As Global Chief Innovation Officer at EY, Jeff Wong helps companies harness disruptive technology and prepare for the future of work.
getty
Non-technical people often ask me about the best classes to take to learn how to work with generative AI (GenAI). The answer I give usually surprises them: Most people don’t need any skills beyond those they already possess to become sophisticated users.
If you’re good at teaching, communicating and asking questions, you’re already well on the path to being able to use GenAI to solve problems, drive efficiency and bring new ideas to life.
Like exercise, a romantic relationship or many other things in life, you get out of GenAI what you put into it. The key to getting the most out of it is writing good prompts.
For people starting out with GenAI, I tell them to talk to it the way they would talk to a child.
Don’t Assume AI Knows—Explain It
If you’re going out to a fancy family dinner and you tell your child to "dress nice," who knows what you’re going to get? They may just throw on their best T-shirt and their least dirty pair of jeans. However, if you tell them to "wear a button-down shirt and black pants," you’ll likely get a better result.
We can apply the same concept to prompt engineering. "Create a social media post about our new app" will get you a result, but not one as good as "Create a social media post that explains why men aged 20-25 should download our new budgeting app."
Set Clear Boundaries
If you’ve ever asked a child to find a toy, their wallet or just about anything, you know they typically need a little guidance. Otherwise, they might wander the entire house until they forget what they were looking for in the first place. GenAI is similar in the sense that if you narrow the search area, you’re more likely to get what you want.
Instead of asking GenAI to explain how cryptocurrency works, say, "Using knowledge from renowned technology publications, explain how cryptocurrency works and why people use it." If you want to ensure you get a simple answer, you could further narrow your request by saying, "Pretend I’m a high school student. Using knowledge from…"
Essentially, knowing where not to look is as important as knowing where to look—especially as you engage with a large language model (LLM) drawing insights from massive datasets.
Have Patience
We’re used to computers being a lot like calculators; if you put in a question, you’ll get the same (or at least a very similar) answer every time. However, it’s important to remember that we’re in the very early days of GenAI, and the LLMs it relies on are still being fine-tuned. Just like when talking to a child, you should expect GenAI to get things wrong sometimes but improve quickly over time.
People underestimate how often humans of all ages get things wrong. When asking another person a question, a certain level of fallibility is accepted. We should expect the same from GenAI. Yet because of how confident it seems (and because of how well it answers some questions), we can get lulled into thinking AI is always right. The truth is that GenAI can be genius-level at answering some questions and quite poor at others—not unlike friends or family.
If you ask GenAI a question and are not sure about the answer you get, ask the system to explain itself. You could also ask the question a little bit differently. The process will make you better at prompting and will make GenAI better at answering. Also, as with kids, yelling probably won’t do any good.
Have Fun
Like playing with a kid, playing with GenAI at this stage of its development should be fun and exciting. It’s a way to view the world through new eyes and get a glimpse of what the future holds. Be creative not only in how you engage with GenAI but also in what you use it for.
Unlike many technical advances, GenAI requires very little technical knowledge to be used effectively. The best leaders will use this time to start training all of their employees on how to develop strong AI prompts so they can get the most from the technology as it develops. Those with patience, curiosity and strong communication skills already have a leg up.
Don’t overthink it, and don’t be intimidated. You already have the skills—just be clear, be patient and have fun.
The views reflected in this article are the views of the author and do not necessarily reflect the views of the global EY organization or its member firms.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
{Categories} _Category: Takes,*ALL*{/Categories}
{URL}https://www.forbes.com/sites/forbestechcouncil/2024/02/15/why-talking-to-children-is-great-training-for-working-with-generative-ai/{/URL}
{Author}Jeff Wong, Forbes Councils Member{/Author}
{Image}https://imageio.forbes.com/specials-images/imageserve/655516d102d93493722f7f19/0x0.jpg?format=jpg&height=600&width=1200&fit=bounds{/Image}
{Keywords}Innovation,/innovation,Innovation,/innovation,technology,standard{/Keywords}
{Source}All{/Source}
{Thumb}{/Thumb}