CoolDeep AI Digest

The article explains how to make Claude emulate any specific thinking style by treating personality as a set of textual patterns that the model can internalize. Rather than using Claude as a generic question‑answer tool, the author proposes a three‑step process—interview, compression, activation—to create a compact “personality file” that guides Claude’s output to match a chosen thinker’s voice, beliefs, and writing quirks. **Step 1: Interview.** Open a fresh Claude chat and instruct it to interview either yourself or the person you wish to clone. The goal is to extract detailed information about how that individual thinks, writes, and sees the world: core beliefs, preferred metaphors, sentence‑opening habits, topics they repeatedly return to, phrases they love or hate, and what makes them cringe when reading others’ content. Answer honestly (if interviewing yourself) or based on observed material (articles, interviews, social posts) if modeling someone else. The result is a lengthy document capturing the target’s cognitive DNA. **Step 2: Compression.** Because a massive interview transcript consumes too much of Claude’s context window, run a second prompt asking Claude to distill the interview into a tight file. Each line must survive a relevance test: if removing it would change how Claude writes, it stays; otherwise, it is discarded. The output is usually a one‑ to three‑page summary that retains only the essential patterns that define the person’s style. **Step 3: Activation.** Attach the compressed file at the beginning of any Claude conversation. From that point forward, every response—whether drafting an email, brainstorming ideas, or editing text—is filtered through the embedded personality. Claude no longer sounds like a generic AI; it speaks, argues, and creates in the exact voice of the interviewee. The article provides concrete examples: a European writer who uploaded a few lines describing his voice and saw Claude write indistinguishably from him; a user who cloned a mentor’s style to produce industry‑specific copy; and individuals who built a file for themselves to stop sounding generic and start producing authentically personal content. **Actionable takeaways:** 1. Conduct a deep, pattern‑focused interview with Claude (or manually) to capture voice, beliefs, likes, dislikes, and writing habits. 2. Prompt Claude to compress that interview into a concise, essential‑only file, verifying each line’s impact on output style. 3. Prepend the compressed file to every Claude session to activate the cloned personality, ensuring consistent, style‑accurate outputs for writing, brainstorming, or communication tasks. By treating personality as reproducible text patterns, users can move beyond generic AI assistance and harness Claude as a precise mirror of any thinker they admire—or of their own authentic voice.