Upskilling of most jobs is required now, including jobs related to AI: The World Economic Forum’s Future of Jobs has stated that by 2020, five million jobs would have become obsolete as AI replaces human workers. This obviously creates new opportunities in the area of engineering, computing, and quantitative techniques A study by David Deming, associate professor Harvard University indicates that the Industrial revolution 4.0 engineers will require soft skills to be successful in their jobs. This is so because these engineers having to perennially move from one project team to another, as work structures become increasingly more transient. They will also have to understand what customers want so that they can program machines accordingly. They would particularly have to be equipped with the soft skill of empathy. Meanwhile, many job holders will be able to avoid being displaced if they re-skilled or at least expanded their skillset. Mathematicians stand to gain by learning soft skills, while white collar workers will find it advantageous to become conversant in high-school level mathematics. The greatest opportunity lies with educationists who must do the upskilling. It can be as difficult to teach engineers how to be emotionally intelligent as it is to teach white collar workers the rudiments of mathematics.

Educationists must now emphasize an orientation known as learnability: This is simply the capacity to acquire new skills on a continuous basis. The best time to start moulding this orientation is at the kindergarten level. Children must be encouraged to learn a new skill on a continuous basis. In addition to learning alphabets, children can learn a broad spectrum of subjects ranging from musical instruments to carpentry to cooking to sports to dance to acting to painting to building and so on. The aim of educationists then is to make learning something new a rewarding experience for children. In the process, learnability can become internalized as an orientation that is as natural as breathing.

Business school education will need re-thinking: The National Human Resource Development Network (NHRDN), in conjunction with the IFIM Business School, Bangalore had undertaken a survey this year, to ascertain what sort of an education should be provided by b-schools so that the focus is  on learnability. An important finding was that practice courses played a vital role in promoting learnability. These courses require teachers cum mentors who enable self-learning and self-direction. In such courses, the teacher-mentor is responsible for the nurturing of a small group of not more than 10 – 15 students. The teacher-mentor learns with his/her students. S(he) comes to know his/her students as individuals as well. “The man-on-mountain” or “conductor of a symphony” approach favoured by proponents of the case method, such as Roland Christensen of the Harvard Business School, is now so twentieth century. The outcomes of a case discussion are expected, known and thought out in advance. The discussions are conducted within the comfort zone of a classroom. By contrast, the outcomes of practice courses cannot be predicted, as each small group project is unique. The problem that requires solution must be unearthed in the field, while formulating a solution requires prototyping and experimentation. This is not to throw the baby out with the bathwater. Practice courses are to supplement academic courses, not replace them. And they are here to stay as part of any curriculum that seeks to cater to industrial revolution 4.0. This is where we say that the learning imbibed through practice courses cannot be learnt any other way. A prospective manager learns to manage by managing. Every time (s)he executes a project (s)he becomes better at his/her craft, and his/her managerial skills improve. The practice creates confidence in a budding manager’s psyche, which then acts as a spur to further success.

AI will replace components of jobs rather than entire jobs: Learnability assists in both the augmentation of existing skills by a workforce, and the acquisition of new skills. Very often, the augmentation of existing skills enables employees to use AI with telling results. In other words, it helps them work smarter and more efficiently. This in turn leads to greater productivity. Thus AI becomes an enabling tool for an employee, rather than a competitor for his/her job. Most employees who have had the good fortune to use “personal assistants” facilitated by AI (such as Amazon’s Alexa) have found their productivity increase in leaps and bounds. And who would not want to circumvent the bumper-to-bumper traffic of our cities by using self-driving flying cars? I am referring to a digital library as I write this and am glad that I do not have to physically visit a library. If AI can provide convenience or speed, so much more work can be done during a working day. After all, AI can perform only certain components of most jobs. We still need people to decide how everything should come together. We also need people to program AI to our specifications. We need people to clean up the mess when bots go awry. Such occurrences may be infrequent but can be calamitous. Consider the disgust experienced by bank customers when an ATM gets stuck. They will seek out bank employees (people) to sort matters out for them and to vent their frustration. Additionally, the machine will have to be set right by people.

We still need humans in all industries: Most AI are rule-based algorithms which have been created by people. AI cannot replace people in jobs that impact lives and are composed of many complex tasks. It can however remove tasks in jobs that are sheer drudgery. This is commendable. Humans should not be wasted on mundane, repetitive, jobs. They will derive job satisfaction from performing higher-order work. Let’s consider the example of medical practice. Nowadays doctors make a diagnosis using AI based systems. But you still need a doctor to put together all the diagnoses and combine them with the doctor’s own assessment. The sudden flash of insight, the ‘aha’ experience that enables a doctor to make the correct diagnosis goes far beyond the piece-meal diagnosis made by AI systems. And then of course, you need the humaneness of good doctors to guide patients through a recovery process. Let’s consider another example: one can learn business management through MOOCs, but these cannot capture the magic of real classes, with face-to-face interaction. 

AI will necessitate new pedagogies for teaching adults how to use algorithms: Hence AI is going to upend education more than employment prospects. According to an independent report, there is going to be an immediate increase in jobs on account of AI. This increase has been necessitated by requirements in the area of STEAM (science, technology, engineering, arts & design, and mathematics). To this, we can consider further increases necessitated by having to teach STEAM fundamentals to adults looking for upskilling. How can adults, who have not used STEAM knowledge for decades be re-skilled? That is the new challenge for educationists. Those educational institutes which can re-calibrate themselves so that they develop teaching methodologies and curricula for teaching adults the basics of STEAM will be in great demand. These basics will parallel what is taught in high school, but in a way that adults can appreciate. The teaching modules will have to be designed so that the adult students spend most of their class time-solving problems, doing exercises, and generally ‘learning by doing.’ Since much of the donkey’s work associated with making calculations is now automated, learning the basics of STEAM can be made very interesting. Educationists only need to know how that is done. Once the initial trepidation to learning the basics of STEAM basics is overcome, adult learners will move forward by the momentum of their own interest.

AI and infringement of citizen’s privacy: A more worrisome issue than that AI may create job redundancies is the encroachment upon privacy that AI makes possible. We know how Facebook gave access to Cambridge Analytics. We may be subject to surveillance in ways we don’t know by governments, employers, and associates. Just because surveillance by AI is possible, does it mean that it is desirable?