Is Prompt Engineering Rocket Science? (Ref: Workshop-1 “WORLD PHARMA BRAND MANAGERS WEEK – 2024 on 20 January)
This blog is in continuation of Workshop-1 during “WORLD PHARMA BRAND MANAGERS WEEK – 2024” culminating day.
At the onset, my congratulations, and good wishes to all those who were present for Workshop # 1 “Medico-Marketing Copywriting 2.0: Mastering the Gen AI Advantage” of WPBMW – 2024.
One of the crucial points we discussed was about PROMPT ENGINEERING.
Read more in this blog.
At the onset, my congratulations and good wishes to all those who were present for Workshop # 1 during the “WORLD PHARMA BRAND MANAGERS WEEK – 2024” physical session!
The Topic: “Medico-Marketing Copywriting 2.0: Mastering the Gen AI Advantage”.
One of the crucial points we discussed was about PROMPT ENGINEERING in our Workshop – 1.
This morning I asked Bard “Can you tell me if prompt engineering is like rocket science and difficult to get help and guidance for accurate responses from Generative AI”?
Bard answered:
Prompt engineering is not quite as complex as rocket science, but it can be challenging to master. It’s like fine-tuning a high-performance race car. You need to understand the mechanics of the language model (the engine) and how to tweak the prompts (the controls) to get the desired results.
Here’s a breakdown of the difficulty level:
Easy:
- Basic prompts: Asking simple questions or giving the AI clear instructions like “Write a poem about a cat” or “Summarize this article for me” usually doesn’t require much expertise.
Medium:
- Refining prompts: Getting more specific with your prompts, like “Write a poem about a grumpy cat in a hat” or “Summarize this article, focusing on the author’s main arguments,” requires some understanding of how the AI works and what kind of prompts will produce the best results.
Hard:
- Creative prompts: Asking the AI to do something open-ended or unusual, like “Write a song in the style of Queen about a robot uprising” or “Generate a code snippet that can solve this math problem,” can be tricky. You’ll need to experiment with different wording and techniques to get the AI to understand your intent.
Expert:
- Fine-tuning prompts: Really pushing the boundaries of what the AI can do, like “Write a script for a movie sequel that no one ever made” or “Create a new scientific theory based on this data,” requires a deep understanding of both the AI and the domain you’re working in.
As for getting help and guidance, there are a few resources available:
- Online communities: There are several online communities for prompt engineers, like the r/PromptEngineering subreddit, where you can ask questions and share tips with other people.
- Blog posts and articles: There are also many blog posts and articles written by experts on prompt engineering, which can teach you the basics and give you some advanced techniques.
- Books: There are a few books available on prompt engineering, such as “Prompt: Building AI Models with Human Brains” by Brian Christian and Tom Griffiths.
Ultimately, the best way to get better at prompt engineering is to practice and experiment. The more you play around with different prompts and see how the AI responds, the better you’ll understand how to get the results you want.
So, while prompt engineering isn’t exactly rocket science, it does take some time and effort to learn. But with the right resources and a bit of practice, you can unlock the full potential of Generative AI and get it to do some amazing things.
I asked the same question to Chat GPT 3.5. The response:
“Prompt engineering in Generative AI is not as complex as rocket science, but it does require careful consideration. It involves crafting prompts to elicit desired responses. While it may not be as challenging as rocket science, achieving accurate and nuanced responses can still be tricky. Fortunately, there are online communities, forums, and resources where you can seek guidance and assistance to improve your prompt engineering skills.”
And the response from Bing?
Explore
Prompt engineering is the process of writing, refining, and optimizing inputs, or “prompts”, to encourage generative artificial intelligence (AI) systems to create specific, high-quality outputs. Generative AI is a type of AI that can learn from and mimic large amounts of data to create content such as text, images, music, videos, code, and more, based on inputs or prompts.
Prompt engineering is the process of writing, refining and optimizing inputs, or “prompts”, to encourage generative artificial intelligence (AI) systems to create specific, high-quality o
Prompt engineering is not rocket science, but it does require some technical knowledge, creativity, and persistence.
There are many resources available to help and guide prompt engineers, such as tutorials, articles, books and online courses. Some examples are:
- Prompt Engineering for Generative AI, a comprehensive guide by Google for Developers that covers different types of prompts, best practices, and strategies for effective prompt engineering.
- What Is Prompt Engineering? Definition and Examples, a Coursera article that explains what prompt engineering is, how it works, and how to use it for different tasks and domains.
- Generative Artificial Intelligence (AI), a Harvard University website that provides an overview of generative AI, its applications, challenges, and limitations, and links to various tools and datasets for prompt engineering.
- Generative AI Defined: How It Works, Benefits and Dangers, a TechRepublic article that discusses the basics of generative AI, its benefits and risks, and some of the most common and innovative use cases for this technology.
I hope this helps you understand prompt engineering and generative AI better. If you have any more questions, please feel free to ask me. 😊
To summarise, the importance of learning prompt engineering is individual and depends on your specific needs and expectations for Gen AI.
However, regardless of your decision, understanding the basics of prompting can still be beneficial.
It allows you to communicate more effectively with the AI, ask more effective questions, and get the most out of your interactions.
In summary, the decision to learn prompt engineering is up to you. Consider your desired level of control, technical comfort, and Gen AI integration to determine if it’s worth the investment.
There’s no one-size-fits-all answer, but hopefully, this information helps you make an informed decision.
The last chapter of the book “AILMENTS IN MEDICO-MARKETING COPYWRITING: THE Rx” has a list of many learning resources.
My final question to Bing Image Creator: if “Prompt Engineering” were a human being, can you create a photo of what would it look like?
The photos you see in the title are the response from Bing.
Good luck and happy reading.
This blog wouldn’t have been written without guidance from my young student Yuvraj Patil