While artificial intelligence has been powering algorithms that we've been using every day for years, the hype around generative AI tools like OpenAI's now-infamous chatbot ChatGPT and image creator DALL-E 2 is at an all-time high.
Meta is now using Shutterstock's massive content library to train its AI capabilities, Microsoft is shelling out billions to its ongoing OpenAI partnership, with plans to integrate society-changing systems to its products, including Bing search and apps like Word, PowerPoint and Outlook -- and some companies are seeing 800% improvement on ROAS (return on ad spend) by using AI technology.
With 2023 promising to be “the year of AI,” MediaPost sat down with Geoff Renaud, CMO and co-founder of Web2 and Web3 ad agency Invisible North to discuss where the hype might lead for brands and agencies, as well as what dangers may be lurking because of these emerging technologies.
MediaPost: Who are you currently working with on the AI front?
Geoff Renaud: We work with a company called Atlas that creates generative gaming and metaverse environments via language commands. My company does brand experiences and now we're building more gamified learning environments and different virtual experiences for brands.
If you tell it to build like a pink diamond-crusted castle inspired by Frank Gehry, Atlas will just pop it out and you can drag it into a virtual world.
MP: What do you think of Microsoft's newest partnership phase with OpenAI?
Renaud: Microsoft's integration of ChatGPT will be leveraged not just for new technology, but for their entire stack. If you're using Teams or PowerPoint, all of these AI tools will be built right into our existing stack and supercharge them.
MP: How are brands and marketers utilizing AI right now?
Renaud: Definitely content and creative. Creatives seem to be obsessed.
I think people are starting to figure out music, world building, and finance. But content, SEO -- creating automatic backlinks, avoiding having to build a whole catalog, etc. -- is currently most common.
What I like to remind people is that AI tools are composable. Let's think about it in terms of customer service. Changing a flight, for example: ChatGPT will allow Delta's basic chatbot to make more nuanced and complex decisions, which can then be blended with a natural voice reader or text-to-speech tool which synthesizes an attendant's voice.
MP: What are the negative implications for this type of technology?
Renaud: Last summer, I got a message from the VP of marketing at a major company on LinkedIn. I got a call shortly after from the same person who I knew from ad trades. We talked for half an hour and at the end of the call they wanted to work with my company, sign a contract, etc.
But then they told me they ask all their new vendors to make a donation to their foundation. My lawyers ended up telling me that it was a deepfake of the real person.
Hackers can now literally synthesize your mom's voice and call you asking for money.
MP: How will this technology be regulated for brand safety?
Renaud: There's going to be a bunch of acquisitions over these AI tools, some of which I think will be pretty well-regulated. But I also think the scams will move just as fast as the technology.
It's exciting that you can hyper-charge your content or music or customer service. But we need to remember that data is the prime currency here.
But we need to remember that data is the prime currency here; as soon as these AI tools receive mass adoption, they will become smarter and smarter and smarter.
We already have such an issue with scams and phishing. Right now, it's usually easy to tell that you're getting a call from a robot, but it will become more difficult to tell and scams will be out of control, especially for the portions of the population who aren't tech-savvy.
I think big players will create guardrails around the AI tools, but a lot of it is open-sourced and composable which will invite unruly global hackers -- it's inevitable.
MediaPost: How would you compare the hype around generative AI tools to Web3?
Renaud: Web3 has major issues with consumer-infrastructure and regulation. I believe in the premise of digital asset ownership and other Web3 applications, but the reality of it has become very difficult to adopt.
Scalability, sustainable business models and consumer infrastructure are major challenges that Web3 still has to overcome.
But AI doesn't have those challenges. Most generative AI is instantly adoptable. It's easy for consumers and developers to interact with. It's going to proliferate and compound way faster than anyone realizes.
MP: How will generative AI be used by ad agencies?Renaud: Overall, generative AI gives you tools to hypercharge productivity.
For agencies, AI is going to proliferate every single discipline or genre. If you're making music for campaigns, you can use platforms that generate royalty-free music. You'll have AI for predicting models for finance and budgeting, content, chat, etc.
MP: Do you think AI tools will begin to fill human roles in the ad industry?
Geoff Renaud: My wife's a copywriter. If it was hard to make a living as a freelance copywriter before, try walking in now and demanding your usual hourly rates for a month of work. It's you versus ChatGPT.