Much of the time, we use tools like ChatGPT to answer questions.
Search engines like Google have trained us for decades to interact with the internet in this question-and-answer format.
“Where can I find the best price for an all-inclusive vacation in Cancun?”
“How do I make coffee with a percolator?”
We bring those same habits to models like ChatGPT, or hybrid tools like Perplexity.ai.
When not asking questions, we simply instruct these tools to write something, or design something.
“Write me an email promoting our latest product launch. Details below.”
“Create a photo-realistic image of two sheep dancing on a beach at sunset.”
What we don’t do very often is pause and ask the model to take the lead.
My own moment of epiphany…
A while back I was working on a presentation, and asked GPT-4 to outline the topic in 5 steps.
It did as I asked. But I wasn’t particularly happy with the result. So I paused and then asked:
“Never mind my requirement for a 5-step outline. How would YOU best organize the information in this presentation?”
The result was way better.
I’d stopped telling it what to do, but instead asked a question.
At that moment, we began to collaborate… a human intelligence working with an artificial intelligence.
How to work with AI tools in a more collaborative way.
I’ve often said that not a word in my blog is written by AI.
Today, I’ll make an exception, in the spirit of collaboration.
Up to now this post has been all me. The following 5 tips on how to work more collaboratively with AI were written by GPT-4.
1. Be Specific in Your Queries: The more specific your questions or prompts, the more likely you are to get useful and relevant responses. Ambiguity can lead to vague or off-target answers, so clarity in your inputs is crucial.
2. Iterative Interaction: Think of your interaction as a conversation where each response from the model builds on the last. If an answer isn’t what you expected, refine your question or ask follow-up questions to narrow down the information.
3. Utilize Feedback Loops: Use the responses you get to inform how you phrase future queries. Learning how to “speak” the language of the model—figuring out which prompts elicit the best responses—can greatly enhance your collaborative experience.
4. Explore Different Approaches: Don’t hesitate to experiment with different styles of prompts or inquiries. This can include asking the same question in different ways or using hypothetical scenarios to gauge different outcomes.
5. Patience and Adaptability: Developing an effective collaborative relationship with an AI requires patience and adaptability. The learning curve can involve understanding the best ways to interact with the model and integrating its strengths into your workflow.
Working collaboratively with AI puts you in a more positive place.
OK, I’m back.
A lot of people feel a little adversarial towards AI. All doom and gloom.
And if we simply use it like a search engine, our relationship is neutral, and passive.
But when we collaborate, that changes our mindset.
Like I said in a recent social media post…
“AI is a tool for enhancing human potential, not eclipsing it.”
In my own experience, I almost always get better results from AI when I collaborate with it.
And the act of collaboration puts me in a more positive place… and that feels good.
