Five Levels of using Large Language Models
Anyone can utilize a large language model (LLM) to enhance their productivity.
While I work with large language models (LLMs) during the day, I believe that most people in the world are still unaware of what they truly are. Here is my mental model of the current understanding of LLMs among the general public, along with my thoughts on the matter so far:
Level 1: Google (60%)
As Google began to roll out AI Overviews, you would already be utilizing a large language model directly from Google, with no prompting required. If you understand what prompting is, you are already at level 2.
Level 2: ChatGPT (20%)
The household name that brings large language models (LLMs) into everyone’s lives. You can either visit the newly launched chat.com, which cost $20 million, or use the ChatGPT app on your phone. You can ask it intriguing questions and engage in conversations with it almost every day.
Level 3: ChatGPT Killer (10%)
Claude, Perplexity, or any of the other countless tools available today can be utilized for specific use cases. You may choose a particular AI tool tailored to your needs, or you might find that your regular software has incorporated an AI feature that assists with rewriting. Some tools even allow you to customize your own AI.
Level 4: Some Low-Code Tools (4%)
You can utilize a large language model (LLM) by copying and pasting API keys to enhance an application you discovered online. Not only can you use a specialized LLM service, but you can also create your own, similar to custom GPTs on ChatGPT.
Level 5: Utilizing the API and Developing Agents (1%)
You would utilize a large language model (LLM) by copying and pasting API keys. Instead of relying on the frontend, you would examine the results of the API calls. Additionally, you would employ various frameworks to develop AI systems capable of performing specific tasks.
