Earlier this week I was using ChatGPT to help me design an agentic workflow. At one step, I wasn’t sure what solution to pick, so I asked, “What solution makes the most sense here, considering scalability and simplicity?” It confidently suggested a specific tool. I made sure to understand the logic behind the recommendation and checked the proposed solution, but that was good enough for me at the moment. I then realized I hadn’t asked for a full list of options with pros and cons. Why? Because, despite knowing ChatGPT’s many limitations, I trusted it. That got me thinking. Who else recommends things we trust just because they said so? Influencers. AI works similarly but without selfies, hashtags, or promo codes. It offers one or two options, often neatly packaged and confidently phrased. We accept, often without realizing we’ve just made a decision. To paraphrase Shrek, trust is like onions. It has layers - it’s not always simple or obvious. AI’s advice can be solid, but can also reflect biases, outdated information, or even designed influence. Imagine if ChatGPT favored Shopify products just because of a partnership (for the record, I’m not familiar enough with the OpenAI/Shopify partnership, and I don’t know how it will manifest in ChatGPT). The more we use AI tools, combined with their growing intelligence and our increasing reliance on them, the more they shape our decisions and narrow the options we consider, especially when our prompts aren’t clear enough. Influencer marketing is a legitimate, proven channel that delivers real results. We need to start thinking about AI in the same way; as a powerful new influencer shaping choices at scale. That’s why I’m really curious to see how AI-aware marketing will evolve. AIO and AEO are just the beginning. What’s next? #AIProductManagement #LLMstrategy #ProductMarketing #InfluencerMarketing #AIO #AEO #ChatGPT
It’s possible that in the near future, we won’t even be making these decisions ourselves. AI agents will choose the tools and solutions they use to complete tasks/goals. Sometimes with our approval, and sometimes without us in the loop at all.