Early in my career, a few months after we launched TechRepublic, I watched something happen in the community that helped me understand the possibilities that communities represent. Someone posted a very specific technical question about a problem most people would never encounter. Two members from outside the US, who were probably the only two people in the world equipped to answer it, did.
I was blown away. That was the power of our community, creating the conditions for collisions that help everyone advance together.
I thought about that moment again today on a panel at the All Things in Moderation conference. Venessa Paech asked which inevitability narrative about the future of community we should stop accepting. Mine was easy: the story that AI will automate everything.
We learned why that’s a fallacy the hard way when I was leading a community for a design software company. We were early in experimenting with agents. The models are good at one thing, ferreting out an answer that already exists somewhere. When a member hits a genuinely new problem out in the wild, something the model has never seen, it has nothing to offer. The collision I watched at TechRepublic was exactly that kind of problem. So the idea that AI will replace this work, or even keep a community’s knowledge base alive on its own, runs into a wall that has nothing to do with optimism or pessimism about the technology. It simply isn’t technically possible.
What a community actually produces is relationships, and those relationships outlast the people who tend them. They outlast the staff and the executives who fund the work, and they keep compounding value long after both have moved on. A model can summarize what a community has already said. It can’t carry a network of relationships for decades.
Which makes the current arrangement worth looking at, honestly. The most valuable members, the ones answering the impossible questions, are feeding what they know into systems that increasingly reuse it, and most of them never see the value they create or get much back for it. That exchange was already lopsided. As contributions get absorbed into models and resold, it tilts further. Any serious version of the future has to make the exchange visible and return something for it, rather than taking it opaquely.
This is what bothers me about how many tech leaders tell their story of a future that is fixed. It tells leaders the outcome is already decided, that the only question left is how fast to adopt. The future isn’t set. Howard Rheingold, writing about The WELL, one of the first online communities, captured something I keep coming back to, that “What it is, is up to us”, “it” being the practice of community building in this context. There are many futures ahead, and we get to (we must) steer toward the one we want.
The best version of AI in this work absorbs the drudgery so people can do the part that only people can do. The worst version is the one where we become the tool the tool uses, which is roughly where mass social media left us. What separates those two outcomes has little to do with the technology and everything to do with whether the people making strategy and design decisions about community today keep a hand on the tiller.
If I had to put one principle at the top of a charter for the next decade of this work, it would borrow from the Hippocratic oath: human well-being comes first, including the well-being of the people doing the work. Build from there, and build around possibilities rather than constraints.
To paraphrase Howard Rheingold, the future of community practice, participation, and platforms is up to us.