🔋10 Tips for Getting The Most out of ChatGPT

Ever feel like ChatGPT just isn’t giving you what you want? It can be frustrating to not get the results that you’re expecting out of such a powerful tool. That’s why we’re giving you the ultimate hacks for getting AI to do exactly what you want. Dive in to this week’s issue to perfect your AI prompts, and learn more about how the New York Times is taking a stand against AI.

What to Expect:

• Prompt Engineering 101: 10 Tips for Mastery

• New York Times Serves Perplexity AI a Cease and Desist

• Level Up Your Running Plan

• More AI Tech, Tools, and Talks

CONCEPT CORNER
Prompt Engineering 101: 10 Tips for Mastery

In the world of AI, prompt engineering is quickly becoming a must-have skill. But what is it? Think of it as talking to AI in its native language. In order to get the most accurate, meaningful outputs from an AI tool, you have to speak in a language that it understands… good prompts = good results.

Without it, your AI output can be vague, irrelevant, or just plain wrong.

📈 Companies have started to hire prompt engineers whose sole job is to craft effective prompts that will yield polished AI-generated results (According to a McKinsey survey of C-suite executives, 7% of respondents have reported hiring for this role).

Why its important: Whether you aspire to become a prompt engineer, or you just like using ChatGPT to help you with your latest task, thoughtful prompt design is critical for generating useful outputs. You may think you already know how to talk to AI, but we have a few recommendations on how to get those desired responses.

⬇️ Here are our 10 tips for mastering AI prompts:

1. Be specific: Vague prompts lead to broad or irrelevant answers. Be clear about what you're asking. For example, instead of saying, “Explain AI,” ask, “What are three benefits of AI in healthcare?” Specificity narrows down the AI’s focus, leading to more accurate and actionable responses.

2. Set the role: Define the role the AI should take to give more context for its response. By giving it a defined role, AI has a better idea of how to respond, ensuring the response aligns with what you’d expect from someone with expertise in the field. For example, “You are a successful Wall Street stock broker, explain how to create an effective investing strategy."

3. Context matters: AI thrives on context. Providing the necessary background helps the AI produce more relevant results. For instance, when asking for advice on a project, mention the project details—industry, goals, or challenges—to get tailored responses.

4. Use constraints: Setting clear boundaries for the AI can help control output quality. Specify requirements like word count, format, or style. For example, "Write a 150-word summary in bullet points" yields more focused and concise results than an open-ended request.

5. Ask for reasoning: Prompt the AI to explain its thought process. This boosts the clarity of the response and helps you understand the logic behind it. For instance, instead of asking for a decision outright, ask, “Explain why choosing option A is better than option B.”

6. Set a tone: AI can write in different styles, from formal and academic to casual and conversational. Mentioning a preferred tone ensures the response matches your expectations. For example, "Explain blockchain technology in a casual tone, like you're explaining it to a five year old,” or “Summarize this article from the perspective of a tax professional.”

7. Avoid vague language: Terms like "good" or "interesting" are too subjective. Instead, opt for clear, measurable criteria. For example, replace “What’s a good marketing strategy?” with “What’s an effective marketing strategy to boost online sales by 20%?”

8. Provide examples: Including examples within your prompt can clarify the expected outcome. For instance, saying “Create an email similar to this template…” helps guide the AI toward your desired structure or tone.

9. Break down complex tasks: If your task involves multiple layers, split it into smaller parts. For instance, instead of asking for a full market analysis in one go, prompt the AI to first gather relevant data, then analyze it, and finally provide recommendations. Avoid overloading the AI at once so that you can ensure a clear response.

10. Explain what went wrong : Prompts may not be perfect the first time. If a response doesn't meet your expectations, explain what was incorrect or lacking and ask for specific improvements. This helps the AI learn your preferences and provide better answers in subsequent attempts. For example, "That response was too general. Could you include more details about…”

AI NEWS
NYT vs. Perplexity AI: Cease and Desist 🚫

This week, The New York Times issued a cease-and-desist letter to Perplexity AI, a search engine that uses generative AI. The newspaper claims the platform is scraping its content without permission to fuel its AI-generated search results, sparking concerns over intellectual property rights.

• 📰 What’s at stake? The NYT argues that its content is being unfairly used to train AI models, potentially violating copyright laws. This action reflects media outlets' growing frustration with AI platforms that use their work without compensation (they’ve already sued OpenAI for the same concerns).

• 🧠 Why Perplexity AI? Perplexity uses advanced AI to generate responses based on search queries, but it relies on external content to do so. NYT alleges that this practice is unlawful since Perplexity has not entered into licensing agreements or paid for the use of its articles. However Perplexity says, “no one organization owns the copyright over facts.”

• 🛠️ Bigger implications: This case is part of a broader legal and ethical battle between media companies and AI platforms. As AI tools become more popular, the question of who owns the content these tools are trained on is becoming increasingly urgent. Content creators want to ensure they are fairly compensated for their work, especially as AI tools scale up. Yet, AI companies believe that there should be open access to information in order to ensure accurate insights.

• 📺️ What happens next? If the NYT’s legal action is successful, it could set a precedent for other media companies to demand compensation or block AI platforms from using their content. This could reshape how AI models access and use publicly available information moving forward, so stay tuned👀 

PROMPT OF THE WEEK
For: Runners

I am a runner who runs [X] times a week. My weekly mileage is [# miles] a week and my usual pace is [X] minutes per mile. I want to train [X] times every week. Create a training plan, based on heart rate, to run a [10k, half marathon, marathon, etc.] race. The race is in [#X] weeks and my goal time is [X].

TECH TOOLS, TIPS, AND TALKS

📖 What we’re reading: Machines of Loving Grace - An Essay about the Future of AI Written by Perplexity CEO, Dario Amodei
📻 What we’re listening to: Huberman Lab - How to Use Memory & Focus Using Science Protocols
💻 What we’re using: NotebookLM - An AI-powered personalized research assistant (stay tuned for next week’s issue where we will cover how to use this tool)

MORE READING

• Google Signs Deal to Purchase Nuclear Reactors - Another AI Giant Turns to Nuclear Energy

• AI Detectors Can Falsely Accuse Students of Cheating - How Schools are Working to Navigate AI and the Classroom

• OpenAI’s Former CTO Starts Fundraising for New AI Start Up - Will She Become OpenAI’s Newest Competitor?

We want to empower our readers with actually insightful knowledge so that they are more confident, informed leaders. Because let's face it, AI could be running the world pretty soon... so shouldn't we at least know how it works? If you are curious about a topic and want to learn more, drop us a message below👇🏼