We’ve talked about our views on AI before (https://designroom.com/the-meat-of-meaning-how-to-create-rich-content/). But because we’re always interested in learning more, this month we sent Anna and Matt to the Marketing Artificial Intelligence Conference (MAICON) in Cleveland. Here’s what they reported back to us.
(Although we wrote this blog unassisted, we did copy the contents into ChatGPT and had it generate the title.)
I had last attended this conference in 2019, and a lot has changed. At a high level, I recall just four years ago much of the discussion was around machine learning. This time there was frequent mention of Large Language Models (LLM) and application uses of AI.
To provide a high-level, loose understanding, a quick internet search turned up, among others, these definitions:
Machine Learning is a subfield of artificial intelligence, which is broadly defined as the capability of a machine to imitate intelligent human behavior.
Large Language Models are a type of AI that are currently trained on a massive collection of articles, Wikipedia entries, books, internet-based resources and other input to produce human-like responses to natural language queries.
After just one day, my head was swimming with all of the things I heard and all of the possibilities my mind started wandering with. During MAICON 2023, I only scratched the surface of what’s out there.
Here are my big takeaways:
- There are LLM companies and then there are Software as a Service (SaaS) companies building applications on top of AI. Some of these applications take the form of action agents, which help automate workflows.
- To create good AI-generated images, you need to input great prompts. Brian D. Fanzo, digital AI artist and keynote speaker, says he speaks with, not at, the AI, including using “please” and “thank you.” The more “story” you put in, the better the image that comes out.
- AI can’t replicate a human’s experience and subject matter expertise. So when thinking about AI use cases, one question to ask is, “Could somebody write this without my expertise?” If so, then maybe you let AI do it.
The best speakers didn’t just address the possibilities of AI, they also covered the things to treat with caution as well. Two talks in particular stood out to me.
Whose Job Does AI Automate? by Cassie Kozyrkov, Chief Decision Scientist at Google
Kozyrkov spoke eloquently on the future of AI-assisted work. Her talk focused on the applications of AI. She sees AI as a tool to make human thinking more efficient—AI is better than humans at some things but incapable of human creativity.
AI replacing human jobs is on top of everyone’s mind, but Kozyrkov didn’t see it as a major concern. From her perspective, AI isn’t replacing humans because it can’t. But it will necessitate a revolution in management. If mindless work is no longer filling most of the workday once it’s outsourced to AI, then new metrics of success need to be implemented. I found this perspective super interesting, and came out of it with a lot of things to think about.
Art in the Age of AI: Decoding Technology, Protecting Artists, and Championing Diversity by Stefan Britton, AI GTM & Innovation at Shutterstock
This was another standout learning opportunity. As a designer, my number one question going in was the state of AI art tools. There’s nothing worse than using a stock photo and then seeing someone else use it, too. The idea of a tool that could create an entirely unique “stock” photo is something I’d love to take advantage of. So going into the day, I was interested in what Stefan would have to say, especially because of his work with Shutterstock.
This presentation more than answered my questions. Britton led the development of Shutterstock’s AI tools, and spoke at length about how to use AI tools and protect artists. Much like Napster, AI image tools are the Wild West right now. Organizations have to take on their own responsibility to use AI ethically, and integrate those ethics into their systems moving forward. I’m excited to use the DALL·E 2 tool and generate some stock photos knowing that the artists behind the tool were compensated fairly.
One central idea ran through the entire conference:
Today is the dumbest version of AI you will ever know.
As AI changes, so will the state of branding and marketing. We’ll be keeping up with it and continue sharing our thoughts on the ways AI can be helpful, harmful, or maybe a little bit of both. Until then, let us know if there is a cool way you are using AI in your branding or marketing efforts.
At designRoom, we make it our business to find real answers and create custom healthcare brands. We believe effective healthcare branding is grounded in research, directed by insight, and driven by strategy.
We love seeing how strategic branding helps the right clients find the right organizations and receive the right care. That’s been our focus for over a decade. Today designRoom is an award-winning, national branding and design firm, known for helping clients build and promote healthy, sustainable brands. And we are super proud of that.