Since the early 2000s, low-cost and high-powered computation , among other things, have let smart algorithms make an impact on our daily lives. Search engines, online shopping cues and traffic guidance are typical examples. But outside tech-business-geek circles, few realized that Artificial Intelligence (AI) had entered their life. AI burst into public discourse only on 30 November 2022 , with the launch of ChatGPT, an easy-to-use tool displaying the wonders of AI.
Since the early 2000s, low-cost and high-powered computation , among other things, have let smart algorithms make an impact on our daily lives. Search engines, online shopping cues and traffic guidance are typical examples. But outside tech-business-geek circles, few realized that Artificial Intelligence (AI) had entered their life. AI burst into public discourse only on 30 November 2022 , with the launch of ChatGPT, an easy-to-use tool displaying the wonders of AI.
ChatGPT allowed any user to log in and access its user interface and type any question in English. On the user’s request, ChatGPT can write poetry, motivational speech, provide advice on improving relationships, summarize laws and regulations that existed pre-2021, and also do homework for students. As of now, it handles social sciences and humanities subjects much better than answering mathematics problems. However, subject matter experts will be able to improve ChatGPT output in most instances. But, as it currently stands, ChatGPT can raise the efficiency of generating reports and topical commentaries, drafting legal or technical documents, and writing software code.
ChatGPT allowed any user to log in and access its user interface and type any question in English. On the user’s request, ChatGPT can write poetry, motivational speech, provide advice on improving relationships, summarize laws and regulations that existed pre-2021, and also do homework for students. As of now, it handles social sciences and humanities subjects much better than answering mathematics problems. However, subject matter experts will be able to improve ChatGPT output in most instances. But, as it currently stands, ChatGPT can raise the efficiency of generating reports and topical commentaries, drafting legal or technical documents, and writing software code.
Subscribe to Continue Reading
While ChatGPT’s GPT stands for ‘generative pre-trained transformer’, AI is classified as a multi-faceted and transformative ‘general purpose technology’ (GPT). In the last 300 years, the world has seen at least five of these. It started with the steam engine, after which we got electricity, the internal combustion engine, computers and internet. They resulted in tectonic shifts in the industrial landscape and improved standards of living. Each GPT created new (and more) jobs that were difficult to imagine before its emergence.
There are no unmixed blessings, however. In their initial stages, GPTs tend do create structural unemployment. Arguably, later GPTs created less of it than earlier ones. Such persistent joblessness is not inevitable. Understanding job evolution and preparing society for such GPTs is the key to guarding against structural unemployment.
The core needs of societies do not change much. However, technology changes the way those needs are met. For example, in India, till the 19th century, the delivery of written letters for most people was enabled by ‘dak hircarrah’ or ‘runners’. These men would run from one place to another with a sack of mail. Then came the postman. Postmen were expected to read and write letters to help receivers and senders of the same. They often used bicycles and had no long-distance running. Fast forward to the 1990s, when e-mails significantly overtook written letters . The new ‘postmen’ were highly trained telecommunication experts who worked in air-conditioned offices. Thus, the basic job did not disappear. It evolved into an avatar which required more skills but less physical work. And then there were ‘new’ new jobs . In the 1970s’ India , it was difficult to envision that by 2000 the country would acquire prominence in a global business, and that too, thanks to professionals called software engineers. This job did not exist in the 70s.
AI will create jobs that are difficult to visualize today. One may only guess. If our communication devices and home appliances get digitally connected via internet-of-things (IOT), and the same handsets also offer access to our banks, then we may possibly need to hire cyber security experts. Today’s physical security guards could be tomorrow’s cyber security professionals, who happen to be in short supply right now. One may also require medical professionals with knowledge of both AI and medicine. India can definitely do with a more AI-enabled legal system, which would require lawyers and judges to have functional knowledge of AI. Thus, jobs will be there. We need to ensure resources are available to perform them.
Past GPT trends suggest that every successive GPT’s time taken to impact society is shortening, while the impact is turning more exponential. Steam engines evolved between 1690 and 1710, with an industrial revolution to follow on its back five decades later. In 1969, Arpanet, the precursor of the internet, took shape in US research labs. The internet itself took shape in 1983. By 2010, the web had become an integral part of our lives. The internet took far less time than the harnessing of steam power to impact society.
We are at least a decade into AI’s GPT cycle. AI’s impact on society may be more exponential and will play out in a shorter time span. Countries that enable their societies to absorb AI are likely to get exponential benefits, while countries which fail will be left behind. India is among the first with a national AI policy. The time is ripe for effective governmental intervention to prepare our workforce so that we can make the most of AI’s benefits. Action must focus on three areas:
Upskilling: Companies need to spend massively in upskilling their existing resources to leverage AI. The private sector may need a governmental nudge to take this up post haste. Such training and upskilling expenses could be treated as Corporate Social Responsibility (CSR) spending.
Intellectual property: To achieve AI supremacy, developing intellectual property (IP) is key, and thus India need to attract AI researchers from global pools. Further, AI research expenses and IP need to get the same depreciation benefits as physical assets, since arguably they will become more important.
Academic infrastructure: All school boards in India need to focus on Science-technology-engineering-mathematics (STEM). Changing a syllabus is easy. It will require a capability-building exercise of re-training relevant teachers for newer ways of teaching. It will also demand an improvement in the teaching infrastructure of schools everywhere. Universities would need to make a larger effort to attract academic and research talent for STEM subjects and particularly for AI.
Rote learning must retire; even an early tool like ChatGPT gives better results than rote learners do. For the country’s future, India needs empathetic creators with a superior grasp of technology.
Deep Mukherjee is a quantitative risk management professional and a visiting faculty of risk management at IIM Calcutta
Source link