Skip to main content

Return to Answer

Post Timeline

added 4073 characters in body
Source Link
l4mpi
l4mpi
  • 5.1k
  • 5
  • 47
  • 60

Updates to address EmmaBees comments:

the CEO shared a vision where “the high-quality Q&A lane” was a top priority, which suggests quality is important to the company.

The same CEO also had "visions" of AI enshittification and various other nonsensical things, and in general produced a lot of hot air and nothing of value from my PoV. Regardless of that, the actions of the company have mostly either shown a disregard for quality, or simply ignored the topic or handwaved it away, all in the name of improving engagement. So I would agree there's a disconnect "between what the company says are priorities vs how those priorities are being carried out in practice". And to put it more bluntly, I don't care what your CEO says and nowadays I don't even read the CEO blog posts because time and time again they turned out not to be worth my time (either it was just hot air from the outset or it turned out to be empty promises).

It seems to me that posting on Stack Overflow can feel daunting for both new users and experienced ones, and some people go out of their way to avoid posting at all. It makes me wonder how a knowledge base can best handle that kind of challenge

I do see that mostly as a feature, not a bug. As you say, SO is a knowledge base. It already contains a LOT of knowledge, and people asking questions which have no place in a high quality knowledge base is the biggest problem in terms of quality.

Some extremely basic questions can have a place here but many of the questions being asked in practice will never be relevant to anyone but the OP, and even OP might not care about the answer anymore tomorrow (think x/y problems, localized debugging questions, other brainfarts). Other questions have been asked literally thousands of times already in some variation and asking a new variant just shows that the person asking has neither a clue nor a willingness to google and no respect for other people's time either, e.g. Java NPE questions (almost all of them can be answered with "some object in your code is null and you try to do something with it, make sure the object is not-null before using it").

All that is to say, having friction for asking questions can be a good thing. IMO for basic questions, we need MORE friction, not less, because most new basic questions that hit the front page are not fit for inclusion in a knowledge repository.

I absolutely agree that “catering” by lowering standards isn’t the goal, and quality must remain priority.

That's great! Sadly you seem to have missed the last decade or so of SO history because throwing quality under the bus already happened, repeatedly, and was doubled down on. Looking at your account it seems you've only been here for a bit over two years so it's expected that you don't know the history, but that means you're missing 10+ years of context which I won't (and can't reasonably) summarize here. The company burned lots of goodwill and various other things such as the two most-respected CMs (Shog and John Ericson) in this time, and changed from a development and governance process that put community input front and center to a process that mostly treats the meta community as an afterthought and sometimes even as a hostile entity.

I guess overall this all boils down to different goals from the company and the community. Prosus bought SE for 1.8 billion dollars and of course they primarily care about getting a return on their investment, so they want more engagement, more ad impressions, change the ad policy to allow more obnoxious ads, push half-baked ideas like AI assistants or sponsored collectives, restrict the data dump to try to secure an "AI training data" revenue stream, and so on. Most of the community wants none of those things and at best tolerates them to some degree because they recognize the company needs to make money somewhere. So from that pov it's entirely expected that content quality is nowhere near the top of the priority list of upper management, even though that is probably short-sighted.

Updates to address EmmaBees comments:

the CEO shared a vision where “the high-quality Q&A lane” was a top priority, which suggests quality is important to the company.

The same CEO also had "visions" of AI enshittification and various other nonsensical things, and in general produced a lot of hot air and nothing of value from my PoV. Regardless of that, the actions of the company have mostly either shown a disregard for quality, or simply ignored the topic or handwaved it away, all in the name of improving engagement. So I would agree there's a disconnect "between what the company says are priorities vs how those priorities are being carried out in practice". And to put it more bluntly, I don't care what your CEO says and nowadays I don't even read the CEO blog posts because time and time again they turned out not to be worth my time (either it was just hot air from the outset or it turned out to be empty promises).

It seems to me that posting on Stack Overflow can feel daunting for both new users and experienced ones, and some people go out of their way to avoid posting at all. It makes me wonder how a knowledge base can best handle that kind of challenge

I do see that mostly as a feature, not a bug. As you say, SO is a knowledge base. It already contains a LOT of knowledge, and people asking questions which have no place in a high quality knowledge base is the biggest problem in terms of quality.

Some extremely basic questions can have a place here but many of the questions being asked in practice will never be relevant to anyone but the OP, and even OP might not care about the answer anymore tomorrow (think x/y problems, localized debugging questions, other brainfarts). Other questions have been asked literally thousands of times already in some variation and asking a new variant just shows that the person asking has neither a clue nor a willingness to google and no respect for other people's time either, e.g. Java NPE questions (almost all of them can be answered with "some object in your code is null and you try to do something with it, make sure the object is not-null before using it").

All that is to say, having friction for asking questions can be a good thing. IMO for basic questions, we need MORE friction, not less, because most new basic questions that hit the front page are not fit for inclusion in a knowledge repository.

I absolutely agree that “catering” by lowering standards isn’t the goal, and quality must remain priority.

That's great! Sadly you seem to have missed the last decade or so of SO history because throwing quality under the bus already happened, repeatedly, and was doubled down on. Looking at your account it seems you've only been here for a bit over two years so it's expected that you don't know the history, but that means you're missing 10+ years of context which I won't (and can't reasonably) summarize here. The company burned lots of goodwill and various other things such as the two most-respected CMs (Shog and John Ericson) in this time, and changed from a development and governance process that put community input front and center to a process that mostly treats the meta community as an afterthought and sometimes even as a hostile entity.

I guess overall this all boils down to different goals from the company and the community. Prosus bought SE for 1.8 billion dollars and of course they primarily care about getting a return on their investment, so they want more engagement, more ad impressions, change the ad policy to allow more obnoxious ads, push half-baked ideas like AI assistants or sponsored collectives, restrict the data dump to try to secure an "AI training data" revenue stream, and so on. Most of the community wants none of those things and at best tolerates them to some degree because they recognize the company needs to make money somewhere. So from that pov it's entirely expected that content quality is nowhere near the top of the priority list of upper management, even though that is probably short-sighted.

Source Link
l4mpi
l4mpi
  • 5.1k
  • 5
  • 47
  • 60

So, did I capture the key points? Did I miss anything major? Are there things you see very differently?

Yes, you seem to have a blind spot (maybe directed by management?) and somehow entirely omitted the fact that SO has a severe problem with low question quality. Let me just quote two parts from the top voted answer of the original post:

There has been a shift to allow unregistered users at day 1 of their programming studies to post their "hello world" questions, in order to maximize site traffic. If you tell them "the answer is actually mentioned on page 1 of your programming book" you are being rude and should be slapped in the face with a welcome wagon. Whereas it is apparently not rude to repeatedly ask unpaid volunteers inane questions that anyone, programmer or not, can answer with a minimum of research.

Overall I feel that the community has always been pushing for quality and the company has always been pushing for quantity.

And a similar sentiment is expressed in various other answers and comments.

You have an explicit point under "where folks agreed" that says

Many of you noted that good, interesting questions are sometimes getting closed

along with similar points that focus on friction while asking questions. But I don't see a single point that says anything about the mountains of trash flowing onto the front page every single day. For lots of folks the major issue is NOT that "sometimes" a good question is closed and perhaps shouldn't have been, or that it's too hard to know where a question belongs, it's that they see a hundred trash questions before they come across a decent one that warrants more interaction than a downvote and potentially closevote.

You're listing the following as a "hard but important" question

How can we make asking questions feel more approachable and predictable for users, while still maintaining high standards of quality and reliability in the content?

But from my pov that is the wrong way around, the question should instead be "How can we maintain high standards of quality and reliability, and is it possible to do so without making the site less approachable?".

With your question, the focus is "approachability", with quality being a secondary criterium. But SE has been pushing for this approachability for more than a decade now, has not presented any decent answers to your variant of the question, and has thrown curators and quality under the bus repeatedly during this time. I do not expect anything useful from your exploration of the same old "welcoming" path either because I don't see anything that's substantially different this time around, and thus I would be very surprised if you come up with something useful that was overlooked by OGs like Shog and John or even by failed initiatives like "Team DAG".

Thus I would suggest a shift of perspective, and have quality as the main focus, with approachability as a secondary criterium. Otherwise this site will turn further into "Yahoo answers, script kiddy edition", bleed more and more experts, and then lose most new coders to autocomplete-on-crack (aka LLMs). But as Lundin says at the beginning of the answer I quoted above, this is beating a dead horse as we've been having this debate since a decade, however SE seems insistent to ride this dead horse all the way to irrelevancy.