or whether nuclear physics is good or bad. Like chemical engineering or nuclear physics, artificial intelligence is just a set of know-how and tools that can provide our society with advanced technological solutions.

The artificial intelligence solutions could be either highly beneficial or hugely detrimental to our society, depending on the nature of those solutions. The benefits and pit-falls of artificial intelligence are highly specific to a particular artificial intelligence solution. ​ No meaningful generalization about artificial intelligence can  be drawn from the nature of these individual artificial intelligence solutions. It is therefore impossible to answer the question of whether artificial intelligence in general is good or bad.

Also, the question of whether artificial intelligence is good or bad introduces a certain sense of, us-versus-them mentality. It is very unscientific and immensely counterproductive to focus solely on the artificial intelligence component of a solution, while evaluating the potential benefits and threats posed by specific artificial intelligence systems.

Instead of thinking about characterizing artificial intelligence as good or bad, it should be replaced by a more effective and pragmatic approach of thinking about how a specific artificial intelligence system is going to impact our societies. Therefore, what are the benefits and pit-falls of implementing that particular artificial intelligence solution, is a better question to ask.

To answer this question of benefits and pit-falls of specific artificial intelligence solutions, the systems thinking approach comes in extremely handy. A decision maker could use this systems thinking approach to individually evaluate the major components of an artificial intelligence solution. It will help understand the key performance characteristics and limitations of each of those components.

This breakdown of strengths and weaknesses of the components of an artificial intelligence system, will in-turn assist in  rational decisions about the overall benefits and risks associated with that particular system.

The systems thinking approach of evaluating an artificial intelligence system, help decision makers re-focus on the possible artificial intelligence solutions in its entirety, instead of trying to focus solely on the artificial intelligence component of that solution. It also prevents individuals and organizations from trying to make ill-informed and futile generalizations about the impact of artificial intelligence. This  system based thinking is not only helpful in methodical designing, testing, troubleshooting, updating and scaling a specific artificial intelligence solution at hand; but also, improve organizational and societal trust in that specific artificial intelligence solution.

TLDR: Attempting to view artificial intelligence as either good or bad is counterproductive. Instead, incorporating the systems thinking approach to evaluate specific artificial intelligence solutions, help decision makers re-focus on that particular solution in its entirety. This approach mitigates the erroneous focus solely on the artificial intelligence component of that solution. It will also prevent individuals and organizations from making futile generalizations about artificial intelligence. Wider adoption of this systems based thinking in artificial intelligence has an important role in improving business and societal acceptance of artificial intelligence.

If your next data-science or artificial intelligence project needs our assistance, feel free to fill-out our consultation intake form.