A lot has been written about Artificial Intelligence (AI) and how it is changing the world. Many thinkers have already raised alerts about the disruptive potential of AI. Elon Musk has recently warned that AI could result in World War III. Stephen Hawking has said that AI could spell the end of the human race. Bill Gates said recently, we should be concerned about AI and the day when the machines will become super-intelligent and could outsmart the humans. At the same time, there is another camp that doesn’t subscribe to the views of the doomsayers.
AI is going to be here, for sure!
It has been a part of our lives for years in various forms. For example, the brightness of visual displays gets auto-adjusted based on ambient light conditions; auto-schedulers carry out a set of predetermined tasks on a variety of machines; when macros are invoked, a set of tasks get done by robots. Siri, Google Now and Cortana have been able to understand our speech and carry out a few tasks based on our voice-based instructions. Video Games have been using AI for quite some time. Based on the behaviour of a player, the computer decides its move. Self-driving cars use tonnes of AI already. Recommendations of music, movies, books, clothes and so on are already getting automated.
Large enterprises have started using AI-based tools while procuring goods and services. The tool helps them carry out e-auctions and help choosing the best suppliers. Chat-bots for customer service have been commonplace. Tasks such as fraud detection, credit scoring and approving loans are getting codified and the machines ae learning fast to make the decision rather than interventions by a human.
How far can we go?
Computer-assisted surgery, drone-based surveillance, accessing places which are hard to reach or risky for living beings have been possible due to robots and intelligent machines. It wouldn’t have been possible for a human being to carry out such tasks with the speed and accuracy that a robot shows; moreover, the mission becomes possible with no risk to human life. How far can we go?
Seeing and moving things have become pretty much possible by robots. It is a matter of time that the machines which require an operator to see and move a few things mechanically, will run by its own intelligence.
Engineers and Scientists are building intelligence in a machine around the other senses of a human being such as hearing, speaking, tasting and smelling. While this research will continue for a few more years, we see Personal Assistants with improved capabilities to interpret speech, speak a few words and make recommendations.
Trying too hard!
Human emotions and behaviours have so far been difficult for fellow humans to understand fully. Hence, it will take time to teach a machine, for sure. However, some data scientists have been over-zealous and based on a half-baked world view, they build models for a machine to take actions.
Machine-recommended actions such as watching a TV show, listening to a song, watching a Youtube video, buying a pair of shoes, reading a piece of news or liking a friend’s post on Facebook could be risk-free. However, let us think of a situation where one builds a model to predict criminal behaviour based on partial understanding of motives behind a crime. Such a system could prompt the police to start surveillance and an enquiry. Will this be helpful? May be, may not be! One can similarly estimate the damage potential if such half-baked information gets into an AI system which is evaluating exam papers, interpreting statements recorded by witnesses in a police investigation, writing the prescription for a patient, emergency response to fire alarms and so on.
AI is our reality, however we need to use it for the right purposes. Most importantly, we must develop a thorough understanding of all possible scenarios connected to an event and build the intelligence in the machine accordingly. Else, the fear of the doomsayers is not far away!
You may also like to read: Workforce trends: What to expect in 2018