Ever since ChatGPT made its debut in late 2022, artificial intelligence (AI) has kept the world abuzz in anticipation of computers doing what we had assumed only humans were clever enough to do. As AI hype threatens to spell under-delivery on an over-promise, champions of this technology have begun to temper our expectations.
Last week, it was Sundar Pichai, who helms Google, a company that launched its own AI tool, Gemini, soon after OpenAI’s chatbot. We should not expect big leaps of Generative AI in 2025, he declared at a public forum. “I think progress is going to get harder,” he said, as “the low-hanging fruit is gone” and the fledgling industry would need “deeper breakthroughs as we get to the next stage.”
For now, AI tools based on large language models would make incremental gains, especially on reliability, but not inspire the sort of awe that some of us may have been primed to expect.
Pichai’s view of the path ahead seems to conform with that of Microsoft’s chief Satya Nadella, who has also spoken of linear AI evolution instead of exponential growth in wizardry. OpenAI’s Sam Altman, meanwhile, has been talking about how far AI could go. “There is no wall,” he recently posted on X.
Whether or not tech barriers must be overleapt for GenAI to marvel the world any further, investors can’t afford to take their eyes off the ball on what it’ll do even with its current level of development. In October, Pichai said that over a quarter of the software code that Google needed was AI-generated.
Last week, he spoke of a decade’s span for programming to become accessible to millions of more people. This leaves us wondering how soon GenAI might turn a big chunk of India’s $254 billion infotech services industry redundant. That a clock is ticking for Indian programmers engaged in low-end work seems in little doubt.
Big players like TCS, Infosys, HCL, Wipro and Tech Mahindra may have sought to play down the impact of AI by presenting it as an opportunity more than a threat (and it’s true that they’re moving on AI), but industry analysts are not so sanguine.
As reported by Mint, many of them expect pricing pressure to weigh upon the sector’s performance as client contracts come up for renewal. The most exposed are revenues drawn from app making and support, apart from outsourced business process tasks.
Together, these two make up about 40% of the industry’s top-line. While TCS and Infosys recently described their pricing power as stable, how India’s software exporters will fare in the age of AI remains a multi-billion dollar question.
One near-term hope has been an edge on work reliability that our software majors could plausibly claim over the quick stuff churned out by GenAI. As AI bots are found to ‘hallucinate,’ they have a general image of clever tools in need of adult supervision.
So long as their output needs to be under close human vigil, clients may find it safer to not just let software specialists deploy GenAI for coding, bug-checks, etc, but also pay a fee for this. That may buy these firms some time while GenAI ascends its whiz curve.
Inspired by ChatGPT o1-Pro’s ability to solve NYT Connections, a game that asks us to group 16 words into four groups of four, Deedy Das of Menlo Ventures had this to post on X: “The rate of progress of [AI] is faster than most comprehend.”
Prowess at a word game won’t drop too many jaws in wonder. But our software majors must race the clock for a clear strategy to ride the AI wave.
Â
#Indian #software #majors #clear #strategy #ride #wave
