AI for Communicators

More than 100 people gathered at the University of Illinois Springfield on Thursday, October 23, for the AI for Communicators daylong workshop. Featured speakers were Andy Crestodina, who made presentations in the morning and afternoon; Jenna Marble, who presented at lunch about AI as a graphic design tool; and Brianna Miller, who conducted an afternoon session designed for people interested in healthcare communications.

Below is a synopsis of what each speaker had to say. Full videos are available to everyone who attended the workshop. Other resources, such as prompts, also are available. A separate email went to attendees in early November about this.

Andy Crestodina

According to Andy, AI is a giant test of our imaginations. Marketing and sales are at the center of AI opportunity.

He described AI as a “skills leveler” among less-educated and more-educated people. He showed slides depicting different phases/evolutions of how we can use AI. By the end of today, he said, we will have progressed in our own evolutions.

“This is not a moment to wait for someone to train us,” he said. “We can turn this into an era of opportunity.”

He asserted that it’s better to use AI for quality and performance, as opposed to using it for efficiency (e.g., to replace other people’s jobs).

He said he is very focused on “using AI to help humans do what ONLY humans can do.” He discussed the concept of writing better prompts. Better prompts = better outputs. He encouraged “talking through” your prompt – in other words, talk to the AI.

He advised us to save good prompts (effective ones that we’ll use in the future) in a “prompt library.” That could be a simple document in Google Drive or whatever method works for you.

He said he mostly doesn’t use AI for writing. He does use it to write YouTube descriptions.

Andy also asserted that “press releases don’t belong on websites,” and neither do PDFs. He’s put together an article, “13 Things to Remove from Your Website Immediately,” on the Orbit Media Studios blog and on YouTube. 

AI doesn’t care and doesn’t have opinions, Andy pointed out. It CAN help us find provocative topics. 

Testing is crucial with AI because sometimes it makes up things (a.k.a., hallucinates). He always tests and validates what AI generates, even if it takes extra time. He spoke about using AI to help write reusable prompts and again warned not to just accept its output without evaluating it for yourself, drawing on your own brain power.

Draft a prompt → Ask AI to revise it →Test prompt. After testing the prompt, either ask AI to revise again, or decide that is your final, reusable prompt.

While he didn’t endorse one AI tool over another, his presentation focused on ChatGPT because that’s what he uses most. He mentioned the “Customize GPT” option, which he uses to tell AI NOT to use specific “AI-typical” words or phrases, such as “unleash” and “treasure trove.”

For ChatGPT to write in your voice, you have to provide it with many examples of your past work. Even so, you’ll still need to evaluate and edit what it generates, and you might have to do this repeatedly. Similarly, AI will provide the sources it used, if you ask it to do that. You’ll still need to check them for yourself to make sure they’re real sources.

In the afternoon session, Andy dug deeper into how communicators can use AI as a tool. For example, he recommended using AI to help find sound bites, which are good for social media promotion.

Andy said he doesn’t want to use AI to find “efficiencies,” but rather to find “deficiencies.”

Jenna Marble

Jenna’s presentation, “Amplify, Not Replace,” focused on generative AI and graphic design.

She specified that what she’s talking about today is “generative AI,” where something new – images, code, music, etc. – gets created.

As a teacher, she starts by teaching students the foundations of graphic design and doesn’t get them involved with AI right off the bat.

She said she tries to resist the “fear and the hype” with AI. She teaches students how to use the tools.

She highlighted a World Economic Forum report about the future of work. It cites AI as transformative, yet still identifies creative thinking, resilience, and other skills that will make a difference in the next 5 years of work.

“Human creativity is more valuable than ever,” she said. 

Then she asked: What’s creative thinking? Examples include asking questions, innate curiosity, and storytelling are key.

Jenna said that if you’re not thoughtful about how you interact with AI, you can get terrible results. (She showed visuals from the abysmal “Willy Wonka” event in Scotland.)

She also cited an article from The Pudding – a data visualization site – called “Can AI make a data-driven, visual story?” The Pudding used Claude in this experiment.

Jenna’s takeaways on the various case studies she presented: 

1. Start with specific intention, clarity and strategy. 

2. Stay grounded in brand values, not just the novelty of AI. 

3. Human intervention and creative thinking are critical.

She concluded by advising everyone about what to look for in a responsible AI tool. She recommended Adobe Firefly for text-to-image.

Remember that you control the tool. Be transparent with clients about using AI. She displayed QR codes with additional resources.

Brianna Miller

Brianna presented a breakout session in the afternoon called “AI Playbook for Healthcare Marketing.”

There are many AI tools, and we’re trying to figure out what’s best and in what situations.

If you’re in the healthcare industry, don’t be surprised that a lot of people have reservations about using AI and about spending money on AI tools. These reservations often revolve around patient care and privacy.

We have to teach AI a lot in the healthcare space, such as maintaining tone and brand trust, avoiding generic content, and maintaining privacy and compliance requirements – e.g., protecting patient data. No patient data ever should go into an open AI tool.

Brianna discussed how to use AI in marketing. Don’t use it for everything. Find where using it makes the most sense. Important points on her list included “Start small with high-impact, repetitive workflows” and “Set guard rails around approved data, tone and approved use cases.”

Brianna offered advice on how to help AI develop the “voice” that you need. Develop your tone, style and personality. For example, would you call it warm and friendly? Give AI examples of your past messaging so it can try to replicate that voice.

She displayed a QR code leading to a workbook about using AI in human-centered healthcare marketing and went through several sections of the workbook.

She also compared weak prompts vs. strong prompts and discussed how to make a custom GPT.

Summary: All three speakers provided a wealth of information and encouragement, calling on everyone to keep experimenting and practicing with AI to see its capabilities, its weak spots, and its evolution. They also reminded everyone that humans/human abilities remain vital and that AI is a tool. It’s up to us to decide how to use it, and to evaluate its effectiveness.

Event attendees can access workshop resources, using the password sent via email.

ATTENDEES: Get Resources