Most of the responses here say "because the moderation is too strict" -- but this is only a symptom of a real problem: the site has been in feature-freeze maintenance mode for about 10 years, with very little new development or improvement since then. The moderation system has aged very poorly -- the tools are inadequate for handling the volume of traffic SO receives, and the process is incredibly opaque and Kafkaesque for new users to navigate.
In its early days, the site had a vibrant community, and was actively being improved by a team that pretty much entirely consisted of power users. The whole team was active on the site and available in Meta and chat -- you could report a bug or propose a new feature and it would often be fixed/implemented and live on the site within a couple hours. The site was expanding and constantly being updated; the staff was focused on expanding the Stack Exchange platform, making the user experience better, and building up the community.
But around 2016, the company abruptly stopped most new development on the core platform. They announced a sweeping "Quality Improvement Project" [0] that was supposed to bring badly-needed improvements to the moderation system, asking the community for ideas and suggestions; announcing some major overhauls that were in progress...and then radio silence. For the next several years, there were no changes to the core platform except bugfixes, minor tweaks, and several UI overhauls that nobody asked for. The company shifted nearly all of its development resources to side projects that were universally failures: Teams, Documentation, video tutorials, another product called Teams, etc. Stack Overflow was (and still is) nowhere near profitability, so they were cutting costs and desperately trying to find some way to monetize the site, while neglecting their core platform and laying off staff.
Over the next few years, the team became more distant and unreachable as staff left and were replaced with new employees who were not engaged users of the platform, and largely didn't interact with it at all. Most of the company's hiring was in sales and marketing; the company is now largely run by the sales department rather than engineering. The site started to decline during this era, and the fun, friendly atmosphere of the early days gave way to with a more grumpy, corporate, and bureaucratic vibe.
Then in 2019...a lot of stuff happened. The company published a blog post blaming the userbase for the site's unwelcoming reputation, rather than the broken moderation process. The company made an announcement that they would no longer be engaging with the community and seeking feedback on new products and features, because they were tired of reading negative reactions to products that weren't useful and changes like UI redesigns that compromised usability. (Infamously, an executive quipped that the users who participate in Meta discussions represent "0.015%" of the site's visitors, and therefore weren't worth listening to -- despite these users contributing around 80% of the site's content). The company also announced plans to illegally change the licensing terms of user-submitted content, fired a volunteer moderator with no warning or explanation over a blatantly false allegation of transphobia [1], issued false and defamatory statements about that moderator to The Register; and fired two staff members who were just about the only employees who still engaged with the community and pushed back internally on the company's bad decisions [2]. A huge chunk of the community (including myself) stopped participating after everything that happened in 2019, and the site has seen a massive decline in quality and engagement since then.
In the wake of this, the company did start to make some changes -- they started engaging more with the community, and started working again on new features and changes, including moderation improvements -- but it's clear that people in the company working on the core site and engaging with the community have limited resources and very little influence with the company, and their work is too little, too late.
Then, the company got bought out by a private equity firm. The new CEO seems convinced that AI is the solution to monetizing Stack Overflow, somehow. This started with implementing half-baked AI features that didn't work, and the latest iteration seems to be trying to convince AI companies to pay for access to the site's content -- never mind that it's all available for free under a creative commons license. He's been trying to lock down the site's data dumps in an attempt to restrict AI companies from being able to download them, most recently with another attempt at illegally relicensing CC-BY-SA content [3]. Never mind that the drastic decline in engagement over recent years means whatever value the site has for AI training comes from its historical content which has long been freely available; not its new content [4]. (The most entertaining part of this debacle is that whenever the company announces some misguided new initiative, droves of ex-employees show up on Meta to tell them what a bad idea it is.)
So, TL;DR: the company neglected their core platform for years, lost or fired their most experienced staff, and lost most of their userbase and engagement. Now they're burning money, running an outdated platform that's not up to the task, and are caught off-guard by ChatGPT.
In its early days, the site had a vibrant community, and was actively being improved by a team that pretty much entirely consisted of power users. The whole team was active on the site and available in Meta and chat -- you could report a bug or propose a new feature and it would often be fixed/implemented and live on the site within a couple hours. The site was expanding and constantly being updated; the staff was focused on expanding the Stack Exchange platform, making the user experience better, and building up the community.
But around 2016, the company abruptly stopped most new development on the core platform. They announced a sweeping "Quality Improvement Project" [0] that was supposed to bring badly-needed improvements to the moderation system, asking the community for ideas and suggestions; announcing some major overhauls that were in progress...and then radio silence. For the next several years, there were no changes to the core platform except bugfixes, minor tweaks, and several UI overhauls that nobody asked for. The company shifted nearly all of its development resources to side projects that were universally failures: Teams, Documentation, video tutorials, another product called Teams, etc. Stack Overflow was (and still is) nowhere near profitability, so they were cutting costs and desperately trying to find some way to monetize the site, while neglecting their core platform and laying off staff.
Over the next few years, the team became more distant and unreachable as staff left and were replaced with new employees who were not engaged users of the platform, and largely didn't interact with it at all. Most of the company's hiring was in sales and marketing; the company is now largely run by the sales department rather than engineering. The site started to decline during this era, and the fun, friendly atmosphere of the early days gave way to with a more grumpy, corporate, and bureaucratic vibe.
Then in 2019...a lot of stuff happened. The company published a blog post blaming the userbase for the site's unwelcoming reputation, rather than the broken moderation process. The company made an announcement that they would no longer be engaging with the community and seeking feedback on new products and features, because they were tired of reading negative reactions to products that weren't useful and changes like UI redesigns that compromised usability. (Infamously, an executive quipped that the users who participate in Meta discussions represent "0.015%" of the site's visitors, and therefore weren't worth listening to -- despite these users contributing around 80% of the site's content). The company also announced plans to illegally change the licensing terms of user-submitted content, fired a volunteer moderator with no warning or explanation over a blatantly false allegation of transphobia [1], issued false and defamatory statements about that moderator to The Register; and fired two staff members who were just about the only employees who still engaged with the community and pushed back internally on the company's bad decisions [2]. A huge chunk of the community (including myself) stopped participating after everything that happened in 2019, and the site has seen a massive decline in quality and engagement since then.
In the wake of this, the company did start to make some changes -- they started engaging more with the community, and started working again on new features and changes, including moderation improvements -- but it's clear that people in the company working on the core site and engaging with the community have limited resources and very little influence with the company, and their work is too little, too late.
Then, the company got bought out by a private equity firm. The new CEO seems convinced that AI is the solution to monetizing Stack Overflow, somehow. This started with implementing half-baked AI features that didn't work, and the latest iteration seems to be trying to convince AI companies to pay for access to the site's content -- never mind that it's all available for free under a creative commons license. He's been trying to lock down the site's data dumps in an attempt to restrict AI companies from being able to download them, most recently with another attempt at illegally relicensing CC-BY-SA content [3]. Never mind that the drastic decline in engagement over recent years means whatever value the site has for AI training comes from its historical content which has long been freely available; not its new content [4]. (The most entertaining part of this debacle is that whenever the company announces some misguided new initiative, droves of ex-employees show up on Meta to tell them what a bad idea it is.)
So, TL;DR: the company neglected their core platform for years, lost or fired their most experienced staff, and lost most of their userbase and engagement. Now they're burning money, running an outdated platform that's not up to the task, and are caught off-guard by ChatGPT.
[0]: https://meta.stackexchange.com/q/285889
[1]: https://meta.stackexchange.com/q/333965
[2]: https://meta.stackexchange.com/q/342039
[3]: https://meta.stackexchange.com/q/401324
[4]: https://fosstodon.org/@JasonPunyon/112792203373096690