By Uma Ganesh
The movie “Matrix” warned us decades ago that humans will one day live in a bubble where they feel they are leading normal lives but in reality they have a plug at the back of their heads with their brains being fed energy by thinking machines. The advent of AI in a seemingly innocuous fashion a decade ago has moved through phases of initial incredulity to fear of autonomous AI to a realisation that AI in an assistive and augmentative manner can be truly useful for mankind. Nobody has however prepared humankind for the near Matrix world that Generative AI could lead us into.
Conversational AI, which uses Natural Language Processing to analyse languages and enable human-like interactions to take place, has been adopted in many business-to-consumer interactions and even governments have been experimenting with chatbots to ease the pain of citizen query handling.
Machine learning and deep learning has also enabled much of the pain of and drudgery of human interaction to be taken over by AI. However the new force that is being unleashed by Generative AI is moving us out of our comfort zone and threatening a deep disruption to the way we work, act and even think in future.
Why is Generative AI different? As ChatGPT has shown in its early days, the ability to create a poem, a song or even a painting can today come close to the product of a thinking human mind and with enough time to soak in ML-led improvements, can enable lazy humans to delegate intellectual pursuits like writing books or movie scripts entirely to Generative AI. Today, a human needs to provide prompts and course corrections to make the output of Generative AI meaningful but that can change dramatically in the days to come, served by exponential growth in technology capabilities.
The scary part of Generative AI in general and ChatGPT in particular is that it has progressed through the development of Large Language Models (LLMs) and “transformer neural networks.” In the next five years, LLMs and generative AI will develop the ability to guide strategic decisions of artistic and business folk and come on par with their cognitive capabilities. A scary scenario that can evolve thereafter is quite akin to the computer software industry where a majority of programmers are simply relying on search techniques to put code together for developing new systems.
Lets not get fooled by the news that ChatGPT could only manage a D grade in an undergraduate economic exam. It’s going to be better in future and in examination systems where “learn by rote” has largely been the norm, machines can easily outperform humans with all our distractions and inherent laziness. However, as an optimist we can still believe that while AI will easily win over natural stupidity, intelligent humans will continue to run business corporations and governments, at least in our lifetime!
(The writer is chairperson, Global Talent Track, a corporate training solutions company, Views expressed are personal)