← All posts
Ai Writing

Ai Writing

The Hidden Cost of Claude's 'Thoughtful' Dialect for Founders and Content Creators

Claude's measured, nuanced dialect is a feature for legal analysis. For founder content, it's a liability. Here's what it produces by default — and the only fix that actually works.

The Phrases That Make AI Writing Detectable

AI-generated content is not detectable because it is robotic. It is detectable because every model has the same defaults — the same transition phrases, the same sentence constructions, the same rhetorical moves. Here's what they are and why they appear.

Anthropic Says Claude Has Functional Emotions. Here's What It Means for AI Writing.

Anthropic's interpretability team found 171 functional emotional representations inside Claude Sonnet 4.5. This isn't philosophy — it's measurable circuitry. And it explains a lot about why Claude writes the way it does.

Why Every AI Model Writes Differently (ChatGPT vs Claude vs Gemini)

ChatGPT writes like a confident LinkedIn coach. Claude writes like a philosophy professor. Gemini writes like a cautious corporate analyst. These aren't random variations — they're predictable dialects. Here's the full breakdown.

How to Tell If Something Was Written by AI (The Patterns to Look For)

AI detection tools are unreliable and easy to fool. But human readers don't need a tool — they can feel it. Here are the structural patterns that give AI writing away, and why they exist in the first place.

The AI Writing Patterns That Make Readers Tune Out (And How to Break Them)

AI writing doesn't fail because it's wrong. It fails because it's predictable. These are the structural patterns that signal 'this came from a machine' to readers — and the specific edits that fix them.

The Philosopher vs The Motivator: Which AI Model Should You Use for LinkedIn?

Claude writes like a philosophy professor. ChatGPT writes like a LinkedIn coach. These aren't random variations — they're deliberate design choices with real implications for what performs on LinkedIn.

Claude's Philosopher Mode: Why It Writes in Essays When You Want Punchy Posts

You asked for a three-sentence LinkedIn hook. Claude gave you a four-paragraph exploration of the concept's inherent tensions. Here's exactly why this happens — and how to work with Claude's architecture instead of fighting it.

The Complete Guide to Teaching AI Your Writing Voice

The hardest problem in AI writing isn't generating text — every tool can do that. It's getting AI to generate text that sounds like you. Here's the complete framework for training a persistent voice model that actually works.

How We Analyzed 10,000 AI Posts to Map Writing Dialects

The methodology behind Bloomberry's AI Dialects research — how we collected, labeled, and analyzed thousands of AI-generated outputs to identify the consistent structural patterns that differentiate ChatGPT, Claude, and Gemini.

How to Get AI to Write in Your Voice (Instead of Generic Content)

Most AI writing sounds like everyone else. Here's the exact step-by-step process to train AI tools to replicate your unique writing voice across LinkedIn, X, and blogs.

AI Tools That Write Social Media Posts Like You

A breakdown of AI tools that claim to write social media posts in your voice — what actually works, what doesn't, and how to evaluate voice replication quality.

Best AI Personal Branding Tools in 2026

A comprehensive guide to the best AI personal branding tools in 2026 — from voice replication to content distribution to analytics.

Agentic AI for Creators: When Your Assistant Starts Owning Outcomes

Agentic AI isn't "write me a post." It's goal-driven execution: generate, schedule, distribute, learn, and iterate. Here's what agents change for creators in 2026.

Voice Consistency: The Practical Checklist to Make AI Writing Sound Like You

A practical brand voice consistency checklist: sentence length, vocabulary, stance, "do not say" rules, and editing steps that make AI output sound like you—every time.

AI Assistant for LinkedIn: What It Should Do (And What It Must Never Do)

What an AI assistant for LinkedIn should do: draft in your voice, suggest angles, format cleanly, and schedule. What it must not do: invent, overhype, or flatten you.

How AI Is Changing Personal Branding for Founders

AI hasn't just made content creation faster — it has restructured what personal branding looks like for founders. Here's what changed and what it means for your strategy.

Your Digital Twin's Memory: What It Should Remember (And What It Shouldn't)

Your AI assistant shouldn't remember everything. Here's the best "memory" model for a digital twin: voice rules, topics, forbidden phrases, and safe personalization.

The "AI Slop" Problem: Why Most AI Writing Tools Make You Sound the Same

Most AI writing tools converge to the same voice. Learn why "AI slop" happens—and how to set constraints, voice rules, and edits that keep your writing distinct.

AI Ghostwriter vs. "Voice Clone": How to Keep Writing Human in 2026

AI ghostwriter or voice clone? Learn the difference, how to protect authenticity, and the exact inputs that keep your writing human (not generic) in 2026.

How to Use AI Writing Tools Without Losing Your Voice

AI can accelerate your writing. It can also flatten it. The difference comes down to how you use the tool — as a replacement or as an amplifier.

Looking for an AI writing tool that actually sounds human?

See how Bloomberry learns your voice →