If you scroll through your social media feeds, you might get the sinking feeling that everyone else is miles ahead of you in the world of artificial intelligence. It feels like everyone is building apps, automating their entire lives, and generating content at the speed of light. But I just came across a breakdown from an AI expert that suggests this anxiety is misplaced.
This savvy professional shared some startling statistics that completely reframe the narrative. According to the data the author presented, 84% of humans have never used AI at all. Not once.
The creator of this post used a brilliant visualization to drive the point home. Imagine a grid where each square represents 3.2 million people. To represent the 8.1 billion humans on Earth, you would need 2,500 squares. The vast majority of those squares are gray, representing people who have zero experience with these tools. A smaller portion is green, representing free users. And the red squares? Those represent the people who actually pay for these services.
You can barely see the red section. It represents just 0.3% of the population. If you are reading this, you are likely in that tiny minority. The author points out that your timeline is a bubble. While your digital peers are posting about AI, most people in the real world, your parents, your neighbors, anyone outside the tech sphere, likely haven’t even tried ChatGPT yet.
However, the expert notes that while adoption is low, the technology itself is accelerating at a frightening pace. We are looking at a 1.85x acceleration in capability in just one year. To stay ahead of this curve, the post’s author suggests shifting focus from standard chatbots to more advanced workflows using Claude. Here is the step-by-step process the expert outlined to move from a casual user to a power user.
1. Set Up the “Cowork” Environment
The first shift the author recommends is moving away from the browser-based, blank chat interface. The goal is to integrate the AI directly into your computer’s environment so it can understand your context.
- Go to the Anthropic website and download the Claude desktop application.
- Select a specific folder on your computer to share with the AI.
- Drop markdown files (.md) that contain information about you or your projects into this folder.
- Use the following prompt to initiate the session: “Read files & question me before starting.”
The expert emphasizes that this method allows the AI to “live” on your computer. By reading your local files, it gains context that a blank browser window simply cannot provide. This turns the AI from a generic chatbot into a knowledgeable coworker that understands your specific history and data.
2. Activate Extended Thinking with Opus 4.6
Most users stick to the default models, but the innovator behind this post argues that you are leaving performance on the table if you do. The newer models have capabilities that go far beyond simple text generation.
- Click the model selector in the interface.
- Choose “Opus 4.6” from the list of available models.
- Toggle on the “Extended Thinking” feature.
- Run your usual prompts and observe the difference.
I found this part particularly interesting. The author notes that enabling Extended Thinking forces Claude to “reason first.” This means the AI spends time planning and checking its logic before it starts writing the answer. The result is a completely different depth of output, making it suitable for complex problem-solving rather than just quick answers.
3. Utilize Official Plugins
Stop prompting from scratch. The original poster highlights that Anthropic has released official plugins that streamline complex tasks. This feature is powerful enough that the author mentions it caused a massive valuation drop in traditional legal software companies shortly after release.
- Navigate to the plugins section of the Claude interface.
- Browse the available options and pick one that aligns with your specific job function.
- Type the forward slash key (/) in the chat box to reveal the “slash commands.”
- Select the appropriate command to execute specific workflows instantly.
This is a massive efficiency hack. Instead of spending five minutes crafting a prompt to explain what you want, you can use a slash command to trigger a pre-built, optimized workflow. It standardizes your output and saves significant time on repetitive tasks.
4. Create a Custom Writing Style
One of the biggest giveaways that text was generated by AI is the generic, flat tone. The expert suggests that you should never settle for the default “AI voice.” Instead, teach the model to sound exactly like you.
- Go to Claude and click the “+” icon to create a new custom style.
- Upload examples of your actual writing (emails, articles, reports).
- Save this as your personal style profile.
- Select this style profile before starting every new chat session.
The author calls this the “Anti AI” approach. By feeding the model your own syntax and vocabulary, the output stops sounding robotic and starts sounding authentic. This is critical for anyone using these tools for communication or content creation, as it preserves your unique voice.
The reality is that 84% of the Earth doesn’t know these features exist. You now possess knowledge that puts you in the top tier of global users. As the creator of this post succinctly put it: You know this exists. Now act like it.
I was genuinely motivated by this perspective. It’s easy to feel overwhelmed by the pace of technology, but realizing how early we actually are is incredibly empowering.
Check out the full post for more details.