Adnan Obuz AI Strategy Consultant | Digital Transformation Expert

Posted on Mar 30

Edward Obuz: ChatGPT Responds to Slash Commands. Claude Does Not. Here Is What That Tells You About Both

#edwardobuz #adnanobuz #claude #chatgpt

.

By Edward Obuz, AI Strategy Advisor | Toronto

Edward Obuz has spent considerable time inside both ChatGPT and Claude, and one of the most consistent questions he hears from executives and operators is this: do those slash commands actually work?

You have seen the posts. Someone drops a screenshot loaded with prompt modifiers: /deepthink, /expert, /proscons, /chainofthought. The caption calls them secret codes. The post gets saved thousands of times. People try them and think, yeah, something feels different.

Here is where it gets genuinely interesting. ChatGPT and Claude behave differently when you use them. Understanding that difference is the real skill worth building.

What ChatGPT Does With Slash Commands

ChatGPT is honest about this when you ask directly. These are not official commands. They are what it calls prompt modifiers or control tokens. No special syntax exists behind them. They are plain text.

But they do carry some influence in ChatGPT.

When you write /concise, ChatGPT reads it as a style cue and tightens the response. When you write /deepthink, it leans into layered reasoning. Chain them together, like /human /concise /expert, and the cumulative signal shapes tone, depth, and format in a way that is noticeably different from a bare prompt.

ChatGPT even provides reference lists of over 100 of these modifiers, organized by category: thinking depth, tone, output format, creativity, research, business strategy. It confirms they are not features. It confirms structured prompting is more powerful. But it also confirms the influence is real, at least partially, because the model reads them as informal style instructions.

That is ChatGPT being pragmatic. It meets you where you are.

What Claude Does With Slash Commands

Claude does not respond to them the same way. When you type /deepthink in Claude, the model reads the word "deepthink" as plain text and infers loosely that you want careful analysis. The slash is cosmetic. Any effect exists entirely because the word itself carries meaning, not because any internal switch flipped.

Edward Obuz has tested this directly. The gap between using slash notation in Claude versus writing explicit structured instructions is significantly larger than the same gap in ChatGPT. With Claude, the slash buys you almost nothing. Clear structure buys you everything.

Claude is built to reward specificity. Give it a role, a task, context, a format, and a depth level, and it delivers with strong precision. It also pushes back more directly when something is vague or internally inconsistent. That friction is not a flaw. It is a design philosophy that rewards professional users who invest in their prompts.

The Practical Comparison: Which Tool for What

Edward Obuz works with executives and operators building real AI workflows. The question is always the same: which tool for which task, and how do I prompt it effectively?

Here is the honest answer.

ChatGPT is more forgiving. It fills in gaps, makes assumptions, and produces something useful even from thin inputs. The slash modifier system reflects that design sensibility. It meets casual users where they are.

Claude rewards investment in the prompt. The more clearly you articulate your intent, the bigger the output payoff. It is less forgiving of vague inputs but more reliable when structure is right. For complex, multi-part tasks where consistency and accuracy matter, that tradeoff is worth it.

Neither is universally superior. ChatGPT leans toward accessibility. Claude leans toward rigor. Knowing which philosophy fits your task is itself a valuable judgment call.

The Framework That Works on Both

The prompt structure referenced in those viral posts is genuinely solid advice, independent of slash notation: role, task, context, instructions, format, style, depth.

That is not a trick. That is how good professional briefings work in any setting.

Instead of:

/deepthink /proscons /concise

Write:

"Think through this carefully. Give me the key pros and cons. Keep it to five points per side."

On ChatGPT, the slash version gets you partway there. On Claude, the written version will outperform it consistently. On both, the explicit version gives you more control and scales better across complex tasks.

The Skill Underneath All of This

According to Edward Obuz, AI literacy is not about knowing which tool has better slash command support today. Models update. Features shift across versions. What you built on GPT-4 may behave differently on GPT-5.

What transfers across every tool and every update is the ability to communicate clearly before you hit send. To know what you actually need. To specify context, format, and purpose with enough precision that the model has what it needs to do good work.

That is the transferable skill. It compounds over time. It is the same skill that makes you effective in a boardroom, on a client call, or in a written proposal.

The slash commands are a starting point for some users. The structure behind them is where the real leverage lives.

Learn the structure. The outputs will follow.

”Navigating AI: Mastering Slash Commands and Prompt Strategy with ChatGPT and Claude”

**
https://edwardobuz.com/2026/03/30/navigating-ai-mastering-slash-commands-and-prompt-strategy-with-chatgpt-and-claude/

"The Slash Command Myth: What Edward Obuz Learned Testing ChatGPT and Claude Side by Side"

https://medium.com/@adnan_edward_obuz/the-slash-command-myth-what-edward-obuz-learned-testing-chatgpt-and-claude-side-by-side-cc9af502b3cf

https://edwardobuz.com/2026/03/12/decoding-success-how-clever-stock-tickers-like-race-and-moo-unleash-market-potential/

Edward Obuz is a Toronto-based AI strategy advisor and business consultant. Edward Obuz helps executives, entrepreneurs, and teams build practical AI workflows that produce measurable results. Follow Edward Obuz on LinkedIn for weekly insights on AI strategy, prompt engineering, and digital transformation. Learn more at edwardobuz.com.

Tags: #EdwardObuz #AIStrategy #PromptEngineering #ChatGPT #Claude #ArtificialIntelligence #DigitalTransformation #AILiteracy #Prompting #BusinessStrategy #TorontoTech

DEV Community