Rapid AI transition

ChatGPT's new version raises old questions

Nearly 18 months after OpenAI launched its artificial intelligence (AI) chatbot, ChatGPT, the Microsoft-backed company has launched its latest large language model (LLM) called GPT- 4 Omni or GPT- 4o. In doing so, OpenAI reaffirms its dominant position globally in the AI space. However, other tech giants like Meta and Google are not far behind. Google, for example, has started powering its search engine with AI Overviews. The main point of departure between GPT- 4o and previous versions of ChatGPT lies in its ease of interaction. ChatGPT was primarily a text-based LLM interface where users could type text-based questions and wait for text-based responses. Besides, ChatGPT - 4 was only available to paid subscribers. GPT- 4o, in contrast, is free, easier to use, and is able to mimic human interaction in real time. It is also a multimodal AI model — integrating audio, speech, vision, and text with minimal time lag, enabling verbal communications with a chatbot that convincingly speaks like a human.

Further, it is capable of translating text, identifying emotion and tone in users’ speech, solving mathematical equations and much more, making it uncannily human-like. The latest technological development augurs well for both OpenAI and its investors like Microsoft. Currently valued at more than $80 billion, OpenAI’s valuation has tripled in less than 10 months. Even though AI-led disruption is already evident in labour markets across the world, it remains to be seen whether generative AI models and LLM interface adoption will have a net positive effect on labour market outcomes in the medium term. AI-induced automation not only promises higher productivity and standard of living, it evokes fear of job dislocation, increasing unemployment, and deepening poverty. The impact of AI is also likely to differ significantly across countries. Developing countries like India, relying on labour-intensive employment, are expected to face less AI-led disruption. At the same time, they may find it difficult to catch up with early AI-driven productivity gains, given their infrastructure deficit and lack of a skilled workforce. The 2024 Work Trend Index Annual Report, recently released by Microsoft and LinkedIn, reveals a striking increase in AI utilisation in the workplace globally, with its use nearly doubling in the past six months. Around 75 per cent of respondents reported that AI performed most of the mundane tasks, thereby saving time, boosting creativity, and allowing them to focus on more important work.

At the same time, concerns regarding the skills gap in terms of AI literacy, the possibility of occupational layoffs, inadequate resources for the reskilling and upskilling of workers, the spread of misinformation, algorithmic bias, and cybersecurity risks remain critical from a policy standpoint. The rapid pace of AI evolution can result in disruption at many levels. While firms and individuals are adopting AI at a rapid pace, the productivity boost could be at the cost of human employment, at least in the initial stage. It’s not clear at this stage how the policy establishment in India and around the world will deal with such potential challenges. But that’s not the only challenge. The creation and dissemination of information and opinions are evolving in ways that were not known until recently and will have far-reaching consequences. Therefore, given the backdrop, while India must actively pursue productivity gains, it’s vital for the policy establishment to remain vigilant.