In the echo chamber of Silicon Valley, artificial intelligence is often marketed as a revolutionary force of autonomy and “whiz-bang” reasoning. But according to two top tech CEOs, the customers actually paying the bills—like a heavy equipment manufacturer in the Midwest—are asking the tech industry to shut up about the hype and show them the receipts.
During a Fortune Brainstorm AI panel earlier this month, two tech executives from multibillion-dollar market cap companies got real with Fortune‘s Allie Garfinkle. Freshworks CEO Dennis Woodside and Xero CEO Sukhinder Singh Cassidy dismantled the current AI narrative, arguing that the gap between what developers find “sexy” and what businesses actually need is widening.
Woodside anchored his argument with a specific example from his own client base: Vermeer, a family-owned company in Iowa that has been building wood chippers and other agricultural equipment for a century. Despite having 5,000 employees and complex operations, he said, Vermeer has zero interest in the Silicon Valley hype cycle.
“They don’t necessarily care about the latest thing that we’re pushing in Silicon Valley,” Woodside said. “They need results, they need to grow their business, they need to serve their customers.”
That’s just one of the 75,000 customers Freshworks has, Woodside pointed out, and “the vast majority” are not opting into AI products yet. “They still want a human in the loop.” The most popular product, he disclosed, is a co-pilot that makes their employees more productive, allows them to make more money or save costs.
‘Please For God’s Sake Tell Us What It Means‘
This disconnect is evident in how software companies pitch their products. According to Singh Cassidy, small business customers are suffering from buzzword fatigue. Their feedback to Xero has been blunt: “Can you please just … stop dropping the word AI and tell us what it is and how to use it?”
Singh Cassidy noted that while customers might use consumer AI tools, their attitude shifts when it comes to business software. “I don’t think they come to us looking for AI other than, ‘please, for God’s sake tell us what it means for us.’”
For Xero’s average small business customer, she added, cash flow is all-important. After that, the top consideration is saving time to spend on their business. What do they do with that time? “Generate cash flow.”
A distant third is intelligence on how to grow. But firms are not looking for a chatbot to take over their company, she added. In fact, the idea of an “AI chatbot loose on all your data” is a source of anxiety, not excitement.
“They want control,” Singh Cassidy emphasized. “They want an audit trail… like ‘what exactly did you just do and how do I know it’s accurate?’”
The ‘Sexy’ Tech Trap
The panel highlighted a stark contrast between what engineers admire and what users celebrate. Singh Cassidy described a recent demo of Xero’s chatbot, JAX. While the developers were proud of the AI’s advanced reasoning capabilities—such as the chatbot’s ability to figure out whether to buy or lease a van—the audience cheered for a much simpler feature: the ability to retrieve tax research from the open web.
“We all think [advanced reasoning] is super cool … it’s not what they’re looking for,” Singh Cassidy admitted.
Woodside agreed, noting that the automation of “L1 support”—the basic, front-line customer service tasks—isn’t finding uptake among his customers. “The stuff that we think is so sexy in the Valley hasn’t really caught on yet.” Companies are terrified that early-stage AI will disappoint their end customers. Instead, the real value for a company like Vermeer lies in unglamorous efficiency: using AI to instantly parse thousands of technical manuals to help a human agent answer a specific question about a wood chipper part.
Dot-Com Déjà Vu
Both CEOs drew parallels between the current AI boom and the dot-com bubble of 2000. Woodside recalled the massive over-investment in fiber-optic infrastructure, based on the false thesis that owning the “pipes” meant owning the economy.
In reality, the value accrued to the application layer, meaning the companies that used that infrastructure to serve customers. Woodside said he was seeing a similar pattern today with the billions being poured into chips and data centers. “The infrastructure becomes a layer that we all access, but the value is going to be on the applications,” he said.
Singh Cassidy warned that the market is currently cluttered with unsustainable business models. “You can raise money on an AI feature today,” she observed, comparing it to the ephemeral startups of the dot-com crash. She cautioned that once the current investment cycle ends, the companies built on singular features rather than deep customer relationships will likely disappear.
For now, the message from the “wood chipper” economy is clear: Keep the buzzwords; give us software that works.
For this story, Fortune journalists used generative AI as a research tool. An editor verified the accuracy of the information before publishing.
