NEA’s Tiffany Luck says enterprises are still figuring out their AI ROI

The recent remarks from NEA partner Tiffany Luck underscore a quiet reckoning in the tech industry: the era of unchecked AI experimentation has collided with cold financial reality. While many enterprises raced to integrate AI into every workflow—often under pressure from investors and competitive FOMO—few paused to ask whether those tools were actually delivering value. Luck’s observation that companies are still struggling to quantify AI’s return on investment suggests a broader disconnect between hype and execution, one that could reshape how businesses approach emerging technology in the coming years. This isn’t just about cost overruns; it reflects a deeper uncertainty about AI’s role in enterprise. Early adopters, especially in Silicon Valley, treated AI as a panacea for productivity gaps, but the Uber example—where aggressive AI deployment reportedly led to inefficiencies or even wasted spend—serves as a cautionary tale. The problem isn’t the technology itself but the assumption that more AI automatically equals better performance. Many companies are now facing a reckoning: their AI investments were often made without clear benchmarks, leading to bloated budgets and unclear outcomes. What happens next will depend on whether businesses pivot toward disciplined experimentation or double down on unproven use cases. Regulators and boards may demand stricter accountability, forcing leaders to justify AI spend with measurable results rather than futuristic promises. Meanwhile, the tech industry’s next wave of innovation could shift from sheer experimentation to refinement—focusing on AI’s most practical applications rather than its most extravagant ones. This moment also exposes a cultural shift in corporate decision-making. For years, the mantra was “move fast and break things,” but as AI’s limitations become clearer, the industry may shift toward slower, more deliberate adoption. The question now is whether companies will treat this as a temporary correction or a fundamental reset in how they evaluate new technology. Either way, the era of AI as a default solution—without proof of concept—is giving way to something more pragmatic.