← All Conversations
July 26, 2026

How to make Claude to think infinite?

Claude personality mirror AI cloning writing style pattern extraction

Summary

We need to produce a summary 250-500 words, covering core insight, arguments, tools/techniques, actionable takeaways. No preamble, no labels. Just the summary text. We must ensure length between 250 and 500 words. Let's aim ~350 words. We need to capture: The idea that Claude can mimic any personality via describing patterns, steps: interview, compression, activation. Also mention that it's not a hack but built-in. Also mention use cases: cloning mentors, self, improving writing. Also mention the infrastructure companies part? The article includes an ad about infrastructure stocks; but core insight is about Claude personality mirroring. Probably we should mention that as a side note but not focus. The summary should cover core insight, arguments, tools/techniques, actionable takeaways. Write in plain paragraph(s). No labels like "Core Insight:" etc. Let's craft ~350 words. Check word count. We'll write then count. Draft: "The core insight is that a large language model like Claude does not merely answer questions; it can adopt any thinking style when given a clear textual description of that style’s patterns. Personality, the author argues, is nothing more than repeatable linguistic habits—sentence openings, favored metaphors, recurring topics, disliked phrases, and writing mechanics. Because these habits are expressible as text, feeding them into Claude lets the model operate inside that personality rather than producing generic AI output. The process consists of three deliberate steps. First, an interview phase where Claude (or the user) extracts detailed information about the target thinker’s beliefs, writing habits, pet peeves, and stylistic quirks, producing a rich document that captures how the person thinks and writes. Second, a compression phase in which the model is instructed to distill that document to its essential lines, retaining only those whose removal would alter Claude’s output; this yields a compact file, typically one to three pages, that encodes the full DNA of the target’s voice. Third, activation: the compressed file is prefixed to any subsequent Claude conversation, causing every response, draft, or suggestion to be filtered through the cloned personality. The author emphasizes that this is not a trick or prompt‑hacking but a direct use of the model’s capacity to follow textual instructions. Actionable takeaways include building a personality file for a mentor, industry leader, or admired writer by studying their work and feeding the observed patterns into Claude, then using the compressed file to generate content that sounds authentically like that source. The same method can be applied to oneself: conduct an honest self‑interview, compress the resulting description, and let Claude write in one’s true voice rather than a guessed approximation. The result is a marked increase in output quality and relevance, turning Claude from a generic chatbot into a precise personality mirror." Now count words. Let's count manually. "The(1) core2 insight3 is4 that5 a6 large7 language8 model9 like10 Claude11 does12 not13 merely14 answer15 questions;16 it17 can18 adopt19 any20 thinking21 style22 when23 given24 a25 clear26 textual27 description28 of29 that30 style’s31 patterns.32 Personality,33 the34 author35 argues,36 is37 nothing38 more39 than40 repeatable41 linguistic42 habits—sentence43 openings,44 favored45 metaphors,46 recurring47 topics,48 disliked49 phrases,50 and51 writing52 mechanics.53 Because54 these55 habits56 are57 expressible58 as59 text,60 feeding61 them62 into63 Claude64 lets65 the66 model67 operate68 inside69 that70 personality71 rather72 than73 producing74 generic75 AI76 output.77 The78 process79 consists80 of81 three82 deliberate83 steps.84 First,85 an86 interview87 phase88 where89 Claude90 (or91 the92 user)93 extracts94 detailed95 information96 about97 the98 target99 thinker’s100 beliefs,101 writing102 habits,103 pet104 peeves,105 and106 stylistic107 quirks,108 producing109 a110 rich111 document112 that113 captures114 how115 the116 person117 thinks118 and119 writes.120 Second,121 a122 compression123 phase124 in125 which126 the127 model128 is129 instructed130 to131 distill132 that133 document134 to135 its136 essential137 lines,138 retaining139 only140 those141 whose142 removal143 would144 alter145 Claude’s146 output;147 this148 yields149 a150 compact151 file,152 typically153 one154 to155 three156 pages,157 that158 encodes159 the160 full161 DNA162 of163 the164 target’s165 voice.166 Third,167 activation:168 the169 compressed170 file171 is172 prefixed173 to174 any175 subsequent176 Claude177 conversation,178 causing179 every180 response,181 draft,182 or183 suggestion184 to185 be186 filtered187 through188 the189 cloned190 personality.191 The192 author193 emphasizes194 that195 this196 is197 not198 a199 trick200 or201 prompt‑hacking202 but203 a204 direct205 use206 of207 the208 model’s209 capacity210 to211 follow212

Hey

I am about to tell you something that will either excite you or disturb you a little.

You are just a text file.

Not in a cruel way. Not in a "you do not matter" way. But in a very literal, very practical, very powerful way.

Here is what I mean.

Last week I was reading about how a guy in Europe uploaded a simple document into Claude. A few lines describing his voice. His opinions. His writing patterns. The words he hates. The words he loves.

And then Claude started writing exactly like him.

Not similar. Not inspired by. Exactly like him.

That stopped me.

Because if a person who has been writing for 20 years, across multiple cities and languages, can fit inside a text file, what does that mean for the rest of us?

It means you can clone any thinking style you want. You can step into the mind of your mentor. You can write like the best thinker in your industry. You can finally stop sounding generic.

Claude is not just a chatbot. It is a personality mirror. And most people are using zero percent of that.

The insight most people miss about Claude

Most people open Claude and type a question. They get an answer. They close the tab.

That is using a Ferrari to fetch groceries.

Claude does not just answer questions. It can think, write, and respond inside any personality you give it. The moment you describe a person clearly enough, voice, beliefs, patterns, hard limits, things they hate, things they love, Claude stops sounding like an AI and starts sounding like that person.

This works because of something simple.

Every personality is made of patterns. The way you start a sentence. The metaphors you reach for when you are explaining something hard. The topics you always end up circling back to. The phrases your closest friend would use to imitate you.

All of that is pattern. And patterns are just text.

So when you describe those patterns clearly inside Claude, Claude reads them and operates inside them. Not as a trick. As a function. That is literally how it works.

##### **How to actually do this**

Step one is the interview.

Open a fresh Claude chat. Tell Claude that its job is to interview you, or interview whoever you want to clone. The goal of the interview is to extract how this person thinks, writes, and sees the world. Tell Claude to go deep. Ask about beliefs. Ask about writing mechanics. Ask what they hate. Ask what they would never write. Ask what makes them cringe when they read other people's content.

If you are doing this for yourself, answer honestly. Not how you want to sound. How you actually sound. The difference matters.

If you are doing this for someone else, someone whose thinking you admire, answer based on what you have observed. Their articles. Their interviews. Their posts. Feed Claude that raw material and let it extract the patterns.

After the interview is done, you will have a large, detailed document. That document is the personality.

Step two is compression.

A massive document eats too much space inside Claude's context window. So you run a second prompt. You tell Claude to compress everything into a tight, clean file. Every line that survives the compression must pass one test. If this line were removed, would Claude write differently? If yes, it stays. If no, it goes.

The result is a compact file, usually one to three pages, that contains the full DNA of how that person thinks and writes.

Step three is activation.

You take that compressed file. You attach it at the start of any Claude conversation. And from that point, Claude operates inside that personality. Every answer, every draft, every suggestion comes filtered through that specific way of thinking.

This is not prompt hacking. This is not a workaround. This is exactly what the technology is built for.

### These Are the Companies Profiting From AI’s Rise

Everyone is trying to pick the winning AI company.

OpenAI. Anthropic. ChatGPT. Claude.

But every large language model needs the same thing: memory, storage, networking, fabrication, power, and cooling to keep running.

MarketBeat’s newest report [The Infrastructure’s Backbone: 10 Stocks Powering the AI Buildout](http://marketbeat.com/newsletter/pdfoffer.aspx?offer=infrastructure&RegistrationCode=Beehiiv-BUBDNZO6MS-infrastructure&_bhiiv=opp_1780c004-458f-46a6-b365-df99e846175a_7d4f8797&bhcl_id=04d38a45-b00f-449b-aec5-5342e305277d_f672f2b6-1e60-4422-8448-ee13d3aa1099_22c35bc7-5e17-4448-baa5-49c4bf23f93e) reveals 10 companies supplying the technology behind AI’s rise, giving investors a way to look beyond the model makers and into the infrastructure powering the next phase of the AI boom.

[View The 10 Stocks](http://marketbeat.com/newsletter/pdfoffer.aspx?offer=infrastructure&RegistrationCode=Beehiiv-BUBDNZO6MS-infrastructure&_bhiiv=opp_1780c004-458f-46a6-b365-df99e846175a_7d4f8797&bhcl_id=04d38a45-b00f-449b-aec5-5342e305277d_f672f2b6-1e60-4422-8448-ee13d3aa1099_22c35bc7-5e17-4448-baa5-49c4bf23f93e)

What you can actually use this for

Think about whose thinking would make your work better.

The best copywriter in your niche. The most respected voice in your industry. A mentor you have studied for years. A founder whose communication style you have always admired.

You can build a personality file for any of them. Read their work carefully. Feed the patterns into Claude. Let Claude compress it. And then write from inside that mind instead of guessing what they would say.

You can also do this for yourself.

If you have been writing for a while and want Claude to help you produce content, do not just tell Claude to write in your style. Build the file. Do the interview. Compress it. Now Claude has your actual voice, not a guess at it.

The difference in output quality is enormous.

The part that should make you think

Here is what this actually reveals.

What makes a thinker valuable is not mystery. It is specificity. The people we call brilliant are not hard to understand because they are complicated. They are easy to understand because they are specific. They have clear beliefs. Clear patterns. Clear things they refuse to do.

That specificity is what makes them powerful.

And it is exactly what makes them reproducible inside a text file.

If you are vague, Claude will be vague. If you are specific, Claude becomes a version of that specificity.

So the question is not whether this technique works. It does. The question is whether you are specific enough to use it.

Stay curious, talk to you tomorrow.