The ‘AI Native’ Trap: Kieranjflanagan on How Jargon Fuels LLM Growth

In a recent LinkedIn post, Kieranjflanagan discusses a common yet potentially counterproductive workflow emerging in companies adopting artificial intelligence tools, particularly large language models (LLMs). He outlines a scenario that he argues highlights a significant driver behind the current revenue growth in the LLM sector.

Kieranjflanagan presents a narrative of an employee tasked with becoming “AI Native” within their organization. This employee is given the objective to create strategy memos using AI, detailing plans and justifications for their effectiveness. The employee then utilizes an AI tool, Claude, to generate extensive documentation.

“Employee goes ‘AI-mode’ working with Claude on memos, pages and pages of technical jargon, impressive sounding words, incredibly detailed graphics, external quotes to convince executives they know what they are doing.”

As Kieranjflanagan describes, the output is a lengthy, jargon-filled document replete with technical terms, impressive vocabulary, intricate graphics, and curated external quotes—all designed to impress executive leadership. The employee, satisfied with the technical prowess displayed in the memos, sends them off to the executives.

The Executive Dilemma: Drowning in AI-Generated Content

The core of Kieranjflanagan’s analysis centers on the executive reception of these AI-generated documents. He points out that this is not an isolated incident; multiple employees are producing similar memos, creating an overwhelming volume of complex, jargon-laden content. Kieranjflanagan suggests that executives are struggling to keep pace with this influx.

According to Kieranjflanagan, the executives, faced with a deluge of these AI-produced reports, eventually resort to using an AI tool themselves to make sense of the information.

“Executive goes to Claude, and puts all memos into Claude and says ‘Summarise all of these into a couple of bullet points'”

The outcome, as detailed by Kieranjflanagan, is that the AI condenses twenty memos into a mere five bullet points. This process leads to the wholesale discarding of the original, elaborate documents. Kieranjflanagan posits that this cycle—employees creating verbose AI-assisted content, and executives using AI to distill it—is a significant, albeit perhaps unintended, consequence of current AI adoption strategies.

Revenue Growth Fueled by AI Workflow

Kieranjflanagan then draws a direct line from this workflow to the financial performance of LLM companies. He argues that this pattern of generating and then summarizing AI content is a primary engine for the sector’s current revenue expansion.

“And that kind of workflow explains 90% of current revenue growth in LLM companies.”

In Kieranjflanagan’s view, the demand for LLM tools is being driven not necessarily by profound strategic insights being generated, but by the sheer volume of content creation and subsequent summarization that organizations are engaging in. This creates a feedback loop where more AI usage necessitates more AI tools, thus fueling growth.

The ‘AI Native’ Paradox

The narrative presented by Kieranjflanagan highlights a paradox for the “AI Native” employee. While their task was to integrate AI effectively, the process described leads to an output that is ultimately bypassed by the very executives it was meant to impress. The technical sophistication and impressive presentation, Kieranjflanagan implies, become secondary to the need for concise, actionable information that AI itself can provide in a distilled format.

Kieranjflanagan’s post serves as a critical commentary on how organizations are currently interacting with AI, suggesting that the perceived value is often tied to the volume of AI-generated output and the subsequent effort required to process it, rather than the inherent quality or strategic advantage of the content itself.