---
schema: 1
type: term
title: '**LLM** (large language model)'
description: >-
  **LLM** (large language model) is a class of AI model — exemplified in this
  material by **ChatGPT** and Notion AI — that writes, reasons, and processes
  large amounts of text. In the material, it is treated as a _tool_ or
  _function_ that augments thinking (a “thought calculator”), useful for getting
  structure and drafts, but **not built for learning**: understanding is
  something you still do yourself.
canonical: >-
  https://liquidgold.cc/channel/everyinc/explore/term/llm-large-language-model-5267e899caa0
markdown: >-
  https://liquidgold.cc/channel/everyinc/explore/term/llm-large-language-model-5267e899caa0.md
llms: null
parent: 'https://liquidgold.cc/channel/everyinc/llms.txt'
channel: everyinc
updated: '2026-10-23T22:12:35.346Z'
language: null
children: []
---


# \*\*LLM\*\* (large language model)

Canonical: https://liquidgold.cc/channel/everyinc/explore/term/llm-large-language-model-5267e899caa0

Parent: [Channel index](https://liquidgold.cc/channel/everyinc/llms.txt)

## Definition

**LLM** (large language model) is a class of AI model — exemplified in this material by **ChatGPT** and Notion AI — that writes, reasons, and processes large amounts of text. In the material, it is treated as a _tool_ or _function_ that augments thinking (a “thought calculator”), useful for getting structure and drafts, but **not built for learning**: understanding is something you still do yourself.

Across the material, **LLM** (large language model) is used in the everyday sense made popular by **ChatGPT**: an AI model that can write, reason, and process vast amounts of text — the engine behind tools like ChatGPT and Notion AI. Linus Lee frames it as a *tool* or *function* that augments thinking — a “thought calculator” — rather than a rival intelligence [^e9ff2088-b9c3-51c2-8d93-456b3cc1bccf@0].The material keeps circling back to what LLMs **can and cannot do** for the people exploring them:One core thread argues that **LLMs were not built for learning**: a plain chat model gives you answers and structure, but understanding still has to be done yourself. That's why the recommended workflow is *ask AI for structure first, then do the understanding yourself* [^e9ff2088-b9c3-51c2-8d93-456b3cc1bccf@0].A writer's perspective treats LLMs like ChatGPT as a new frontier that blurs the line between human and machine intellect — the fear is that AI can **write, reason, and even learn** at speeds and scales far beyond human capabilities, which unsettles a self-image built on intellect [^3e5745ab-279e-5154-9c22-9015dba70a32@0].At the same time, the handouts argue LLMs **amplify** rather than replace: a writer can generate drafts in seconds with ChatGPT, freeing them for strategy, creativity, and emotional depth; professionals in medicine, psychology, and philosophy use LLMs to summarize research and generate hypotheses while keeping the nuanced, high-stakes decisions for themselves [^3e5745ab-279e-5154-9c22-9015dba70a32@0].In coaching episodes, the LLM becomes a personal coach or life assistant — drawing on a long-term personal context the AI can reference — with clear caveats about what you should and should not hand over [^3269b2ab-ef7b-5c97-8d40-6498fd0cf1ca@12].- **LLM** here means the class of AI models behind ChatGPT and Notion AI — models that write, reason, and process text at scale.- The channel treats LLMs as *tools that augment thinking* — thought calculators, not rivals.- A recurring limit: LLMs were **not built for learning** — ask them for structure, then do the understanding yourself.- The term carries philosophical weight too: LLMs challenge our definition of human intellect and identity.