Artificial intelligence, for many, sounds an alien or ultra high-tech concept, when the truth is that it not anymore so. There is already so much of AI in daily life, just that most of us do not realise that.
In the words of Prof Raj Reddy, one of the early pioneers of AI and the founding director of the Robotics Institute at the Carnegie Mellon University, AI was actually an attempt in the early ’50s to ‘automate tasks usually thought to be uniquely human’. Though the progress made in the last 50-odd years is ‘probably automation of just one per cent of what humans can do’, we are still on the verge of a revolution, he says.
And that revolution riding on the waves of AI, a creation of the inventiveness of man, could change everything. AI-driven computers are already beating world champions at chess and self-driven cars are being tested by Tesla and similar firms, while attempts at automated learning systems are fast leading to a situation where language will no longer be a barrier for anything, think Google Translate or Microsoft Translator and their ilk.
“In the future, there will also be personalised apps and guardian angel apps, which will do things that you do not know to do, which can warn you if your heartbeat is abnormal (wearables like Apple Watch are already capable of ECGs and much more) or even order medicines for you, or if there is a scarcity of water or food, help you navigate such problems,” said Prof Reddy while talking at a workshop on Artificial Intelligence at the International Institute for Information Technology Hyderabad (IIIT-H).
For a sample of how much AI is there already in our daily lives, just look at the advertisements that pop up on Facebook or on Google. Try searching for a camera, and you will have camera ads popping up, even in the spam box of your Gmail account. On the other hand, AI is helping the Telangana Police, and police forces across the world, use the Facial Recognition concept with great success and to identify suspects from grainy CCTV visuals. AI is helping air traffic controllers keep flyers safe in the air, so is it helping recruitment agencies find potential candidates from job websites, just as it plays the role of a matchmaker on matrimonial sites.
If this is not enough, Facebook, which is believed to have 2.38 billion monthly active users as of the first quarter of 2019, has, for long, been using AI for different experiments, and according to news reports, with much success as well. Using natural language processing and speech techniques among others, Facebook had used ‘proactive detection’ AI technology more than two years ago to scan posts for patterns of suicidal thoughts, apart from using AI to detect ‘troubling’ posts and to ‘prevent suicides’.
There are many more applications, some which are not that encouraging for the common man, like how cab aggregator Uber was found to be using AI. In reports that came out about three years ago, it was revealed that Uber, through its mobile app, could keep a watch on mobile phone battery levels. And it was also found that users with low battery levels were ‘victims’ of sudden surge pricing.
“People with low battery fear getting stranded somewhere so they are much more willing to spend more. Users with fully charged phones have the flexibility of being able to wait longer to see if the surge pricing changes after 15-20 minutes” was how technology experts explained the alleged price surge when battery levels were low. Obviously, AI was telling Uber about the battery levels and, in turn, triggering surge pricing, tech experts argued.
So, will our lives be gradually and completely taken over by AI? Prof Reddy does not think so.
“Even 100 years on, there will be many things that the computer or AI won’t be able to do, because that is the power of the human brain,” he notes, adding that at the same time, the doubt whether AI would make humans more lazier was subjective, since AI hinged upon the ‘principle of least effort’.
What is AI
AI, for the uninitiated, is sometimes called machine intelligence and is used to describe machines/computers that mimic cognitive functions that humans associate with other human minds, such as learning and problem-solving.
Modern machine capabilities generally classified as AI include understanding human speech, competing in strategic game systems, autonomously operating cars, intelligent routing in content delivery networks, and military simulations.
A glimpse of what AI can do
From city level road audit systems that locate potholes and water logging, to sports analytics, the applications of AI appear to be endless. According to CV Jawahar, head of the Centre for Visual Information Technology (CVIT) at IIIT-H, the buzz at the 2019 IEEE Conference was on ‘Computer Vision and Pattern Recognition (CVPR): Inverse Cooking’, an AI technique that uses food images to generate recipes!
On AI uses in the habitat, including agricultural land, urban areas and issues plaguing them, P Krishna Reddy from the IT Centre for Agricultural Research of IIIT-H said there were already data science based systems to provide agricultural advice to farmers. Two such tools – eSagu, an agro-advisory system that works at the farm as well as the village level by using coloured photographs of the crops, and Crop Darpan, a smartphone-based application that serves as a crop diagnostic tool for farmers were already developed.
In health, robot-assisted surgeries were just a tip of the iceberg. According to Radha Rangarajan, head, Innovation and Strategy, at the Ojas Medtech Incubator, AI uses both structured and unstructured data in the healthcare context from medical notes, recordings from electronic devices, physical examinations, to clinical laboratory reports and images to learn features. AI was useful in drug discovery and could speed up drug development. However, it would not replace doctors, she said, adding that it would only help them make better decisions.
KS Rajan of Research Lab for Spatial Informatics says AI was being used for managing crowds. This was through the deployment of smart CCTV cameras, as was done during the last Godavari Pushkaralu.
Tackling fake news
The AI-related work at IIIT-H’s Information Retrieval and Extraction Lab has a lot to do with fake news, the buzzword of the year.
According to Vasudeva Varma, professor and Dean (Research), the Lab has recently developed the Fake-o-Meter tool, which ‘overcomes current limitations of existing content verification solutions’. The Lab used Deep Learning methods and Natural Language Processing to create the Fake-O-Meter, a mechanism to segregate real news from fake news.
The model could be expanded to diverse domains (political domains – world or specific geographical divisions, financial domains, public health; arguably to any domain of interest). The model’s applicability was only limited by the availability of past data of a specific domain.
Past data is used to ‘teach’ the model which, then, applies itself to future items of the same domain, he said, adding that media houses stood to gain immensely if this automation became intrinsic to their news production process.
The Fake-O-Meter, in simple terms, is a web engine where one can copy the tweet or headline of a news story and submit it to get a colour-coded response on its authenticity in percentage terms. News that has a low probability of being fake is coded green with lower percentages. On the other hand, higher percentages are coded red indicating high degree of fakeness.
At present, efforts were on to identify the content more accurately, whether it was posted with ill intent or not, with the challenge also being on how fast the meter could detect fake news, said Prof Varma.
AI and jobs
Will AI displace humans from workplaces? Statistics from different surveys and studies are triggering waves of panic, but experts like Prof Nimmi Rangaswamy, anthropologist, Social and Human Applications for AI (SAHAI), says the statistics need demystification and substantiation to know exactly what they mean.
Though there were jobs, like IT testing, that were sitting targets for automation, Rangaswamy says there are certain limits to automation from an ethical standpoint as well as in cases where creativity was needed.
A report on the future of jobs in India, by Ernst and Young India and NASSCOM said nine per cent of the workforce would be redeployed in new jobs that do not exist today, while 37 per cent would be deployed with radically changed skill sets. Studies have indicated that five to 10 per cent of existing jobs would be automated while about 60 to 70 per cent of the workforce would have to reskill themselves.
Predictable physical activities, like packing jobs, data processing and collecting, accommodation/food services, manufacturing, transportation and warehousing were the ones with high possibilities of automation. A McKinsey study had even said 69 per cent of jobs in India were threatened by automation, especially in the quality testing sector.
However, this statistical tyranny required validation, which only time can bring.