/startuppedia/media/media_files/2025/12/05/copy-of-copy-of-web-22-2025-12-05-13-35-44.png)
Anupam Mittal
In a candid LinkedIn post, Shaadi.com founder and investor Anupam Mittal sparked a fresh debate by remarking that “the most overrated asset in 2025 is a large language model.”
Anupam Mittal reasons why LLMs won't save you
At a time when GenAI dominates every pitch deck and conference stage, Mittal argues that the hype has begun to eclipse reality. He said that while the underlying technology is transformative, its strategic advantage is evaporating.
“You can buy it, rent it, or build a decent version quickly,” he wrote, pointing at how rapidly LLMs are becoming commoditised.
And once a technology becomes widely accessible, Mittal notes, “it stops being a competitive edge.”
Instead of chasing novelty, he urged founders to redirect their energy toward the unglamorous grind of problem-solving.
“The real money is never made in the commoditized tool. It’s made in the relentless application of that tool to solve a deeply boring, high-friction, real-world problem,” he added.
For Mittal, technical architecture is secondary to the mindset of the entrepreneur.
“When I look at a pitch, I don’t care about the model architecture. I look for the founder who is maniacally focused on the grind, not the glamour,” he shared.
According to him, value is not in chasing the Gen AI trend, but it lies in building an Execution AI company.
He went on to reframing the startup world’s obsession with “GenAI” and introduced what he calls the only type of “S.E.X.Y.” that truly scales:
S – Solving real problems
E – Execution with brutal efficiency
X – eXtraordinary focus
Y – Your purpose, your ‘why’
Mittal concluded the post with a provocative question for builders: “What is the most boring, high-friction problem you see GenAI completely failing to fix right now?”
Indeed, it is a challenge that cuts through the noise, pushing entrepreneurs to trade hype for hard problems and algorithms for accountability.