“The hottest new programming language is English.” – Andrej Karpathy
Why It Matters
The number of programmers just increased from 26 million to 8 billion
Problem
You need to learn how to direct AI.
Solution
Prompt engineering helps you get accurate and reliable results.
Players
Prompt Engineers
Prompt Engineering Tools
Prompt Engineering Use Cases
- Programming • Write, test and debug code
- Writing • Find writing prompts, generate stories and perfect your writing style
- Research • Find, analyze and cite scientific papers
- Image Generation • Generate abstract and real-life objects and settings
- Search Engine Optimization • Research keywords and write SEO-optimized articles
- Marketing • Build marketing strategies
Predictions
- We’ll use text-to-audio models to create music. We’ll prompt the style, tempo, instrumentation and more.
- WavTool lets you use prompts to make, edit and compose tracks.
- MusicLM is an unreleased text-to-music model by Google that can generate consistent short and long tracks.
- Cassette lets you create royalty-free music by prompting genre, mood, length, instrumentation and more.
- Text-to-video will democratize animation and filmmaking. People will craft the story, lines, characters and locations. AI will take care of the visuals.
- Prompt hackathons will help to find new prompt writing techniques. This will boost creativity, innovation and adoption of prompt engineering.
Opportunities
- Use role-playing to get more accurate results for complex tasks. Pretending to be an expert at “topic” helps AI give you relevant, focused advice.
- Use advanced prompting. Help AI understand the context and perform better at complex tasks.
- Use constraints to get more targeted results. Prevent AI from going beyond the context of your tasks.
Risks
- Hallucinations • ChatGPT lies. It accused a law professor of harassment and cited a fake article as evidence. You may miss inaccuracies or inconsistencies if you don’t know what a quality answer looks like.
- Platform Risk • OpenAI’s rate limits can block you in the middle of solving problems.
Key Lessons
- A prompt consists of the task, instructions, context and parameters. “Task” is an action. “Instructions” are the steps needed to do the “task”. “Context” helps to understand the background and the end goal of a “task”. “Parameters” are settings that directly affect the result.
- Use detailed prompts to do complex tasks. Share specific context, format and style that the result should have. Justin Fineberg says if your prompt is only one sentence long, your output is never going to be good.
Hot Takes
Haters
“Prompt engineering is not programming.”
Backend-GPT is a language model that handles database and backend business logic. Similar to programming languages, you can prompt functions to work with data. Anyway, it’s too early to say that prompt engineering is programming. That’s why it’s a hot take.
“Prompt engineering will become obsolete once AI learns how to build its own prompts.”
We’re on track to this. AI can perfect itself without relying on human guidance. Tools like AutoGPT turn AI into an autonomous prompt building and execution robot.
“Tools built on top of APIs are just wrappers. They have no moat and can’t form a sustainable business.”
This critique has been around since the release of GPT-3. Danny Postma says if you can generate years of revenue in weeks or months, seize the opportunity.
“Marketplaces are winner-take-most markets. Why would we need more prompt marketplaces?”
These are micro-marketplaces which can optimize for niche use cases and audiences. While a micro-marketplace might not be the best option for most users, it's best for some. That's what matters.
Links
- Introduction to Prompt Engineering • Learn how prompts work.
- Framework for Reusable AI Prompts • Build reusable prompt templates.
- Awesome ChatGPT Prompts • 100+ role-playing prompts for ChatGPT.
Related Reports
