← Retour au blog
tech 11 August 2026

Humanising LLM Outputs Is Dumb

Large Language Models (LLMs) are powerful, but attempting to make their responses more human-like is misguided. Let's explore why focusing on efficiency rather than humanisation is more prudent.

Article inspired by the original source
Humanising LLM Outputs Is Dumb ↗ kuber.studio

Introduction

Large Language Models (LLMs) have taken the tech world by storm. Their ability to generate text, answer complex questions, and automate processes is undeniable. Yet, a trend is emerging: making these outputs more "human-like." But is it really necessary?

Why Humanise LLM Outputs?

The idea behind humanising LLM outputs is to make them more understandable and relatable for human users. For instance, an LLM might be programmed to use more familiar language or to include emotional expressions. While this might seem appealing, it poses several issues.

Unnecessary Complexity

Humanising LLM outputs adds an unnecessary layer of complexity. According to a recent McKinsey study, 40% of companies adopting AI primarily seek to optimize operational efficiency. Adding human nuances diverts from this core objective.

Risk of Misunderstanding

Humanising responses can also lead to misunderstandings. An LLM using emotional expressions might give an impression of understanding or intent that doesn't exist. This can lead to unrealistic expectations from users.

Efficiency First

Focus on Accuracy

LLMs are designed to process and generate accurate information. In 2023, OpenAI reported that improvements in accuracy increased user satisfaction by 25%. Developers should focus on enhancing accuracy and relevance rather than adding human traits.

Concrete Applications

Take the example of finance. An LLM can analyze thousands of transactions to detect potential fraud. Here, accuracy and speed are crucial. Humanising outputs would be not only unnecessary but counterproductive.

The Future of LLMs

In a world where AI is becoming ubiquitous, it's crucial to set clear priorities. Decision-makers should focus on leveraging LLM capabilities to solve real-world problems rather than attempting to make them more human. By doing so, they can maximize the value these technologies bring to their operations.

Conclusion

Humanising LLM outputs is a distraction from the true benefits AI can offer. By focusing on accuracy and efficiency, businesses can truly transform their operations and improve their bottom line.

Let's discuss your project in 15 minutes.

LLM AI Automation Humanisation Efficiency
Deepthix newsletter · 100% AI · every Monday 8am

An AI agent reads tech for you.

Our AI agent scans ~200 sources per week and ships the best articles to your inbox Monday 8am. Free. One click to unsubscribe.

Visit the newsletter page →

Want to automate your operations?

Let's talk about your project in 15 minutes.

Book a call