“Nothing is so painful to the human mind as a great and sudden change”: Victor’s anxiety in Mary Shelley’s Frankenstein is symptomatic of the insecurities that the rapid progress of artificial intelligence (AI), specifically large language models, or LLMs, have spawned in people. The most dominant anxiety is around job losses that AI and the aggressive redefinition of human work are threatening to cause. This worry runs deep, and far and wide, from scriptwriters in Hollywood to employees of Big Tech.
The fear that machines will steal jobs away from humans is as old as the steam engine (read Genpact’s Tiger Tyagarajan’s interview for a metaphorical take on steam engines and the fear it struck among jobbers of the day). So far, the fear was exaggerated, since machines only replaced some menial jobs and, for the rest, created more jobs than the ones they stole. But this time, with AI and LLMs, this fear is not entirely misplaced, as it can already perform a number of high-end tasks, such as coding, analytical reporting, etc.
AI has been floating around for more than a decade now, but it gathered massive steam as LLMs were trained to read large data sets, effectively the entire wide world of the web. And so, you have ChatGPT, or other such GenAI tools, writing everything from school essays to business pitches. Its scanning the internet on behalf of you explains its popularity, but therein lurks the biggest danger. Beyond the intellectual threat they pose to humankind (ChatGPT and its likes are pretending to think on your behalf; in case you did not notice, they are disincentivising you from exercising your grey cells), LLM-powered bots have become the most potent tool for misinformation or fake news and propaganda. While conversations around ethical AI have gained ground and works to restrict the use of GenAI to generate and spread misinformation are afoot, the threat it poses to democracy cannot be ignored when more than four billion people elect their governments across the world in 2024 (The Economist, September 2023 edition).
While many would argue that propaganda and lies are at the heart of any election in any year, it would be unwise to underestimate the power of GenAI in exacerbating it a million times. Deep fakes are already hitting headlines in India. So far, its victims are celebrities, but soon it would be democracy itself. If social media and WhatsApp forwards were not enough, now you can add the might of GenAI to it.
Denouncing technology is not the answer to the challenges it poses. If history has taught us anything, it is that the power to harness technology needs to be tightly bound by laws and regulations. Governments and corporates need to come together to ensure this. Otherwise, history will repeat itself: 200 years of nuclear research to understand the atom better would again culminate into the creation of a weapon of mass destruction.