Despite knowledge of human-like automation, artificial intelligence still has some shortcomings.
The concept of artificial intelligence began with the concept of making machines work like humans. In the current, artificial intelligence has improved to the point that artificial intelligence machines and robots can paint, write poems, and easily do many things that man can do. Recently, artificial intelligence technology has been able to solve an important problem with the concept of protein folding, with which researchers had worked in vain for a long time. So it can be said that artificial intelligence is in the midst of people and machines. However, despite its expertise in human-like automation, artificial intelligence is still reaching out to people with simple but vital aspects.
While artificial intelligence robots and machines are capable of solving difficult problems in mathematics, physics, and technology, they are often unable to solve some simple problems. For example, ‘Rani went to the store and chose a red dress. He paid 500 rupees. – There are no direct indications in this statement that Rani has bought a red dress. A person would easily understand it because he has background knowledge that when someone chooses an item and then pays the money, it means that the person has bought the item. However, the artificial intelligence machine cannot automatically draw conclusions that Rani is buying a red dress. It has no background information that the combined act of selecting and paying money refers to a purchase. This background information can be called common sense. Common sense grows within man through experience and practice to preserve this experience, for which artificial intelligence is still incapable.
Dave Gunning of DARPA has said in an interview with Forbes: “Lack of reason prevents an intelligent system from understanding the world, communicating naturally with people, behaving reasonably in unexpected situations, and learning from new experiences.”
Some might say that this common sense can be realized by downloading everyday databases to the machine. In 1984, such an operation was launched under the name Cyc. The basic problem with this function is that shared knowledge has its own set of exceptions and different ideas that would be impossible to put into a machine.
As a human child grows up, he learns new things and gradually adapts to the environment. The AI machine is made by feeding loads data sets and once deployed, it no longer has the ability to learn, which prevents the machine from achieving adaptability. Machines equipped with artificial intelligence are unable to simultaneously learn from the environment and automatically adapt to new changes in knowledge. Machines can be updated from time to time. The ability to learn continuously over time, taking into account new knowledge while retaining previously learned experiences is called continuous or lifelong learning.. Such an ongoing learning task has been a long-term challenge to neural networks and thus to the development of artificial intelligence.
If Artificial intelligence-based robot sent to an unknown environment, its human creator may have no idea what the robot might encounter at that location. So a human developer would not be able to enter any database on this into the robot. As a result, the robot is unable to adapt automatically to the environment of an unknown location. It cannot produce immediate ideas or instincts that are different from a person who can always adapt to unfamiliar environments by adding his or her newly acquired knowledge to previously preserved ones.
Artificial intelligence is also unable to connect between cause and effect, which prevents it from understanding the basic dynamics of the world. An artificial intelligence robot can always perform a controlled action, but it cannot internally understand the reason for the action or what it may have an impact in the future. For example, according to the instructions in the morning machine may serve breakfast, but it may not understand the connection of breakfast to the morning. Reason reasoning is an integral part of the human intellect that shapes how we understand and interact with the world. We know that dropping a vase causes it to break, drinking coffee makes us feel energized, and regular exercise makes us healthier.
Brenden Lake of New York University said: “Our minds build causal models and use these models to answer arbitrary questions, while the best artificial intelligence systems are far from imitating these qualities. “
Researchers are constantly striving to turn artificial intelligence into human-like qualities and functions. Robots like Sofia provide clues to the pace of improvement and progress so that machines can fix these shortcomings in the future as well. The concrete aspect in this situation is that being a bridge between humans and machines, artificial intelligence works very quickly to catch up with intelligence very soon.
Share this article