Artificially yours

Stepping into Artificial Intelligence (AI) guru Vijay Mukhi's Mumbai office, you might expect smart-suited robots to play ushers—you'd be disappointed. But spend an hour here and you'd realise that AI's potential lies in the high-precision analysis of existing data, making it innocuous and powerful.

"How old is this man?" Mukhi asks me, pointing at an image of himself on his giant desktop. Before I can produce my answer, Mukhi's already bid the AI to announce not just his age (59), but his height, build, complexion, ethnicity and even state of mind/emotion as portrayed in the photograph! Most of this seems to have been achieved by the method of elimination—for example, the AI first denounces Mukhi's possibility of being Caucasian and then African, narrowing it down to the possibility of Hispanic and Asian. At this stage, Indian, as an ethnicity, is too specific an identification. But let's backtrack a little. What is Artificial Intelligence?

Nature of the beast

Oxford Dictionary defines Artificial Intelligence as: the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. As far as Mukhi's arsenal of AIs go, they can do this, and plenty more. The software translates and detects tone/mood of text; analyses sentiment in text, images and videos; detects face, age and ethnicity of celebrities; compares and correlates a single face in separate images; detects song genre; translate text in 50 languages in a second, detects frame changes in videos; recommends fashion wear based on your software-registered sartorial history and even detects the composition of a lung vis-a-vis a CT scan image. In this impressive volley of skills, the only area where the AI faltered was the age and ethnicity of Indian celebrities—its efficiency muddled by the disparate body statistics of Hollywood celebrities stored in its database. Turns out, a vast, diverse range of data to infer from is the key to an AI's formidable powers.

"What I'm doing here is feeding a gigantic database of information – on a relevant topic – to different neural networks (a computer system modelled on the human brain and nervous system) that are good at different kinds of detection. Microsoft might be good at one thing and Google at another, so each will process the data, draw their inferences and hand it to me, something people can access for a small fee," explains Mukhi. Organisations across the globe now use AI to maximise efficiency/accuracy and minimise labour cost.

Areas of disruption?

Mukhi mentions technology, transportation, fashion, medicine and even journalism as some of the areas to face disruption in the impending AI revolution. He points out how driverless cars is a reality in many countries, which are operating on the basis of the data inference and analysis that we witnessed. IT engineers are already losing their jobs to AI and middle-bracket positions in transport, fashion and medicine are being traded for software assistance.

"Imagine a software that can work out your brand-specific size, your clothes compatible with your physical appearance, much faster than you change out of trial rooms!" says Mukhi, explaining the role of AI in fashion. The same AI can be used to keep a tab on your health, "Going to a doctor only when some doubt arises". Mukhi claims that just as doctors have to go through months of training to master their speciality, a well-trained AI can go from being an expert on one body part to another, depending on the information you feed it.

Writing on the wall

According to Mukhi, these developments mean that the West's requirement for Indian labour will reduce, and if India keeps functioning the traditional way, the high-cost structure would make its exports incompatible with the rest of the world. "Which means we will soon be out of jobs." Even as he complains about India not "seeing the writing on the wall," Mukhi's associates prepare to haul on board one of the city's leading bridal couture designers by explaining to her how AI assistance can help the brand tap into previously unrecognised target audiences and serve existing ones better.

"It is all very well for a TCS or some other IT company to say that they are growing, while the numbers say otherwise. A Snapdeal today rakes in much less than what it would, originally . We should have anticipated this fall," he cautions, implying that for the other sectors, there's still some time to readjust. Unless "the government and bodies like NASSCOM stand up to the truth, there is no way forward," says the man who points out how Accenture PLC, a global management consulting company that risked the "one-time investment" of AI around five years ago has a turnover higher than "Infosys and Wipro put together".

Reason to struggle

Mukhi makes no secret of the crisis that comes in tow — "mechanical positions will be the first to go". So it's best to reorient one's skill sets so as to be part of the creative workforce or the decision-makers. "For example, those taking crucial editorial and managerial calls at a newspaper won't be at risk, but the copy desk, whose editing skills can be mimicked and bettered by AI, will be in trouble," he says.

Invasion incoming?

Haunted by narratives of computer apocalypse in popular culture (Her, Black Mirror), I ask Mukhi if AI could actually take over human mind and behaviour one day. "AI is an autonomous body with a mind and intelligence of its own, distinctly different from what is understood to be human intelligence" Mukhi deliberates. "Will AI be able to brainwash us? I can't tell you, because I don't know."