Run, the Robots are coming!
Run, the Robots are coming!
But hey, maybe they will LOVE you!
It is quite fascinating to see some of the recent headlines since the launch of ChatGPT. As it is in global politics, it seems like there is a dearth of centered opinions. On one side the doom-and-gloom crowd are at their Sunday best trying to insist that AI will take over our jobs, our lives, our children’s brains and soon overrun us, and human beings will merely be but slaves. Some of them are writing open letters for shutting down AI research for 6-months, or forever. Some countries are banning GenAI based products. While others are sure that the next war will be between robot armies.
On the other hand, there are those who believe that AI can find a cure for all diseases affecting human beings (and the Chan Zuckerberg Initiative is spending billions on it, among others), help us in achieving world peace, while some people are trying to develop new gods for the AI world!
Whichever side you are on, and even if you could not care less about such bombast, the thought of how AI will affect your job now and in the coming years, and also how it would impact the landscape of jobs, would have crossed your mind for sure. If it has not, maybe it should. We know that from automatic temperature control to AI supported maps, our lives have become easier and much more convenient with the rise of AI. But will take over our jobs or will it work alongside humans?
Recently a lot of folks have been saying, “AI will not take away your job but someone who knows how to use it will”. I believe there is quite a bit of truth to that. Having said that, the impact will keep growing larger over the years. At the limit, should we complain if AI does most of our work a decade from now, the government pays us a decent universal basic income, pays for health/education/food/clothing/shelter – and humanity works on finer and higher pursuits as opposed to fighting over limited resources in an endless rate race?
But let us look at the immediate future first before we also start dreaming about AI-generated utopia (and not just in the Metaverse😊). The legendary McCulloch and Pitts paper of 1943 (“A Logical Calculus of the Ideas Immanent in Nervous Activity”) to “Attention is All You Need” (2017) artificial intelligence indeed has come a long way!! AI now can perform many tasks that were previously thought to be the exclusive domain of humans. “According to McKinsey Global Institute, at least 30% of work in 60% of organizations can be carried out successfully using automated processes – something that will likely revolutionize over half of the current jobs once implemented. As a result, the firm’s calculated scenarios suggest that anywhere between 75 million and 375 million workers will have to change occupation by 2030, with automated alternatives likely to either completely or partially displace many jobs.” Research after research is showing us that it is true that AI can automate many tasks that were previously done by humans. This can include everything from data entry and customer service to manufacturing and healthcare. In some cases, AI has even been able to outperform humans in certain more ‘advanced’ tasks, such as diagnosing diseases or identifying patterns in large data sets. But I think that there is still a long way of saying the sentence “Taking Over” and I will try to justify my case in the next few paragraphs.
AI is still reliant on humans to create, train and maintain it. They still require human oversight and intervention to ensure that they are functioning properly even if AI algorithms can be programmed to perform certain tasks. Additionally, AI systems are only as good as the data they are trained on, and they can only make decisions based on the information they have been given. But what about driverless cars, even in the advanced economies where people usually follow rules, those countries still haven’t been able to implement driverless cars. Now imagine driverless cars in countries like India. This means that they are often limited in their ability to adapt to new situations and make nuanced or creative decisions. Thus, the turmoil will affect many but create a new set of skill-based roles which will be higher in need than before. “World Economic Forum fine prints report reveals that an estimated 75 million existing roles will be displaced by automation by 2022, which is a lot of jobs lost in a very short amount of time. But at the same time, there's the potential for 133 million new roles to emerge "that are more adapted to the new division of labor between humans, machines and algorithms," ultimately resulting in net job growth.”
Another important consideration is that AI systems cannot replicate many of the complex social and emotional skills that humans possess. (Are you still humming that “Nirma” song? Now that I mentioned it, you surely are 😊). While they may be able to understand and analyze language, they do not have the ability to understand the deeper meanings and emotions behind human communication. This means that they are not able to fully replace human interactions in many settings which require Zero to One thinking. What do I mean by that? Our CEO wrote a paper with some of his students back in 2014 where they predicted performance of Hindi movies based on basic social media information. Nearly a decade back social media presence of Indian movies was limited but yet they could create a simple model which predicted with more than 80% accuracy the gross box office earnings of a movie. Now, for a movie studio, if they realize this ‘formula’ they will keep repeating it – and other studios using similar algorithms will keep doing the same too. Same genre, similar looking actors, similar story lines, same tropes for success…we will, as humans using AI, will start behaving closer to mindless automatons if we go down that path. Thankfully, human beings have a natural defense against such an outcome of AI-driven homogeneity – and as kids of the 21st century love to say, “I AM BORED”. After reading two formula driven books, seeing five formula-driven movies and reading a hundred formula-driven social media posts written by GPT, we will get bored of the same. We will crave the human randomness which allows us to go from 0 to 1 and create something completely new, without hallucinating like GPT models.
For certain optimization tasks, data extraction and labeling tasks, summarization, and even pattern recognition, AI will have a deep impact. But to solve Human Problems, with all their randomness and boredom, it will still take a human being. It is important to recognize that AI is not a replacement for human intelligence and skills. While AI can handle certain tasks more efficiently and accurately than humans, it lacks the creativity, emotional intelligence, and problem-solving ability that makes us uniquely human.