Chatbots Built With Trusted Knowledge Create Deeper Human Connections

Chatbots Build Deeper Human Connections When They're Built With Trusted Knowledge

I wasn’t going to share this because anything that remotely sounds like AI is exhausting and worn out. Not only that but I had to wonder if anyone would really want to read about this. Well — here we are. I’ll let you decide if there’s something unexpected to think about. I confess that by the time I got to the bottom line, I changed my mind about chatbots.


There’s never been a website chatbot I’ve been willing to waste my time on. You know those perky non-humans that are here to answer all your questions —except the ones that you need answered. So when Mark Schaefer introduced his MarkBot creation, my first reaction was a hard pass.


You’ve heard me mention Mark, a well-respected, trusted marketing and business strategist, in my past newsletters. His announcement sat in my inbox for a day, something I typically don’t do. Once an email is read it’s either moved to a folder for the next to-do or deleted. MarkBot was in limbo. Eventually, curiosity won.


Mark had fed his vast knowledge store of books, podcasts, blog posts, speeches, and classroom lectures to ChatGPT. Surrendering his entire intellectual property to an LLM that could do whatever it wanted was a bold decision. There were several months of training involved which would be interesting to see in action. Mark didn’t share how teaching the LLM evolved so I could only look at the results and speculate.

When Is It Time for a Human Conversation?

I promised myself that my time with MarkBot was going to be short and put an end to my curiosity. No rabbit holes allowed.


The conversation started with my asking MarkBot to imagine that he’s a small technology services company that specializes in working with C-suite leaders on strategy, technology decision making, and what comes next. Creating a unique, engaging presence that encourages leaders to work with you is your priority. Where would you start? I intentionally was short on context and details. Would MarkBot have the skills to ask insightful questions?


Mark’s response outlined five strategic actions that we could take to build engagement and start a meaningful conversation with leaders. At the end of the response, Mark offered to help shape our messaging so, of course, I couldn’t resist. This is where AI chat starts to build a connection with its user.


MarkBot asked six spot-on questions about our business that would help him deliver a messaging framework that serves our company and our clients. A funny thing happened. I knew this wasn’t a live human conversation with real Mark. He was simply asking trained strategic questions that would lead to the next step in our plan.


But I found myself deeply intent on answering each question because I didn’t want to disappoint Mark. Yes, that sounds as weird as it felt. By the way, none of the chats with MarkBot are available to real Mark. (That’s his claim anyway so I’ll go along with it.)


It was at this moment where I was challenged to think with purpose and clarity that I realized the usefulness of a well-trained chatbot — but not for the reason you might think. The practical outline for a strategic plan was actionable, and I probably could have stopped there. But I wasn't ready to call it done. Instead MarkBot created my desire to talk with real Mark. I wanted a human conversation.

It's About Human Connections, Not Problem Solving

In his book Supercommunicators, How to Unlock the Secret Language of Connection, Charles Duhigg, explains why we’re drawn to people most likely to solve our problems.


It’s not because they’re necessarily the most interesting or smartest person you know. More likely, it’s because you anticipate that you will feel smarter after talking to them. They might not even give you the best advice, but you feel better after the conversation. How does that happen?


These trusted go-to folks understand that the goal of a conversation is to connect. It’s not to immediately problem-solve, sell, or leap into expert mode.


That’s exactly what happened with MarkBot. Sure, it gave me some valuable information. Was it the smartest resource I could have turned to? Probably not. If we chatted long enough, it would likely become less useful because it's knowledge is limited. What it did was build that need for connection with a human who could take the conversation farther.

The Bottom Line

I couldn’t help but wonder if there will be more well-trained, useful tools like MarkBot in the near future.

→ What role will they play in building the lasting relationships our clients want with us?

>→ At what point will they lead us and our clients to feel the need for human connections?

→ Chatbots that are trained not on our personalized preferences but on our values will create a natural connection between the artificial and the human.

→ Instead of replacing us, will these smart tools become our collaborators, doing the initial discovery work that we do now?