Lou Cohen, chief digital officer at EY, said many marketers are yet to take advantage of the benefits of artificial intelligence.
Cohen, who is also a professor at New York University, Yeshiva University and Baruch College, said marketing is at a “tipping point” as investment shifts from general digital innovation to AI transformation.
Cohen said marketers who focus on what outcomes AI can deliver and understand how to use AI in supportive ways will have access to deeper levels of audience segmentation, targeting and testing. He was interviewed by CMO Insider at Business Insider’s studio in New York City.
Ultimately, Cohen said, the marketing department will embrace new opportunities. “Marketers aren’t afraid to try things,” Cohen says. “We’re going to learn more from what went wrong and what didn’t work than what went right.”
The transcript below has been edited for clarity.
We are at an interesting tipping point. In today’s marketing environment, you really need to understand how to leverage AI. If you don’t, you’ll end up working for it.
Gain efficiency and operational benefits. But if you think about the results AI can deliver from a marketing perspective, you can become smarter about how you segment your audience for different campaigns. You may be able to make your advertising more efficient. You’ll be able to test faster and deliver higher quality content to the right audience in the right place.
However, most marketing teams are not yet ready to take advantage of this potential. In other words, the digital transformation investments of the past 15 years are now transitioning into AI transformation.
It’s a bit unknown now. Marketers are not entirely happy with this, as they are very concerned that it will cause hallucinations or provide inaccurate information. Marketers aren’t afraid to experiment. We learn more from what fails and what doesn’t go well than from what goes well.
My colleagues have come up with a great way to evaluate the quality of content using AI. Paste an article written by one of our partners and we’ll give you recommendations on how to make that content better. But we will never use content created by AI. You shouldn’t say never. But it is certainly possible to use AI to enhance content and provide feedback to content creators.
Hallucinations are real. The challenge is that we, the users of these technologies, do not yet understand the difference between probabilistic and deterministic results. Probabilistic is the probably correct response that the AI is trying to give us. Deterministic means “1 plus 1 equals 2” and, without a doubt, 1 plus 1 always equals 2.
For example, when you perform a search on Google or Bing, you get a definitive response. You will receive information that we believe is likely to answer your question. LLMs, ChatGPTs, Llamas, and Geminis of the world give probabilistic responses. The model brings together various sources of information to determine the answer it thinks you should get based on the prompt.
This means that even if you’re using these tools for the purpose they were designed for, you still need a search engine just to navigate to what you’re looking for, or to find a needle in the internet’s haystack. But the LLM gives us another opportunity. They can be assistants. This was part of the original idea behind these AI tools: to help people perform different tasks.
I think of these LLMs more as marketing assistants who give me real-time ideas, feedback, and suggestions, rather than performing tasks for me. This is where humans leverage AI to get faster and better results than they could on their own.
