At the recent AMPLIFY conference in Washington in June, artificial intelligence seemed to be topic one. Even in sessions on other media issues, AI was part of the undercurrent.
One of the most sharply focused sessions was titled “How AI Will Impact Digital Marketing In 2023.” The presenter was Alan Hirsch, then VP of Sales and Marketing for Contentware, an AI-based marketing tool. The session covered how AI will create marketing messaging, help zero in on the right audiences, and help manage campaigns. Hirsch has since moved on to knk Publishing, a platform that unifies the management of various publishing operations, from finance to sales, marketing and content.
Fox Tales was struck by several things Hirsch emphasized, including the importance of media brands having experts vet AI content, and the importance of style and voice. We recently interviewed Hirsch to further illuminate these things and more. Here’s a transcript of our conversation.
Fox Tales: To start with some of the points of your session, how will AI assist in automated marketing messaging?
Alan Hirsch: AI, right now, can create complete posts, including imagery—not just text—for email and social media. A tool like Contentware can create 40-100 unique posts for email and social, all optimized for each channel. Review is certainly required, but the AI has gotten so good, there’s often very little editing to be done.
Fox Tales: How specific can we expect such AI functions to be and how widespread is its use now?
Hirsch: I would say that right now, adoption is not widespread at all. I’m not sure what people are waiting for! AI is nowhere near perfect, but people need to realize the AI you use today will be the worst AI you ever use. It continues to improve at warp speed. There’s so much value in the time saved from using AI right now that it’s more than worth it to use, even if it’s not perfect.
Fox Tales: Moving to what we mentioned above, why should media brands have experts vet AI content before accepting it for publication?
Hirsch: Oh, review is absolutely essential. It still can make terrible mistakes. First and foremost, you should not be using AI for areas about which you don’t know a lot. It would be like doing a math calculation and not knowing the order of magnitude to check your work (is the answer 10,000 or 10,000,000?) And, very obviously, it’s not human. Our job is to take AI content and make sure it’s “human.” AI lacks wisdom and humor, to cite two good examples. We humans are responsible that the content we publish, AI generated or not. It speaks to our brand.
Fox Tales: Can AI be trained in style and voice?
Hirsch: It absolutely can be. But you will still need to review the content for voice and tone.
Fox Tales: Talk about the most common pitfalls, and how to minimize them using Contentware or other products and processes.
Hirsch: I’d say the most common pitfall is people use it once or twice, don’t like the result, and give up. It takes a small amount of time to learn how to use these tools, which is why people should be starting now. AI is a tool. You have to learn how to use it. The second biggest pitfall: They haven’t even tried to use it.
Fox Tales: You’ve moved from media to restaurants. That’s a fascinating combination. Does the one inform the other in a business sense?
Hirsch: Well, the full arc is, I’ve moved from publishing to restaurants to AI-powered marketing tools and now to publishing-management software, where I’m charged, among other things, with bringing AI to our process. And, I think the answer is yes, every one of them feeds the other. What I learned about AI at Contentware and how it can supercharge your work efforts is something I’ll be trying to bring to knk’s software and processes. The data and technology requirements for running a restaurant and a publisher are really not that different, from the 30,000 foot view. Order entry is order entry.