Upcoming initiatives on Stack Overflow and across the Stack Exchange network (March 2024)

Thanks for following up on the previous announcement. This definitely helps answer the questions about what's coming next for "us", the community of Stack Exchange users, especially those who curate and moderate content and help maintain the overall quality of the network as a resource.

I think the blogpost starts out a bit on the wrong foot with the overemphasis on generative AI (especially in the very first sentence) that seems to cue "here's more GenAI stuff you didn't ask for!", but if a reader sticks with it, I think the other initiatives discussed there (those bulleted here as well) sound like a promising direction. I'll be eagerly awaiting further details: a bullet point like "better onboarding" doesn't say very much, and I'm curious to know which of the plethora of ideas offered by the community will get attention. A renewed Staging Ground, on the other hand, is a clear win for the people who have already put a lot of time into making it work (including former employees).

On the AI side of the blog, both the "guiding principles" and OverflowAI alpha takeaways sound to me like a more measured approach to AI tools that would have been appropriate for Stack Overflow to take from the outset. My confidence in SO's approach would have been much higher if you started there, but we can't undo the past and I'm glad you're getting there. Don't forget that your primary value here is an excellent, human-made, human-curated resource! You are wonderfully positioned to be an answer to many of the limitations of AI (and those limitations have been well-known for a long time). Don't lose sight of that chasing fads!

I await the exact meaning of "better onboarding" with bated breath.

I hope it means being intentional and zealous about earnestly promoting contextually-relevant Help Center pages to their intended audience in the site UI/UX (sorry for the word dump in that sentence, but I chose each of those words with care).

I hope it means adding basic logic to have the system automatically catch potential basic mistakes like posting non-answers, and then popping up friendly tips concerning those mistakes and guiding to a proper path of site usage before those mistakes are committed. (an example of what I refer to as "JIT / Just-In-Time guidance")

Such changes would offload a lot of annoying work from human volunteers, and remove a lot of frustration from both sides of curators and newer users.

Zoe's comments don't quite inspire confidence though.

And reading the relevant part of the blogpost doesn't inspire confidence either.

We believe if it were easier to participate, more users would.

Better onboarding to me does not mean more users participate (unqualified statement). It means more users participate and are automatically guided through texts which teach them the site rules, guidelines, and norms. I.e. The Help Center and Meta.

We'll start by getting more users onto the platform; a project has already kicked off with experiments to optimize the sign-up flows.

We're focusing on simplifying and making the platform more approachable, enabling everyone to have those "aha!" moments more frequently.

Assuming "approachable" means having fewer negative experiences on the receiving end of curation / guidance comments (what I'll call "desired output"), and that those "aha!" moments are the moments where people come to understand what Stack Exchange is, what its mission is, and what its methodology is, along with its rules, guidelines, and norms (what I'll call "efficiency", since it is what enables each individual to avoid such negative experiences), then you're choosing an approach to increase desired output where you attempt to increase input (add more users) and leave efficiency (desired output divided by input) untouched.

desired_output = input * efficiency

I can't explain why I think this input-focused approach is poor. But I do. I think it's shallow and lacking in taste/perspective (excuse my lack of skill and interest in persuasive writing with sweet words).

Maybe it's because I see a lot of low-hanging fruit in increasing efficiency (see what I wrote at the top of this post and all my other meta writings about SE).

Maybe it's because I went through a certain stream of education which is concerned about efficiency in design.

Maybe it's because I have a morbid interest in reading comments on Reddit by the people who probably didn't have what I think is a good onboarding experience on SE, and ran into a pitfall or shot themselves with a footgun, and then decided they hated it and would go to AI and never come back. The natural question is- where are you going to get all this input for the equation? I'm convinced that it's largely these types of people who left. Good luck getting them back without changing the story with respect to educating newer users about what this platform is and how it works, and waiting for such an actual story change to become reflected in peoples' mental images of the platform and their external discourse about it (if you increase input without changing efficiency, you'll have done absolutely nothing to change the balance of that discourse- the proportion of people who understand and appreciate the goals and methodology of SE). Or maybe there's no way to get them back. Maybe if they learned what SE really is and how it works, they'd realize they never wanted to take part in the first place. To me, this naturally leads me to think that it's wiser to focus on the efficiency part of the equation.

We will continue this effort by acquiring new users and providing lower-effort ways for users to participate. By simplifying how users interact on the platform and providing guidance to learn community norms, we can help users feel successful while maintaining a high bar with content health.

I don't see how "lower effort ways to participate" has any direct consequence of "providing guidance to learn community norms". The bars to participate are already incredibly low. The actual issue is that the guidance on how to do so properly is not reaching the ears of the people who need it. Though I am interested what ideas you have about "lower effort ways to participate", and have a couple of ideas myself.

As for discussions, I await the return of downvotes (one of the primary/foundational curation mechanisms).

As we look forward to a new fiscal year, we are focused on initiatives that support community health and growth by making it easier for users to find the information they need

I wonder what exactly the information being referred to is. The statement is especially weird since the AI search experiment is "ramping down". Since I care about user onboarding, my mind immediately jumped to hope that this is about making it easier for users to find Help Center pages and Meta Q&A they should read, but I'm going to try not to get my hopes up with assumptions.

Thank you for letting us know about your priorities. It is good to know what to expect.

If you are looking for feedback, here is mine:

Staging Ground sounds promising and like a step forward. I'm encouraged to hear a focus on onboarding; that sounds like it could be a step forward, depending on what that means, and I wait to see what specifically you all have in mind. The work on image hosting sounds necessary.

The remainder of the items sound neutral to me -- neither positive nor negative. I do not expect moving to the cloud to make any notable difference to the user experience, either way, and I am skeptical it will make much of a difference to reliability, in terms of my ability to use the network when I need it (I find the network already more than reliable enough, and I don't expect to see any end-user-visible value from this move). I am not optimistic about the Product Advisory Council making an appreciable difference either way, though I appreciate that CMs expect it to be positive, so we shall see. I don't think Discussions contributes to the Stack Exchange mission or is particularly valuable in its current form and so, while it's unclear exactly what work on Discussions is planned, I don't expect value from more work on Discussions. I think the AI stuff is not a good use of time, so it's good that it will be ended, but I'm skeptical of plans to devote resources to ramping it down or new applications of AI.

Overall, it sounds to me like 3 out of the 7 priorities are ones that I would agree are a good priority. I'm glad to see work on those. It sounds encouraging that there will be work on improving the platform.

Try Discussions for all technical topics on Stack Overflow.

Share insights about all your favorite technologies.

• Seek advice, expertise, and best practices from others.
• Dive into meaningful conversations with others in the community.
• Communicate your opinions about specific areas of practice.
• Registered users can create new Discussions and reply to existing Discussions.
• While voting is supported and post scores are tracked, there is no reputation impact to voting within Discussions.

The people who run Stack Exchange have finally conceded that comments have their use. Beyond requests of clarification, detail, evidence of research, and informing users their questions are either off-topic or a duplicate; comments–I presume–will not to be locked, transferred to chat, edited or deleted by moderators or staff members. There's also a good chance that the number of flags, especially on Stack Overflow, will drop significantly as users who insist SO and other Stack Exchange sites are not forums will refrain from raising flags every time they encounter a meaningful banter between two or more users in their field of expertise.

How does this groundbreaking experiment work? In a video entitled Community Discussions, the femalr voice-over, which I swear is generated by an AI, describes the simple process:

…you see a discussion where someone is asking for opinions or thoughts on which approach they should take. You click into the full discussion and read through the comments. Then you want to add in your perspective. You share your thoughts on what they can consider, thinking about how they can trade off between future proofing their approach and future expense associated with it. You post your input and are part of the discussion.

Will this experiment gradually be extended across the network?

By the way, how is “community discussions” significantly different from an online forum or message board? How will SE differentiate from Reddit?

We were told years ago that comments were worthless, they added only noise. Stack Exchange was different, it focused on answers not chit-chat. We've all heard (and mostly accepted) this argument, haven't we? Examples:

I have been told time and again, as have others, that the Questions and to a far greater extent Answers are what matter in the long run; comments are ephemeral. Source: @KorvinStarmast

A former Community leader, much respected, posted in June 2018 the following reflections: [emphasis mine]

Try not to provide full answers in comments; if you end up working a problem out in comments, please move it to an answer. We know you're trying to help, but the system expects answers to questions. If we're reiterating that comments are ephemeral (and they are), we have to caution against leaving good information in them that needs to last, too

Instead, six years later, comments will occupy a higher role: they will be productive, meaningful, an immediate way to share opinions and experiences in any of the ramifications a single post might generate.

Returning to the announced initiatives and the related Stack Overflow article, I read the following [emphasis in bold mine]

As we consider integrating AI into our platform through partnerships and new features, we remain committed to preserving the essence of Stack Overflow: a space driven by human connection and genuine knowledge sharing.

and further along…

• The combination of community-driven verification with AI generated excitement about improved trust and reliability of technical content; however, user trust in the accuracy and relevancy of content is paramount. Transparency is key.
• AI-powered search features helped developers cut down on time in low-complexity situations, but sometimes ran into hallucinations when dealing with answers to more complex problems.
  Source: Community Products: Reflections and looking ahead March 27, 2024

SE cannot pretend to value human-generated content when visitors and users alike will soon be provided with solutions generated by AI, which in turn will (miraculously) be curated by the human community. Why would developers, programmers and engineers sacrifice their precious time and years of experience on checking the validity of AI codes? It's plainly absurd.

Am I seeing irony where none exists?

Only now does the company fully appreciate its legacy and success lies with the user base, complete with their human interactions and their behavioural flaws. It has also paid tribute to the incredible generosity of its communities on numerous occasions. Yet, at the same time SE seeks to promote AI generated content, as if it were the deus ex machina, in the present crises. Madness.

We will launch the 1-rep voting experiment on Stack Overflow for 4–6 weeks, along with account creation prompts was posted a couple of weeks ago on Stack Overflow Meta. I'm surprised that this so-called "experiment" wasn't mentioned here as one of the initiatives, as it was derived from If more users could vote, would they engage more? Testing 1 reputation voting on some sites.

The "experiment" is currently paused. Ref. Pausing the 1-rep voting experiment on Stack Overflow: reflecting on the feedback and rethinking the approach

I kindly request that someone from the CM team add it and make an update about the 1-rep voting "experiment" / "initiative".