Software-defined everything (SDE), the Internet of Things (IoT) with its developments from previous years to present, and the artificial intelligence (AI) boost in 2018 when companies prioritized budget for investments in machine learning, deep learning and chatbots show that their growth across various industries could indicate a quicker penetration in the mainstream by the end of 2019.
How can IT handle the penetration of AI, as so far it becomes obvious that cannot be done by IT alone and other departments would have to contribute to lay the foundation of what is the major task of blending artificial intelligence in our lives.
Since the IoT is growing exponentially the implementation of AI seems as the next step to take. Without artificial intelligence, IoT would not be able to handle the massive number of connected devices to their networks, and the risks of using them securely soar. When the communication and interaction between the internet nodes become sluggish, it has errors and even stops there comes the opportunity for major change. It’s named AI which implies automation of processes, natural language processing (NLP), introduction and adequate adoption of robots.
The power of AI goes beyond the limits we know from using our raw knowledge accumulated through history or in combination with any other device created by humanity and we want to grow as a society, the economy and the social structures, therefore advancing this technology seems obvious.
The issue with creating rules – to what extent we define, state and apply them, what are the rules determining the execution of functions of AI-enabled tools? Are they influencing results? The creation of rules seems beneficial when we know beforehand what we want to achieve, what outcome we define qualitatively which might extend to the coverage of applications of a specific tool.
Besides creating rules humankind have to make a choice and decide for what type of AI the results and outcomes can unfold freely based on its algorithms and for what type it becomes fundamentally restricted.
Even for AI-enabled tools, their creators would have to apply rules across the applications, as specific checkpoints in development. At the border of cognitive sciences and artificial intelligence, specific rules can be determined given the interferences between different fields such as organizational behavior, philosophy, health sciences, biology, neurology sciences. At every development checkpoint, specialists will be able to cooperate and create and validate these rules according to the criteria they chose together.
At first stage, we would advance the creation of AI-based tools for which results if not known at least would be perceived or guessed and we can design the AI to unfold and perform its tasks using all included algorithms. Humans would not delegate the power of decision to the AI-driven tools. When such AI is created, let it perform to the point of decision making. How to define decisions and how do we know which party makes the decision? Can there be decisions which AI-enabled tools can perform? Most likely, such decisions will exist and their results unravel within a framework. Decision making by AI-driven tool emerges as part of the design and the system acknowledges the necessity for the decision. Then the party as designed to make the decision intervenes, the status of the system changes and the operations are handed over to machines for processing until the next decision appears.
The monitoring and acting upon decisions and other aspects of the AI development create AI-driven tools which empower humans to conduct business with far superior results while managing the process of using the AI tools with astute awareness.
Author: Cory Popescu