Keeping It Human in a Digital World (10)

We all know good human relationships depend on good communication. In fact, how well two people communicate is one of the best predictors of whether they can establish a successful and productive relationship.1 Surprisingly, your relationship with AI follows this same principle.

The better you communicate with AI, the better your results.

This is because AI was trained by ingesting vast volumes of content created by humans. In this way, it learned our language and a lot about the nuance, intent, and depth underlying our communication. And so now, getting the most out of AI means having a conversation that feels strangely human—but with a few important twists.

The Best Part? You Already Speak AI

The good news is that you already know the basics. You don’t have to take a class, you don’t have to learn to code, and you don’t have to look up commands in a user manual. We all can tap into the power of AI simply by speaking or writing in the same language we use every day, especially if we are English speakers.[i] Or, you can choose to communicate visually by drawing a sketch, sharing a photo, or even giving AI access to your camera so it can see your expressions, the room you are in, and the objects in front of you.

While AI is increasingly embedded in software that you may have used for years, the most direct way to experience frontier AI models today is through chatbots provided by their creators.[ii] This interface might be unpolished but connects you more closely to AI’s power.[iii] There are many chatbot options to choose from, and you can find links to these on my website.

The building blocks of AI communication are the requests, or “prompts,” we make into the system. An entire field called prompt engineering has sprung up to optimize these instructions. Prompt “influencers” have also proliferated online, hawking special formulas that can, at times, sound more like sorcery than conversation. Some of their approaches offer inspiration and are worth a try, but I’ve found that all the noise can be overwhelming to people. AI works best as a natural conversation; it’s rapidly getting better at understanding our intent on its own, and you can even ask AI for help along the way. For these reasons, most people learn fastest and get better results when they ignore formulas, charts, and spells and instead dive in, learning through hands-on experience and direct conversations with AI.

That said, there are a few fundamentals and techniques that make these conversations most productive, and this chapter walks you through both. I will be covering this in two sections that go hand in hand. First, we’ll talk about the fundamentals that underlie a productive relationship with AI. Then, I’ll share some practical advice on having useful dialogues with AI by taking you through prompting basics.

AI Fundamentals: Core Principles for a Productive Relationship

The journey to AIQ requires not only gaining experience in prompting the tools, but also understanding the underlying principles that help us tap into AI as a valuable collaborator. This section explores principles that shape a foundation for lasting success.

Treat AI as a Partner

Work with AI as if it’s a thought partner that learns about your needs and preferences over time.

One of the hardest things for people to wrap their heads around is how AI works best as a collaborator. Traditional software delivers a definitive and (assuming it’s well-tested software) correct answer, the same answer every time. Conversely, AI is always changing. It’s continually trying to understand what you want and refining its responses accordingly. It will even give you a different answer if you ask it the same question twice.

This is why feedback makes AI work better. Think of this like the back-and-forth that happens when you work with a team member to develop or hone a concept or project. It’s as if you are learning and exploring together.

AI works best when you have a “discussion” in which each party, you and the AI, continually refines “thinking,” gives or responds to feedback, and asks clarifying questions. My research revealed that a key reason newcomers fail to extract value from AI is that they frequently settle for an AI’s first response instead of engaging in deeper dialogue to achieve responses that add significant value to their work.

When a team member doesn’t get something right, you would tell them. Do the same with AI, and you will see it respond to your guidance. Experiment with probing how AI will approach a problem before you ask it to solve your challenge, just like you might with a respected colleague. Or ask AI to ask you questions to refine your thinking.

When you bring on a new team member, you may help them succeed by providing them with product manuals, samples of your writing style, or policies and training guides. AI is no different—the more background information you give, the more successfully it will respond. You can upload digital material in many formats directly to many AI chatbots.