Master writing prompts: use AI effectively for your brand

There are masses of articles out there that tell us how the world of X, Y, or Z industry is changing because of AI.

So where do you start?

Without doubt, AI is transforming the worlds of HR and internal communications.

I’m using it to generate content, to synthesize qualitative research findings using Otter.ai, Miro.com and Adobe. There are examples of AI streamlining processes like Enboarder, an AI-enabled tool for onboarding new employees, or Synthesia, a generative AI tool that turns text to video in minutes. The latter is a little cold, but effective.

Understandably, there will be aborted ideas along the way.

The automated job description generator Recrooit has already fallen by the wayside, although others are available. Perhaps this is an area where generative AI just doesn’t get it – it’s in the nuance.

During a recent client project, we used one of the world’s most powerful generative AI tools to create a film script for a big internal event.

The result was excellent, but the machine is still learning and it needed a little human intervention along the way.

Sometimes AI needs help.

Sometimes, it’s simply not the right moment to use AI.

We must always ask ourselves if AI is sacrificing the quality of experience at the altar of process excellence. And if it is, then it’s not good AI.

What’s key, for us as communicators and HR professionals, is understanding the basics.

There are three starting points:

✅ Prompts

✅ Content

✅ Data

AI tools are just that: tools.

That’s why an understanding of prompts is so important.

In effect, English (or your native tongue) has become the new coding language.

Because whether you’re using Firefly to create imagery; ChatGPT to create content; or Remesh for data, unless you know how to ask, you’ll get the wrong answer.

Total clarity is a must when you’re writing prompts!

So, this month, read this guide to understanding prompts, from the employee experience software provider Simpplr. If you’re going to use generative AI, you’ve need to get your prompts clear and succinct.

All this takes time. I know from bitter experience when creating images that the wrong prompt delivers the wrong result, and AI tools need to work with you over time to learn more about you – just as you will learn about it.

Creating content with AI

With content, whether you’re producing podcasts, film or the written word – for whichever channel – the secret is honesty and authenticity.

The Team’s newsletter is still created by hand, which is why you’ll spot a rare typo. But that’s because we believe humans still trump machines when it comes to certain communications.

There are plenty of times when using AI tools to quickly share information makes perfect sense. For example, when your communication is simply about sharing knowledge, insight and fact, then as long as you’ve double checked any claims being made – fill your boots!

But when it’s about capturing hearts and minds, tread with care.

I’ve mentioned Synthesia above. If you’re on a low budget, then such tools are a a great way to enliven communications – when appropriate.

Podcastle has the ability to turn text into a podcast.

Ad Creative can help you develop social media stories fast. You can move from key messages to fully formed stories fast.

They’re all there to play with, but beware. As with all content, you need to know where you’re headed first.

Data

Our friends at Remesh.ai are the gold standard for collecting data, insight and sentiment. For many years – long before the latest rush to AI – I’ve advocated that employee satisfaction surveys are dead.

It’s usually open for a month, data is analysed over another 4-weeks, then presented to senior executives, by which time it’s nearly three months old.

With tools like Remesh, communicators can host large-scale conversations with employees or customers that allow participants to see in real time the views of other users and vote on that content to give an accurate interpretation of sentiment.

And because it’s dynamic, facilitators can shift course to ask questions about sentiment.

Otter.ai and Adobe have both introduced tools that mean you can quickly look at the sentiment behind transcripts and all the data sat on your servers.

Think about it.

You have a vast array of PDFs, word documents, excel spreadsheets, and keynotes that you’ve built over time.

What if you could quickly search that intelligence to generate insights?

With tools like Virtual Brain, you can do exactly that.

We know that data and insight is everything. It should be driving all our communication efforts.

For longer reads, EY’s latest Tech Trend contains some excellent insights (jump to page 37 for a handy list of emerging AI-enabled HR tools.)

And for an even longer read, download IBM’s CEO guide to generative AI.