Future of Communities
Last updated
Last updated
Thereās a lot of buzz around AI and people ask what the future of online communities looks like in light of these new innovations. Itās early and hard to know what awaits us.
Here are six predictions based on what (we think) we know today.
Right now, people are rushing to figure out how to cram AI into communities. Most use cases involve tacking on chat bots, automating the answering of questions, or generating massive amounts of mid-quality content.
Many communities will abuse this technology to astroturf content, discussions, and answers which will lead to unsatisfying results for members. Metrics like Solution Rate and Time To Solution will become less relevant as the quality of the solutions decline.
Owners/admins of communities will harm themselves and their members by haphazardly adding AI into the mix without fully considering the consequences in the short term. Long term, we hope to see more thoughtfulness about how itās deployed and the value to the user.
Ironically, the explosion of AI capabilities will likely drive even more emphasis, differentiation, and value of talking to real people with insights gained through experience and intuition.
Thatās not to say that the contributions of AI to communities wonāt be valuable, itās that weāre going to vastly over-do it in the beginning and āAI fatigueā will set in pretty quickly with most folks. If you thought bots were annoying before, you havenāt seen anything yet.
We can argue about the implementation details, but we believe that Elon and others are philosophically correct in saying that the future of social media and communities is the ability to filter by verified humans. Bots will become ubiquitous and humans will become novel/valuable.
We wouldnāt be surprised at all if we find ourselves all signing up with a new utility service that specializes in verification that is then portable and integrated across the web in pursuit of discovering and engaging with real people. Talking to bots only will get old, real fast.
To that end, we are likely to see further pushes into small, private group chats as one prevailing form of community simply because you can control for known humans only and the content will not be indexible/trainable by LLMs.
The last point will become a major factor in where people choose to participate going forward. The idea that everything you say can and will become part of a training set will not sit well with people. Privacy will become a defining characteristic and value prop for platforms.
This will have a downstream net effect of there being less publicly-available, indexable, quality information in open communities and the internet as a whole. It will change the dynamic of how people search for and find information.
This will be both a challenge for community builders to overcome and an opportunity. As has almost always been the case, communities with open access to high-quality information win in the long run. Itāll be a delicate balance to strike, but a winning formula will emerge.
Members are concerned that their contributions will become less valuable, while businesses are intrigued by the potential efficiencies/cost savings. Itās easy to think that AI will write all the content, but thatās not how it works.
LLMs and bots are only as good as the dataset theyāre trained on. They will always be reliant on humansā perspective and insights to generate content. For example, when Apple ships a VR headset, AI arenāt going to instantly know about it and write all the content.
Only after humans have written volumes about its features, functions, advantages, disadvantages, and so on will LLMs be trained on the data and able to answer queries in a community support environment.
tl;dr ā quality human contribution will always be in fashion.
If you agree that genuine, verified human connection and content creation are still valuable in an AI-driven world, then it stands to reason that companies with established communities are best positioned to offer the experience.
Itās not zero sum. You can provide both an authentic experience and deploy clever AI in your community that holds immense utility for your community members. No one has nailed this balance as of yet, but itās only a matter of time before a new standard is developed and copied.
āAI Discovery Optimizationā will be the new SEO. In pursuit of your companyās products and services being surfaced in generative results, the most content wins (as well as paid advertising, of course). Companies with large groups of customers creating content to be trained on win.
We have to move past the phase of AI being a technology into it being a service that powers products and features inside of applications. A new breed of community platforms is coming that with that aim.
Communities probably wonāt look too different going forward, instead itāll be more about AI being tucked in neatly into core functionality that makes usersā lives easier: help them discover solutions faster, ask better questions, get answers faster, and build human relationships.
That's why we're building OpenPurpose. Weāll see what comes of all this in time.