Research Suggests AI May Be Uncontrollable in the Future, Raising Concerns About Its Potential Impact on Humanity

According to a new research, it was found out that there is no surety that AI can be controlled in the future. One of the researchers in the study says that as AI cannot be controlled, the developers should stop developing such systems. Dr. Roman V. Yampolskiy, a researcher in this study, says that many AI experts admit that AI is going to create problems for mankind in the future but they don’t explain this in detail. They don’t properly tell us how AI is going to be destructive, what its impacts will be and how we can remain safe from its harmful effects.

In his book called AI: Unexplainable, Unpredictable, Uncontrollable, Dr. Roman V. explains that AI will be responsible for changing our society in the future, and not always for the good. Many AI experts, including him, believe that AI is going to be the biggest problem mankind will ever face in this world. There is no saying what the outcome of AI control over the world will bring. Most probably, it will be the extinction of human beings. He says that he has read much scientific literature related to AI’s effects on humans, and there is not one proof that tells us that AI can be controlled. Many researchers believe that they can control AI but there is no solid confirmation for this.

Dr. Dr. Roman V. Yampolskiy says that humans can never fully control an intelligence model after making it. AI models or softwares are always going to create risk factors for the people and it is important to develop a solution before it’s too late. The problem that is in the way of making AI models safe and controlled is that AI is capable of making decisions, irrespective of if they are useful or useless, and this creates a lot of safety issues for mankind. AI also cannot tell its thot process behind the decisions it has made. So, if we cannot understand its decisions, we are unable to detect the problem and find a solution to it. For instance, many AI models are able to make decisions on health, finance, investing and employment. But they should be able to answer how they came up with those decisions in order to know if they are biased or not.

Photo: Digital Information World - AIgen

Read next: Emerging Cybersecurity Trends You Should Be Aware Of

Previous Post Next Post