In today’s newsletter:

1. 📖 Prompt Engineering & the Different Types of Prompting

2. 🤖 Biden’s new Executive Order about AI (Plus, Other Countries Worried about AI)

3. 🎤 One prompt you can use at work today

Read time: 5 minutes

1. Prompt Engineering & the Different Types of Prompting

I recently attended a great talk by thought leaders in the AI space.

One of the speakers was Elvis Saravia (founder and lead AI scientist at DAIR).

He wrote a comprehensive guide called the “Prompt Engineering Guide” that covers the topic in-depth.

Prompt engineering is this idea of tweaking and optimizing your prompts (i.e., the questions you ask a tool like ChatGPT) to improve the output that you get.

Some AI experts believe that this whole discipline of prompt engineering will eventually go away because AI tools will keep getting smarter.

But in the meantime, it sounds like this technique is needed to maximize the value we get from those tools.

In his guide, Saravia covers different techniques of prompting, including:

• Zero-shot Prompting:

• Few-shot Prompting

• Chain-of-Thought Prompting

• Self-consistency

• Retrieval Augmented Generation

You can read the full guide (including a description of those definitions) by clicking on the following link: https://www.promptingguide.ai/ 

What does this mean for leaders?

A couple of takeaways about this.

First, as a leader, it’s a good idea to familiarize yourself with those prompt engineering terms because they’ve been popping up a lot recently, and they can help you with understanding whether you should adopt a particular strategy.

Second, keep in mind that this just reinforces the concept that there is still no business playbook to apply generative AI at work. The key is to just start using it and learn from the feedback.

Ethan Mollick wrote a great post about the two paths to prompting when working with AI, and he summed it up pretty nicely with the following comment:

2. Biden’s new Executive Order about AI (Plus, Other Countries Worried about AI)

President Biden recently issued a new “Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence” which highlights the importance that the US government is placing on this field.

You can read the full exec order by clicking here.

And here’s a summary post by the White House on their X account:

What does this mean for leaders?

Here are 8 main takeaways from the exec order for leaders:

• Risk Mitigation: Implement standardized testing and risk management for AI safety and legal compliance.

• Innovation and Competition: Support a fair AI marketplace, address anti-competitive practices, and protect small businesses.

• Workforce and Rights: Train workers for AI-era jobs and deploy AI responsibly to enhance workers' rights.

• Equity and Civil Rights: Prevent AI from exacerbating discrimination and ensure inclusive practices.

• Consumer Protection: Uphold consumer rights against AI-related fraud and bias, especially in critical services.

• Privacy: Secure personal data and employ privacy-enhancing technologies in AI applications.

• Federal Governance: Align business AI practices with evolving federal workforce and governance standards.

• Global Standards: Adhere to international norms for responsible AI as the U.S. leads in setting global standards.

Overall, business leaders should prepare for a landscape where AI is tightly integrated with ethical considerations, safety standards, and global leadership initiatives.

How seriously is the rest of the world also taking AI?

Very.

There was an AI Safety Summit on November 1, 2023 where more than 25 countries (including the US, UK, EU, and China) signed a “Bletchley Declaration” to establish a close working relationship among those countries.

3. One Prompt You Can Use at Work Today

Here’s a practical ChatGPT Prompt you can use at work:

For example,