Hype, déjà vu and smashed keyboards

Summary: Everyone says AI will change everything. But our readers aren’t buying the hype so easily. From chatbot frustration to tech-bubble déjà vu, this week’s responses reveal a calmer, more grounded view of what AI can and cannot replace. The gap between headlines and lived reality is wider than it seems.

The inbox was unusually unanimous this time. “Yes, exactly!” wrote one reader after another in response to Dhirendra Kumar’s latest Editor’s Note on artificial intelligence.

But this wasn’t just about irritating chatbots. The responses carried something deeper. Readers were pushing back against the breathless market narrative that AI will replace everything overnight—jobs, software companies, business models, even human judgment. The bot frustration was just the most visible crack in that story.

Anantharaman Iyer put it bluntly: his experience with chatbots has been “very very annoying. It has not solved a single problem.” Mandar echoed that frustration, saying bots “only give stereotypical answers” and can’t really handle random, real-world problems. Praveen Godbole confessed that after one such exchange, he ended up “smashing” the keyboard.

If you’ve ever tried resolving a telecom or banking issue through a chatbot, you know the feeling. Prasanth called it “refreshing” to read a perspective that mirrors his own concerns about AI bot services. Arun Johary described robotic agents that either escalate to humans or leave issues unresolved, though he fairly admitted he has had “2 successful interactions” recently.

H M Prabhu offered some historical perspective. AI has been a buzzword before, he reminded us, in the 80s, then expert systems and fuzzy logic in the 90s. “Not that there has been no progress,” he wrote, “but the hype has been too much.” That line about what keeps businesses running, systems that don’t randomly fail, resonated deeply with him.

There was also an unmistakable sense of déjà vu. Prabahar Ayyappan compared today’s AI panic with the arrival of core banking systems. Handwritten ledgers gave way to intranet systems, then CBS. “No bank has been closed due to the advent of technology,” he noted. The basics remain. Technology evolves. Panic subsides.

Vivek Banzal, who has been in IT since 1986, made a similar point. The real utility lies in optimising and integrating, not in tearing down everything to disrupt for disruption’s sake.

Panic in markets, perspective in portfolios

The anxiety isn’t just technological. It’s financial.

Praveen drew a sharp parallel with the tech bubble 25 years ago. Back then, he invested in technology funds after their NAVs collapsed, exited three years later with more than 100 per cent returns, and now sees “a scenario like that emerging again.” He is investing in tech funds once more, hoping history rhymes.

Seshachari Madabhushi observed that since ChatGPT arrived, people are “unnecessarily reshuffling their portfolios and incurring losses.” His conclusion: “Old is gold.”

Anantharaman and Prasanth both found the article timely precisely because of the current wave of predictions that AI will replace everything. Rajen Agrawal agreed with the broad thesis but felt the last two sentences reflected “an avoidable personal mindset.” Fair enough. When writing about hype, one must guard against becoming equally absolutist in the opposite direction.

Vikas Basandrai raised another angle altogether: energy. AI, he pointed out, is an energy guzzler with massive capex and still no clear profitability. It’s a question that doesn’t fit neatly into demo videos.

And then there is the economic irony that Prasanth flagged. If AI displaces jobs at scale, who will be left to consume the products of these very corporations? A marketplace without consumers is not progress. It is a loop that eventually collapses in on itself.

Not doom, not denial, but nuance

Not everyone saw the situation in black and white.

Siva S disagreed slightly with the sentence framing. Customer support, he argued, is just “a fraction” of AI’s application. In research, medical scanning, coding and legal processes, AI is already proving itself. He pointed to Bajaj Finance using AI at scale to track and convert calls into significant disbursements. His view is that AI will improve through feedback loops and human-in-the-loop systems, getting better over time.

Arun Johary, too, sees enterprise software staying power, but believes business models will evolve toward usage- or outcome-based structures.

Even Mandar, who is sceptical about the adoption of aggressive AI, spoke from within the system. As a system analyst, he says clients often ask them not to use LLM-generated code because errors, once copied everywhere, can cascade through the system. That caution itself is revealing.

Across the responses, one theme recurs: integration, not obliteration. Optimisation, not overnight extinction.

No one denies that AI will change how we build, test and deploy software. But very few among you believe that decades of battle-tested systems, with millions of bug reports and institutional memory embedded in them, will vanish because a demo looks impressive.

If anything, your responses confirm what the original note tried to argue. There is a difference between technological evolution and technological erasure. The former is real and ongoing. The latter makes for good headlines and volatile stock charts.

And in markets, as in customer service chat windows, headlines and demos are rarely the same thing as lived reality.

Credits

Anantharaman Iyer, Praveen Godbole, H M Prabhu, Arun Johary, Rajen Agrawal, Seshachari Madabhushi, Prasanth, Vivek Banzal, Prabaharan Ayyappan, Siva S, Vikas Basandrai, Mandar

Many of our best stories begin in the inbox. If you have a view, we’re listening. Write to us.

Also read: When the system works against you

This article was originally published on February 17, 2026.