UiPath Founder Daniel Dines Challenges Readers to Disprove His 2028 Book

Daniel Dines, the founder of UiPath, has released a new book titled The Work That Remains, which outlines a distinct philosophy for integrating artificial intelligence into enterprise operations. The core argument is that organizations should empower AI agents to handle preparatory tasks while reserving consequential decisions for human oversight and relying on deterministic automation to execute approved actions. This framework is anchored by a specific, testable prediction regarding the state of enterprise technology by the end of 2028.

Dines predicts that by the close of 2028, any large regulated enterprise managing high-consequence workflows autonomously will do so within an engineered environment rather than by deploying general-purpose agents into legacy systems. He distinguishes his work from typical vendor literature by explicitly inviting scrutiny. “If broad enterprise deployment proves otherwise, this framework weakens,” he writes, clearly identifying the evidence that would invalidate his central thesis.

The operational model can be summarized as a three-step process: AI proposes, humans decide, and automation executes. In this structure, agents are responsible for planning and preparation, people retain ownership of decisions with significant impact, and deterministic systems perform the actual execution of those decisions. To support this division of labor, Dines introduces the concepts of the map and the rails. The map serves as a governed description of how a business actually functions, while the rails represent the machinery that executes stable processes with precision.

A counterintuitive aspect of this approach is the inversion of traditional expectations. While AI agents were initially viewed as the quick win and legacy automation seen as expensive and rigid, Dines argues that the agent’s role at the point of consequence proved more difficult than anticipated, whereas reliable code became increasingly affordable. This perspective aligns with broader industry observations; reports from July indicate that Forrester estimates 75% of projects stall before reaching production, a gap in orchestration that the book aims to address.

Transparency is a key feature of the text. “I run an automation company. That is my declared bias,” Dines writes, acknowledging that firms like UiPath benefit from his conclusions. He urges readers to evaluate the argument on its merits rather than dismissing it due to his commercial interests. This viewpoint has been developing publicly for some time. In July, he discussed how reducing junior hiring deprives enterprises of future leadership and introduced the concept of the cartographer role, which is now expanded upon in the book.

Regarding workforce strategy, the argument focuses on talent density rather than total headcount. Dines advocates for a shift toward fewer passengers and more drivers, creating a workforce where high talent density becomes a mathematical necessity rather than just a corporate philosophy. The specific workflows highlighted in the book, such as claims adjudication, loan approvals, and payment processing, correspond directly to UiPath’s market expansion through its acquisition of WorkFusion.

The narrative also traces the company’s roots back to Europe. After spending five years at Microsoft in Seattle, Dines returned to Romania in 2005 and founded the company from his apartment in Bucharest. Today, UiPath trades on the New York Stock Exchange. The book cover notably credits Claude and ChatGPT, a detail that either demonstrates the argument in action or serves as a clever illustration of the collaborative potential between human vision and AI assistance.