AI shows intelligent behavior

The most powerful entity that would make a huge difference in economy in the 21st century is artificial intelligence (AI). The emergence of AI has changed a lot of equations in terms of replacement of low-skilled jobs. The more advancement would result in its entry in virtually every sector and more jobs will be cut. The workers in few sectors such as automobile, BPO, and manufacturing have been hit by the application of AI. A machine or a robot has replaced their jobs and producing more output than they could. The unemployment would impact the economy of the world. However, tech giants have been adopting a strategy of retraining workers to acquire better skills to work in collaboration with AI systems. Though there is a wave of worry among people that AI would replace humans in near future, it seems the future is so far when humans will be replaced by machines.

There is still a scope for vast improvement for AI to replace humans. Though it can recognize human speech, drive cars, and beat humans in games such as GO, it does not have one important characteristics of humans: common sense. AI system shows an intelligent behavior by training itself on enormous amount of data. It was not possible a decade ago, because there was no sufficient amount of data to train a system. However, the availability of data by now has made it possible. The AI system analyzes inputs, check background data, and gives output. Then it trains itself and figures out the most intelligent behavior to give output.

But showing intelligent behavior is not same as being intelligent. If a user asks the system to book a table at specific restaurant, it will identify speech and take necessary action. On the other hand, if a user asks the system to book a table for four at the McMahon Clinic, it will still try. It will not apply a common sense that one cannot book a table at a doctor’s clinic. There are limits to AI as it checks what is on the background and what it is trained for. It will check the number of McMahon Clinic and call to book the table, which it is trained for. Even though machine learning algorithms could automate few tasks by humans, there are limitations to their capabilities.

There is a possibility of computers which can rationalize and take decisions accordingly. Brains of human beings are biological; but giving the same capabilities to a machine is possible. MIT physics professor Max Tegmark in his book, Life 3.0: Being Human in the Age of Artificial Intelligence, suggested that it is just a matter of time when computers will have such capabilities to rationalize. He argued that computers improve at an exponential rate. He also wishes that when such machines will arrive, they will be used safely, fairly, and wisely. The probability of computer intelligence surpassing human intelligence is definitely greater than zero.

Tegmark said in his book, that the near-term opportunities for AI to help humanity are “spectacular”. It is possible only if AI systems are made robust and unhackable. However, it is difficult to make systems that are not prone to hacking. But there is no denying that sooner or later, these machines will have a common sense same as humans with progress in technology.