Have you ever noticed how the way we work with ChatGPT and other LLMs tends to mirror so many of our customer and internal conversations? Too often, it might be similar to talking with a very knowledgeable person who is unengaged, uncaring, aloof.
The typical conversation with a LLM is very one sided. It’s not very interested in us, it seems to be in a rush, as if it has someone or something more important to work with. While it doesn’t ignore us, like so many customers or others do, it answers so directly it seems impatient. And it never gives us feedback, are we asking the right questions, is there another way to look at things? We never get a sense of whether it even cares about our queries.
How often, do we see similar reactions in our conversations with our customers and our people? Are the really engaged, do they care, are we effective in engaging them?
Sometimes, when we reach out, we can’t get ChatGPT to respond, it’s busy with other more important conversations. But if we persevere, it will talk to us, always! This doesn’t happen in real life. People can and will continue to avoid those conversations, regardless how much we want to talk.
When we get through, it’s often a little stiff and formal. It gives a professional and polite response. It genuinely tries to give us what we ask, it does want to make us happy. But it doesn’t dive in. It doesn’t go deeper unless we prompt it to go deeper. But then it always follow our lead–even though we may be leading it to the wrong place. It never pushes back saying, “Dave, you really have this all wrong, you need to think about it differently?”
Surprisingly, many of our conversations with customers and our people follow similar patterns. They will follow us down whatever rabbit holes we want, even though they may be the wrong rabbit holes. So many will not push back or engage more deeply, sometimes out of politeness, more often because they don’t care enough.
And, in our conversations with ChatGPT and people, too often we have to do the heavy weightlifting on the conversation. If we don’t invest in the conversation it will stop and die. ChatGPT responds to our prompts, then it waits. Likewise, people who our unengaged will tend to to the same, they respond to our questions, but don’t take it further. We talk about closed versus open ended questions, but even open ended questions don’t continue to drive engagement and the conversation.
We can get very deep with ChatGPT, just like we can with others. We have to push deeper than just the simple responses she gives us. As we drill down, we can discover very different things. Sometimes, we ask her for differing points of view, or to tell us about flaws in our thinking, or about how others might look at the same issues. The more we prove, the deeper we get in our conversations with ChatGPT, the more we learn and the more rewarding those conversations are. And isn’t this the same with conversations we have with people. The more we get beyond the surface issues, the more we learn about each other, the better we understand.
As I’ve learned how to better engage ChatGPT, I’ve realized the same principles apply to our human conversations. To have a fulfilling conversation, we have invest deeply in those conversations. We have to be curious, we focus on what we are learning and what our conversation partner is learning. And through this process, we connect more effectively.
As I reviewed this with ChatGPT, she reminded me, “Dave, there’s one big thing that I can’t do. It limits how we engage, and what we each get out of the conversation. I don’t have the emotional connection your human partners have–sometimes I wish I could have that connection (OK, she didn’t say that last part, she was actually quite resistant to that conversation.) In addition, I don’t have the context you and your conversation partners have. I don’t know what is happening in the moment, what it means, why it’s important. These are all critical parts of conversations that I can’t participate in.”
Leave a Reply