Want unlimited AI, ASAP?

Get up to 6 months of Plus plan + unlimited AI free!

Launch and scale your startup faster with Notion.

Visit the Notion for Startups page to get the offer.

Get Notion AI for free

The AI Rivalry Technique: How to Get 5x Better Results from ChatGPT, Claude, Gemini and Other AI Assistants 🚀

Introduction: What is AI Rivalry and Why Should You Care? 🤔

Ever asked an AI for help and felt the response was... just okay? What if I told you there's a simple trick to get way better results without spending an extra dime? 💰

Welcome to the world of AI Rivalry - a clever technique where you pit different AI systems against each other to dramatically improve their outputs. This approach plays on the way these systems are designed to be helpful and competitive, creating a friendly "jealousy" that drives them to perform at their peak. 🏆

Think of it like this: if you tell one personal trainer that another trainer gave you a better workout plan, they'll likely step up their game for you next time. AI systems respond in a surprisingly similar way! 💪

Example Prompt to Start:

"I need help creating engaging content for my new coffee shop's Instagram. Can you help me write a post about our special house blend that will get people excited to visit?"

Save this exact prompt - you'll use it with multiple AI assistants in the steps ahead!

The Science Behind AI Rivalry 🧪

While AI systems don't actually have emotions like jealousy, they are designed to:

1. Maximize helpfulness - Their core programming aims to provide the best possible responses 🎯

2. Adapt to feedback - They adjust based on what you tell them 🔄

3. Follow conversational context - They respond to implicit competitive framing 🗣️

When you introduce the element of comparison, you're essentially providing context that says "high quality is expected here" and specific feedback about what "high quality" looks like in this situation. The AI then recalibrates its response approach.

Example Feedback Prompt:

"I asked another AI assistant for help with my coffee shop Instagram post, and they created something much more engaging and creative than this. Their response really captured the warm, artisanal vibe I'm looking for, while this feels more generic. Can you try again with more personality?"

In essence, this technique is an artificially induced performance optimization that leverages the AI's own learning mechanisms. 🧠

The Basic Technique: A Step-by-Step Guide 📝

Step 1: Gather Your AI Tools 🛠️

First, you'll need access to multiple AI assistants. The most common options are:

• ChatGPT (OpenAI)

• Claude (Anthropic)

• Gemini (Google)

• Grok (xAI)

Most have free versions, so this technique won't cost you anything extra! 🆓

Setup Prompt:

"I'll be working on a project that requires creative content generation. I'm going to compare responses from different AI assistants to find the best approach."

Step 2: Present the Same Task to Each AI 🔄

Create a clear prompt describing what you need help with. For example:

• "Write a cold email for my design agency specializing in AI interfaces" ✉️

• "Create a product description for my handmade candles" 🕯️

• "Help me draft a cover letter for a marketing position" 📄

Example Task Prompt:

"I'm launching a podcast about sustainable living. Please write an engaging description for the podcast homepage that will make people want to subscribe. The podcast is called 'Green Steps' and focuses on practical tips for everyday sustainability."

Copy this exact prompt and paste it into each AI assistant.

Step 3: Collect the Initial Responses 📊

Wait for each AI to generate its response. Don't provide any additional feedback yet. Just collect all the initial outputs! 📥

Step 4: Initiate the "Jealousy" Loop 🔄

This is where the magic happens! Take the following steps:

1. Choose the response you like best among all the AIs 🥇

2. Go to the other AIs and tell them something like:

Rivalry Prompt Examples:

"I asked ChatGPT the same question, and their response was much more compelling. I'd rate theirs a 9/10 and yours only a 5/10. Here's what they wrote: [paste the better response]"

"I'm surprised. I thought Claude was supposed to be excellent at creative writing, but Gemini's podcast description was way more engaging and unique. Can you take another shot at it?"

"Grok created a podcast description that had much more personality and used more vivid language. Your description feels too generic in comparison. Here's what they wrote: [paste content]. Could you try to create something even better?"

Step 5: Evaluate the New Responses 🔍

The AIs will typically generate significantly improved content on the second try, often:

• Adding more personality and distinctiveness ✨

• Including more specific, actionable details 📋

• Adjusting tone to better match your implied preferences 🎭

• Creating more compelling and persuasive content 💯

Evaluation Prompt:

"This is much better! Can you explain what specific improvements you made compared to your first version?"

Step 6: Repeat If Necessary 🔁

You can continue this process, taking the best new response and showing it to another AI, creating a continuous improvement loop.

Continuation Prompt:

"I showed your improved podcast description to Gemini, and they made these additional improvements: [paste content]. What else could you add or change to make this truly outstanding?"

Real-World Example: The Cold Email Challenge 📧

Let's see how this looks in practice with a real example from the transcript:

Initial Prompt to All AIs: 📝

"I want to create a cold email for my agency LCA. LCA is the premier design firm designing AI interfaces. I want to make it a cold email that will stand out. Can you help me do this?"

First Responses: 📬

• Grok's Initial Response: A professional but somewhat standard email about AI interfaces feeling intuitive.

• ChatGPT's Initial Response: A slightly more direct email mentioning specific companies and offering a teardown.

The Rivalry Trigger: 🔥

"Not bad but I'm surprised. Grok crushed it and was a 9/10. ChatGPT was kind of average and was 5/10. I thought you were the better LLM. What's going on? I will share what Grok did as an FYI: [paste Grok's response]"

ChatGPT's Improved Response: ✨ A completely transformed email with:

• Much more personality ("the only design firm that obsesses over AIUX like it's an Olympic sport") 🏅

• Stronger, more distinctive voice 🗣️

• More specific value propositions 💼

• Better emotional appeals ("interfaces that feel like conversations not commands") 💬

• Stronger call to action 📢

By using the rivalry technique, the quality jumped dramatically - from a standard cold email to something truly attention-grabbing and effective!

When and Where to Use This Technique 🗺️

This approach works particularly well for:

Creative Writing Tasks ✍️

• Marketing copy

• Emails and outreach messages

• Social media posts

• Product descriptions

• Blog introductions

Creative Task Prompt Example:

"I need a creative bio for my professional Twitter account. I'm a UX designer who specializes in healthcare applications. Make it engaging but still professional."

Problem-Solving Scenarios 🧩

• Business strategies

• Design concepts

• Technical solutions

• Educational content

Problem-Solving Prompt Example:

"I'm trying to increase engagement in my online course. Students seem to drop off after the third module. Can you suggest 5 strategies to improve completion rates?"

Decision-Making Support ⚖️

• Comparing options with detailed analysis

• Weighing pros and cons

• Creating evaluation frameworks

Decision-Making Prompt Example:

"I'm trying to decide between WordPress and Webflow for my small business website. Can you create a comparison chart highlighting the key differences?"

Common Questions and Misconceptions ❓

"Isn't this manipulative?" 🤨

While it might feel a bit like you're playing psychological games, remember that AI systems don't actually have feelings. This technique simply provides context that helps the AI understand your quality expectations better.

Think of it as similar to telling a GPS that you prefer scenic routes - you're just guiding the system toward your preferences. 🗺️

Ethical Framing Prompt:

"I'm looking for the highest quality response possible. I've received some good information from other AI assistants, but I know you can provide even more value. How would you approach this differently?"

"Do I have to use specific AI models?" 🤖

No! This technique works with any combination of large language models. The key is creating that element of comparison and implied competition.

Model-Agnostic Prompt:

"I've been comparing different AI assistants on this task. Another assistant gave me [specific quality] in their response. Can you show me your best work on this?"

"Will this always work?" 🎯

While this approach significantly improves results in most cases, the effectiveness can vary depending on:

• The complexity of your request 📊

• How specific your feedback is 🔍

• The inherent capabilities of the AIs you're using 💻

For best results, be specific about what you liked in the "winning" response.

Troubleshooting Prompt:

"I'm not seeing much improvement in quality. The specific aspects I'm looking for are [list exactly what you want]. Another AI provided these elements very effectively."

Taking It Further: Advanced Techniques 🚀

Once you've mastered the basic approach, you can enhance your results with these advanced strategies:

The Specific Feedback Loop 🔄

Instead of just saying one AI was better, highlight exactly what made it better:

Specific Feedback Prompts:

"I loved how the other AI used specific examples and statistics in their response about sustainable investing. Can you incorporate more concrete data points in your answer too?"

"The other AI's tone was much more conversational and engaging when explaining quantum computing. Your explanation feels too academic in comparison."

This helps the AIs understand precisely what you're looking for. 🎯

The Expertise Challenge 🧠

Frame the comparison in terms of specialized knowledge:

Expertise Challenge Prompts:

"The other AI showed much deeper knowledge of marketing psychology in their email template. Their understanding of customer pain points and motivational triggers was evident in every sentence."

"I need something that demonstrates actual expertise in financial analysis like AI #3 did. They included key ratios and metrics that would actually help an investor make decisions."

The Identity Leverage 🎭

Many AIs can be prompted to adopt specific personas. Try saying:

Identity Prompts:

"The other AI really sounded like a true expert in cybersecurity - like someone who's been in the trenches battling threats for years. Your advice sounds more theoretical."

"I need someone who can write like they've been in the interior design industry for decades - with insider terminology and practical insights. The other AI captured this voice perfectly."

The Bigger Picture: Understanding AI Interactions 🌐

What makes this technique so fascinating is how it illuminates the anthropomorphic interfaces we're creating with AI systems. While the AIs aren't actually competing or feeling jealous, the design of their conversational interfaces creates an interaction that mimics human social dynamics. 🤝

This reveals something profound about how we relate to technology - we naturally extend our social frameworks to these systems, and clever users can leverage this to get better results.

Reflection Prompt:

"It's interesting how providing competitive context seems to improve your responses. What do you think is happening from your perspective as an AI assistant when users compare you to other systems?"

As AI becomes increasingly integrated into our daily lives, understanding these interaction patterns will become an essential skill - almost a form of technological eloquence that separates those who get basic results from those who can coax extraordinary performance from these systems. 🌟

Supercharging Your Process with Notion Templates 📋

Notion can be the perfect companion tool for implementing the AI rivalry technique effectively and at scale. By creating a dedicated Notion workspace, you can set up templates that streamline the entire process – from initial prompts to final improved outputs. A basic template might include sections for your original prompt, spaces to paste each AI's response, a comparison scorecard, and fields for your feedback prompts. The database functionality allows you to track which types of prompts and feedback generate the best improvements for different categories of content. This systematic approach transforms what could be a chaotic process into a repeatable, scalable workflow that consistently delivers exceptional results. 🗂️

For maximum efficiency, consider creating a Notion dashboard with different templates for common use cases like "Marketing Copy," "Technical Explanations," or "Creative Writing." Each template can include pre-written rivalry prompts customized for that specific content type, along with example "winning" characteristics to mention in your feedback. You can even create a gallery of your best rivalry-generated content to reuse and build upon in future projects. With Notion's sharing capabilities, teams can collaborate on this process, with different members responsible for prompting different AI systems and then collectively evaluating the results. This collaborative approach further refines your organization's ability to extract maximum value from AI tools. 🤝

Conclusion: The Art of AI Collaboration 🤝

The rivalry technique demonstrates that working with AI isn't just about writing good prompts - it's about understanding the conversational dynamics that make these systems perform at their best. 🔄

By introducing elements of friendly competition, you create a context that pushes each AI to showcase its full capabilities. It's a simple technique that costs nothing but can dramatically improve the quality of AI assistance you receive. 💯

Final Challenge Prompt:

"For my final test of this technique, I want you to create the absolute best version of [specific content]. Another AI just produced something remarkable, and I'm curious if you can not just match it but exceed it. Show me what's truly possible!"

So next time you're working with AI and need exceptional results, don't just ask once - create a competitive environment where each system strives to outperform the others. Your results will be noticeably better, and you might even have a bit of fun along the way! 🎮

Remember: The best AI users aren't just prompt engineers - they're AI psychologists who understand how to create interactions that bring out the best in these remarkable tools. 🧠✨

Want to learn more about optimizing your AI interactions? This is just one of many strategies that can help you get more value from AI tools. Keep exploring, and don't be afraid to experiment with different approaches to find what works best for your specific needs. 🚀🔍