3 Reasons why Artificial Intelligence is not a Technology

Artificial Intelligence Blog

Definitely maybe. The term “Artificial Intelligence” droolingly excites us and also draws some mystical connotations to its usage. Nevertheless, “Artificial Intelligence”, as a terminology, does not do justice to the word “Technology” in the current context. It’s not about the semantics of the word, but how the word is used (and misused).

The reason why I shy away from classifying AI as a technology is the lack of consensus on the specifics. I would look for something more concrete before calling it a Technology. I want to specifically look at the following 3 areas.

1. Definition

Scientists and Engineers are exceedingly smug about using the right terms in the right context supported by documented terminologies and definitions. And then we love to have lengthy debates about the exceptions 🙂

Typical example: Should chatbots be classified as AI? Is interactive A2P messaging also AI then?

The very fact that we can have lengthy discussions on the above indicates that such terms are loosely defined and include conjectures. Loose definitions are definitely not becoming of any technology. Yes, we have attempted to define AI and are verbalizing the categories of AI. But again those terms do not clearly define the boundaries of what AI includes, but instead sounds more like vision statements.

2. Value

A technology needs to sustain itself by creating a distinguishing value which is typically done by solving a problem or fulfilling a need. Unfortunately, only an unflattering few have been able to create value of AI beyond programmatic automation. And that’s not enough of a distinguishing factor.

Sample this: Should Robotics Process Automation projects be classified as AI projects?

A detailed discussion on this topic is here; but as always we are free to pick our answers under the context of the question specifics. The more specific we are in defining AI, the more objective would be the value from it; and too broad a definition will liquidate both the purpose and value of AI.

3. Community

For a technology to grow and touch the horizons of universality, it is usually supported by dedicated communities who painstakingly standardize the processes and the inter-workability.

The very fabric of internet was built on the foundations of standardization. The mobile telecom revolution was incubated by industry-backed standardization committees (even though many companies chose to ignore or tweak the standards).

Specifications like AutoML are definitely initiatives in the right direction for AI, even though I would have preferred more congruity in its application. Facebook and Microsoft are also working on how Neural Networks can exchange information. Nevertheless, AI is mostly driven in silos, seeded by academicians/researchers and taken up by the technology firms but still lacking a strong middle incubation ground. We need an IETF-analogous task force for AI!


The good part about us that we are comfortable with the evolution of terminologies. A couple of decades ago, AI was pseudo-synonymous with Neural Networks. Now, the umbrella of AI has expanded to include much more specific elements such as Neuro-Linguistic Programming , Machine Learning, Speech Recognition, Computer Vision etc. We can continually expand the umbrella of AI, but let’s just consciously create a distinguishing value of AI.

Leave a Reply

Your email address will not be published. Required fields are marked *