Why Humans Still Need Humans
The Story Behind Stack Overflow's Decline
A few weeks ago, I came across this interesting chart from the Pragmatic Engineer’s Newsletter on the ‘Monthly questions asked on StackOverflow’.
It was a telling graph because while I had strong suspicions that overall engagement had been on the decline for years, I certainly didn’t realize it was this bad until I saw the graph above, which was originally derived from Matt Gravel’s X post. Matt is also a top-ten all-time contributor to the platform.
Now, it’s easy to assume from a cursory glance from the visual that the advent of Generative AI was solely to blame for the rapid fall, but if we peel the onion layers back, there appears to be something else that’s been driving this, and perhaps it was simply AI that finally put the nail in the coffin.
As a quick primer: Stack Overflow (SO) is a Q&A platform for developers/programmers and software engineers, launched in 2008. It’s a community-driven hub where developers ask, answer, and vote on technical questions from programming languages to software design. Users post queries, and others provide solutions, with the best answers upvoted for visibility. Its gamified system rewards contributions with reputation points, badges, and privileges, fostering engagement.
You see, back in the day, SO was one of the hottest online software communities because it was based heavily on the engagement of its members, and it was part of a wave of programming forums. More importantly, it was hinged on the fact that anyone and everyone could join the community as long as they followed the rules and guidelines set by the moderators. These were enough to understand and abide by: don’t use profanity, don’t speak off topic, refrain from any abuse, etc.
From this, SO was able to develop a pretty harmonious and valuable community. You would come across folks from around the world who would provide genuine guidance and advice, completely pro bono. This self-regulated group of people ultimately gave way to its popularity and rise, hitting ~200,000 questions around 2014. I was an avid user myself.
However, over the last several years, beginning from 2018 (notwithstanding COVID in 2020-21), you could see a progressive decline in questions being posted on the once popular forum. But why?
One of SO’s most powerful features was its reputation system. It was the thing that gravitated developers and the like to its platform. It allowed users to earn rep points and badges for asking thoughtful questions and giving helpful answers. This mechanism encouraged people who joined to be on their best behaviour. It wasn’t the Wild Wild West like forums that had gone before it. Because of this, intrinsic value was created, and users like me felt comfortable participating in discussions without the fear of backlash or abuse. Furthermore, anyone could join, ask, and/or answer any question they wanted, and only those that were helpful were naturally upvoted, which implied they would be placed at the top, thus getting more views.
If you’re thinking this is like Reddit, then you’re correct. This system has proven itself to work.
But, over time, the rules changed. SO became much more self-governing, and those with enough reputation points became the gatekeepers. They took charge of moderating incoming questions and answers. Ultimately, this meant that posting had to be parsed and filtered by the moderators who would be the decision makers to ensure it was “worthy” enough to be allowed through. This overregulation devolved the SO community culture such that it was no longer a safe space and eventually transitioned into an elitist culture.
While unfortunate, SO is a cautionary tale that gamification is a double-edged sword. What made the platform great had easily led to its timely downfall as well. The culture of real human discussions, even if it goes in weird directions, is exactly what a good conversation looks like. It’s not about being transactional.
Given all of this, I think there’s something to be said about humans connecting with other humans.
It’s easy nowadays to ask questions to an LLM (like GPT-x or Claude, etc.) and get an immediate answer. But it’s also important to understand that SO wasn’t just a Q&A platform; it was a thriving community of people who engaged over thoughtful questions and answers, and of course, people who genuinely wanted to help.
While LLMs have certainly provided a much easier approach to getting answers, I feel human connectedness is still incredibly valuable.
It’s why Reddit continues to thrive for this very reason. The irony is that AI relies heavily on authentic discussion like this — Anthropic is being sued by Reddit for scraping its data without consent.
It’s clear that forums like Reddit will be extremely valuable in this new age of AI because humans (including me) want to banter with real, genuine, and authentic people, even though it might take us a little bit longer to get to the answer.
🙋🏻♂️ Survey
To build the best newsletter for you all, I’d love to get some feedback. If you have some time, please fill out the following anonymous survey.
🤩 Deals, Discounts & Affiliates
You can find carefully selected offers and deals here.
🙏🏼 Connect with Me
Are you new to the newsletter? Subscribe Here
Check out my YouTube channel (and subscribe!)
If you’re a founder, apply here (Metagrove Ventures) for startup funding or contact me directly at barry@metagrove.vc
If you think this could be helpful and informative to others, please share it :)
Thanks for reading.
Barry.