Artificial Intelligence - The most happening technology of the era and some misconceptions surrounding it. Let us examine what AI actually defines - the myths and the facts.
AI is interconnected with terms like machine Learning, deep learning, and intelligent machines and understanding them is confusing. Let us navigate into these terms and debunk the myths associated with it.
Myth #1: AI means intelligent robots
Fact: Robotics and AI are two different domains serving different purposes. Robots are devices controlled by actuators and sensors to perform a wide range of tasks, such as building, carrying or dismantling products in factories.
AI is a software solution that has been programmed in a way that it is capable of making decisions based on its activity and also to learn from its mistakes. Although some robots may eventually be enhanced by AI algorithms, the "intelligence" part is just one additional ability AI may possess.
Myth #2: AI, machine learning and deep learning are one and the same
Fact: They are all parts of AI but they're three different things. Basically, machine learning is the method through which AI learns from different outside sources, used to differentiate data. Deep learning is another aspect of machine learning. It is based on neural networks used to inform AI about the probability is of making the right decision.
Myth #3: AI learns on its own
Fact: There is no AI-powered system which has application in the real world that can learn from nothing without human intervention. Systems that have to deal with hidden information or probabilities of any kind cannot be interpreted by AI tools. They require to be fed with data by humans.
Myth #4: Chatbots are a form of AI
Fact: Chatbots only make use of some forms of AI. They are programs that interact with humans through an interface or medium that can be voice or text. Bots are pre-programmed, and they would require several technologies that would allow it to understand human behavior. For instance a voice or text recognition software, analytic tools etc..
Myth #5: The power needed for future AI is unsustainable
Fact: AI requires a lot of additional computing power to be trained to perform tasks. With demand for AI in enterprises, the deep learning process will grow exponentially making it almost unsustainable, requiring a huge amount of energy. So with companies ending up buying more electricity, AI can be of aid in power generation by replacing old grids with new AI-powered microgrids that can distribute electricity more efficiently.
Myth #6: Immense amounts of data are required to train AI
Fact: The data required to train an AI depends on the type of application needed to be performed by AI. Certain formerly trained neural networks need only be retrained for specific functionalities or areas. However, there is a basic data structure that is fed with some general data with only the endpoints required to be fed for that specific use-case.
Myth #7: AI to replace existing business models
Fact: AI is integrated into the existing business models and so they need not have to replace the existing Business Intelligence solutions. BI solutions are highly adaptable and customizable. Besides companies would not look to change their business models for fear of workflow. AI can be integrated into the system through the web at different stages into the existing processes hence a replacement is not required.
So it might be exaggerating to say so.
Myth #8: Neural networks work like biological networks
Fact: Despite many kinds of research, we are not able to bring AI anywhere near to the functioning of the biological neural network. The neurons in the body perform more than just transmitting signals. There are different types of neurons that have different paths and use different biological techniques to transmit messages. Whereas machines intake data in the form of bits.
Myth #9: Beware!!! AI will eventually become intelligent enough to be hazardous to humans
Fact: But no. Simply put, AI has reached nowhere near in acquiring the intelligence needed to understand what is happening around it so as make decisions. Each algorithm concentrates on a particular task and does not go beyond that and because of that, it cannot relate itself to other tasks to make independent decisions. They have to use ultra super computational powers to find solutions to simple issues. It is nearly impossible to gain an understanding beyond the perception level of the program and so you can rest assured that AI will only remain to help us and not ruling us.Published: Dec 12,2018 11:34:02 AM IST, Updated: Dec 12,2018 11:38:25 AM IST