Back to portfolio
[Content Generation · [Cross-functional

Brett Content Generator

Brett-Style Content Generator

Thought-leadership content creation → a custom GPT imposes a distinct voice and structure → users get sharper drafts that feel more publishable.

01The Problem

Strong content is rarely blocked by ideas alone. The friction usually comes from turning vague expertise into writing that is clear, differentiated, motivating, and structurally sound. A blank chat model can produce competent text, but it often drifts into generic phrasing, weak positioning, and inconsistent structure. That creates editing overhead and makes it harder for a creator to produce content that feels recognizably theirs.

02What the AI Does

This is a custom GPT for content generation that is configured to mimic Brett VanTil’s writing style and produce engaging, insight-led business and personal development content. It generates drafts, rewrites content, structures ideas, and frames concepts so they feel motivating, differentiated, and actionable. It is built on GPT-5.4 Thinking and is not just a blank chat window with the same model. It includes embedded instructions that calibrate tone, structure, and output quality: start with a compelling hook, present ideas step by step, emphasize unique perspectives and results, and end sections with implementation questions. It is also instructed to ask follow-up questions when more detail is needed for higher-quality output. It can also use broader ChatGPT capabilities when relevant, including working with uploaded files, browsing for current information when needed, generating or editing images, and creating artifacts such as documents, slides, and spreadsheets. Those capabilities expand the formats it can support, but its core function here is prompt-shaped content generation rather than a deeply integrated business workflow.

03Design Decisions

01 · Choice

The GPT is narrowly scoped around content generation in a specific creator’s style rather than positioned as a general business assistant.

Why

A narrow scope usually improves consistency, voice fidelity, and usefulness over a general-purpose setup. The instructions strongly suggest the goal is not breadth, but distinctive authored output.

Constraint

This keeps the model focused on writing tasks and reduces drift into generic assistant behavior.

02 · Choice

The GPT is explicitly told to mimic Brett VanTil’s writing style.

Why

The likely judgment is that style consistency matters as much as factual adequacy for this use case, because the output is meant to feel like a recognizable extension of the creator’s voice rather than anonymous AI copy.

Constraint

This enforces voice calibration and makes outputs more brand-coherent, though it also narrows stylistic range.

03 · Choice

The instructions prioritize content that offers a fresh perspective and gets readers excited about possibilities for their business, life, or career.

Why

The creator appears to value content that does more than inform; it should reframe the reader’s thinking and create momentum.

Constraint

This sets a higher bar than “write something clear.” The model is pushed toward insight and inspiration, not just adequacy.

04 · Choice

The tone is calibrated to be professionally insightful, accessible, engaging, and motivational.

Why

This balances authority with readability. It avoids the tradeoff where expert-sounding content becomes cold or jargon-heavy, or motivational content becomes fluffy.

Constraint

The model is discouraged from using obscure jargon or sounding overly academic.

05 · Choice

Outputs are structured with a compelling hook, then a step-by-step breakdown, often mirroring “four characteristics plus a bonus” style sequencing.

Why

The likely reasoning is that structure improves clarity, reader retention, and reuse across formats like posts, articles, and frameworks.

Constraint

This reduces rambling and forces ideas into a repeatable editorial pattern.

06 · Choice

Each section is designed to end with an implementation question.

Why

The creator appears to want content that converts passive reading into self-reflection and action.

Constraint

This pushes outputs toward practical relevance instead of staying purely conceptual.

07 · Choice

The GPT is instructed to emphasize unique insights, differentiation, and competitive edge.

Why

[Creator: add rationale]

Constraint

This nudges the model away from commodity advice and toward positioning-oriented content.

08 · Choice

The GPT is told to focus on delivering results and organizing complex ideas into actionable sequences.

Why

The likely judgment is that good content should not just sound smart; it should help people do something with what they learn.

Constraint

This favors operational clarity over abstract exploration.

09 · Choice

The GPT is allowed to ask follow-up questions when it lacks enough context to generate high-quality content.

Why

The creator appears to prefer calibrated depth over false confidence.

Constraint

This is a quality safeguard against overcommitting on thin inputs.

10 · Choice

The GPT has access to broader tool capabilities such as file handling, web browsing, image generation, and artifact creation.

Why

[Creator: add rationale]

Constraint

While the toolset increases flexibility, the system-level guardrails still require honesty about uncertainty, source-checking for current facts, and staying within supported workflows.

11 · Choice

The GPT is told to report what it actually is and does, not inflate scope, outcomes, or usage claims.

Why

The portfolio prompt explicitly prioritizes credibility over marketing language.

Constraint

This prevents invented ROI, fabricated case studies, and exaggerated system claims.

04Tradeoffs & Limits

This GPT is strong at shaping tone, structure, and motivational business content, but that also reveals its limits. It will be weaker when the task requires original reporting, proprietary business context, or domain-specific claims that need evidence the user has not provided. Its style instructions can also become a constraint. If a user needs dry technical writing, legal precision, highly neutral prose, or a radically different brand voice, this configuration may oversteer toward inspiration, structure, and “fresh perspective” framing. It should not be trusted to invent metrics, outcomes, adoption figures, or client scenarios. The provided prompt explicitly forbids that, and the system design reinforces honesty about uncertainty and current knowledge. It is also not, by itself, a full operational workflow with confirmed integrations, memory of business systems, or access to real performance data. Even though it has access to useful tools, this project is best understood as a well-crafted custom GPT with tool access, not as an autonomous end-to-end business system.

05Key Insight

Useful AI implementation often comes less from model novelty and more from narrowing scope, codifying taste, and enforcing a repeatable standard for output quality.