🧠 How AI Redefines Language Itself

Sign Up | AI Toolkit | Meet The Team

Welcome back AI prodigies!

In today’s Sunday Special:

• 📜The Prelude

• 🗣️Linguistic Relativity

• 🤖LLMs as Relativity Machines?

• 🔑Key Takeaway

Read Time: 7 minutes

🎓Key Terms

• AI Agents: Virtual employees who can autonomously analyze, execute, and fulfill without you lifting a finger.

• Large Language Models (LLMs): AI Models pre-trained on vast amounts of high-quality datasets to generate human-like text.

🩺 PULSE CHECK

📜THE PRELUDE

Beneath every ChatGPT output lies a subtle disclaimer: “ChatGPT can make mistakes. Check important info.”

The world’s most popular conversational chatbots, like ChatGPT, are powered by LLMs: probabilistic machines designed to predict the most likely next word within a sentence.

Although LLMs are made of software, they don’t function in the predictable manner of traditional software. In other words, LLMs often produce different outputs when given the same input. This ability allows them to generate an infinite variety of outputs. However, they’re far from arbiters of the truth.

The elusive nature of language has long been recognized by some of the world’s greatest thinkers. German philosopher Friedrich Nietzsche described truth as “illusions we’ve forgotten are illusions.” What we call truth is often just a convention hardened by repetition.

LLMs expose our relationship with the truth. Their outputs are typically not objectively true, but they often feel true because they match our expectations of what credibility sounds like: logical, eloquent, and error-free.

How does language shape the way we act and think? Can AI-generated language exert the same influence? How can we harness the benefits of LLMs without letting them dictate our thinking?

🗣️LINGUISTIC RELATIVITY

⦿ 1️⃣ The Power of Language.

Language is the filter through which we interpret everything. Yet, it’s not clear how much language influences our thought processes and, by extension, our behavior.

⦿ 2️⃣ Language Determines Thought?

Advocates of Linguistic Determinism argue that a person’s language determines their way of thinking. In the 1930s, American linguistic anthropologist Edward Sapir and his apprentice, Benjamin Lee Whorf, proposed the Sapir-Whorf Hypothesis, which states that the grammatical structures and verbal agreements we choose to use within our language influence how we perceive the world. In other words, language influences our thoughts and dictates our perception of time. Consider how different languages determine time:

In English, verbs often change form depending on when something happens, a feature known as Future-Time Reference (FTR):

• Past Tense: “I ate breakfast yesterday.”

• Present Tense: “I eat breakfast every day.”

• Future Tense: “I will eat breakfast tomorrow.”

In Mandarin, the verb itself often stays the same. Instead, time words are added to indicate when something happens:

• Past Time: “Wǒ zuótiān chī zǎocān.” → “I yesterday eat breakfast.”

• Present Time: “Wǒ měitiān chī zǎocān.” → “I every day eat breakfast.”

• Future Time: “Wǒ míngtiān chī zǎocān.” → “I tomorrow eat breakfast.”

English speakers view time as distinct moments: past, present, and future. Mandarin speakers view time as more continuous and less divided.

Memory works differently across languages too. English speakers often encode the agent of an action, even in accidents (e.g., “She broke the vase!”). Japanese speakers use more agentless phrasing (e.g., “The vase broke!”). As a result, English speakers are more likely to remember who caused an accident, while Japanese speakers are more likely to remember how it happened.

⦿ 3️⃣ Language Doesn’t Determine Thought?

Critics of Linguistic Determinism argue that thoughts aren’t trapped or influenced by language. Canadian cognitive psychologist Steven Pinker popularized the idea that humans possess a language of thought called Mentalese: an internal sense of how the world works that exists independently of any spoken language. In other words, we form ideas in our minds, and spoken language is just how we label those ideas to share them.

Again, consider how different languages think about time. English speakers think about time horizontally: “last week” is behind us and “next week” is ahead of us. On the other hand, Mandarin speakers think about time vertically: “shàng gè xīngqī” conceptually translates to “above week.”

Does this mean English speakers and Mandarin speakers actually view or experience time differently?! Not really. They both understand time in a basic way. For instance, they both know that “a few days from now” means something happening in the future. The difference is mostly in how they talk about time, not how they think about time.

Psychological research supports aspects of Mentalese. According to the Dual-Coding Theory (DCT), our minds encode information in two types of imagery: verbal and nonverbal. In simpler terms, we don’t rely only on language to think; we also use mental pictures.

⦿ 4️⃣ The Verdict?

Language doesn’t control how we think, but it influences the way we think. Yes, we don’t need words to think. However, language influences how we frame, process, and recall experiences. For instance, speakers of weak-FTR languages like Mandarin tend to engage in longer-term financial planning compared to speakers of strong-FTR languages like English.

🤖LLMs AS RELATIVITY MACHINES?

⦿ 5️⃣ LLM Usage: Scale and Scope.

The world’s most popular LLMs, including OpenAI’s ChatGPT, Google’s Gemini, Anthropic’s Claude, and xAI’s Grok, process approximately 20 billion prompts from over 1 billion people every week. In the U.S., 52% of adults report using them.

That doesn’t even account for LLMs embedded directly into websites as built-in features. Amazon’s Rufus sits at the bottom of each product listing, assisting 274.3 million customers per day to make informed purchase decisions.

LLMs are also increasingly being applied to complex daily workflows within corporate environments. From mid-2023 to mid-2025, the percentage of U.S. workers using LLMs rose from 8% to 28%.

This rapid adoption is mirrored by the growing presence of AI-generated text across the internet. For example, 54% of LinkedIn posts longer than 100 words are likely AI-generated.

LLMs are woven into the fabric of how we create, connect, and collaborate. Given the influence of language on society, it’s critical to understand the nature of an LLM’s outputs. Unlike human language, which is inseparable from intent and authorship, AI-generated language lacks a speaker. This raises fundamental questions about trust, meaning, and authority in a world where so much of the language we consume may originate from LLMs.

⦿ 6️⃣ LLM Outputs: Lacking Intent.

French sociologist Pierre Bourdieu argued that language is a form of symbolic power. In other words, the authority of language rests not only on what is said, but also on who says it. Every human utterance carries intent, whether it’s to entertain, persuade, or inspire. AI-generated language lacks this foundation. It’s fluent and coherent, but not authored; there’s no motive, accountability, or psychological state behind the words.

AI-generated language is rarely marked and often indistinguishable from human language. Once we assume a human authored the words, we instinctively wonder why they wrote them, whether we believe them, and how we should respond to them. Of course, there’s no inner motivation to interrogate. We’re projecting intent onto AI-generated language that inherently contains none.

⦿ 7️⃣ Why Does This Matter?

Language isn’t just about words; it’s about context, connection, and consequence. When LLMs produce fluent, even poetic AI-generated language with no underlying consciousness, it challenges three core assumptions:

1. “Can meaning exist without intent?”

2. “What does authorship mean now?”

3. “Can creativity exist without a creator?”

As AI-generated language continues to blend into human language, distinguishing between intentional expression and algorithmic output becomes increasingly challenging, affecting everything from how we assign credibility, how we engage emotionally, and how we interact ethically with language.

🔑KEY TAKEAWAY

LLMs challenge fundamental assumptions about meaning, authorship, and creativity, forcing us to reconsider what it means to communicate authentically and engage ethically in an increasingly automated world blended with human expression and machine communication.

📒FINAL NOTE

FEEDBACK

❤️TAIP Review of The Week

REFER & EARN

🎉Your Friends Learn, You Earn!

You currently have 0 referrals, only 1 away from receiving 🎓3 Simple Steps to Turn ChatGPT Into an Instant Expert.

Share your unique referral link: https://theaipulse.beehiiv.com/subscribe?ref=PLACEHOLDER