last day (117 days later) » 

4:06 PM
room topic changed to Discussion about Prompt Design + New site creation: Discussion with staff/others about the creation of the Prompt Design site, as well as the idea of a new way to create/define sites (no tags)
room topic changed to Discussion about Prompt Design + New site creation: Discussion with staff/others about the creation of the Prompt Design site, as well as the idea of a new way to create/define sites. Relevant post: meta.stackexchange.com/q/390463 (no tags)
So to address Seb's initial point
For those who can't see that (and since it's being addressed directly here), Seb said:
there's two potential options here:
1. This new site is a gold rush thought of < 1 mo ago. At which point the executive decision is correct (there wouldn't be time to go via A51), the motive is lame, and they know the site will die
2. This new site was thought of as developing genAI trends showed up, which means it may potentially have been on the cards for 6mo or even longer. At this point, why not A51 it?
Prompt Design was first imagined in February. I know this because I was the first to hear about it outside the group who thought about it initially, which I will clarify was not anyone a part of the C suite.
@SpencerG So why not put it through Area 51 as soon as it was imagined? That's the point of Area 51. Setting aside all the known issues and complaints with A51, it's what we have.
Well, one big thing about A51 is that proposals need a pre-existing community in order to get off the ground. Without one, it'll become quickly apparent that one doesn't exist in the early stages of a proposal with a lack of activity, and about a week in, the proposal will have died.
5
I don't suspect this, particular, site proposal would've had much buy-in, but I may be wrong.
4:15 PM
@ThomasOwens There is a number of reasons, it wasn't so that it could line up with the we are developers conference, the original launch date for this was actually July 10th. We originally were going to share the Mod update on May 30th, as it the announcement was written in early May, but given the mod action just starting we wanted to give that some air time, cause the optics would have been worse. So the whole timeline got moved around.
2
That's kind of the point of A51, though. The company has something that most people don't - sales and marketing people who could have used their contacts to sell the community. Or other company business partners.
@SpencerG So it's a trendy marketing thing. I'm still not convinced that you have enough good, high quality, on-topic questions to populate a site. Look at the history of Software Engineering, for example. Highly opinionated questions don't do well. But also, questions that will inherently go out of date very quickly (which, my understanding of prompt design, is that good prompts will become useless with the next generation of models) also don't do well.
2
By bypassing A51, you have no user base, no demonstration of the existence of good questions, no commitment to participate. I guess I don't know what you do have.
(You meaning the company, not you personally. Writing out "the company" is verbose.)
2
@ThomasOwens Marketing and Sales is something that we are leveraging here. The problem is that we think a lot of the most interested users for this site are not currently on the network. While there may be many, who would be interested the likely majority are elsewhere. So the process of doing promotion and sending them all to Area 51 one to participate in that process just had a lot of issues with it just from a user experience.
Which its definitely legit to say that our fault and we could have fixed it, but we didn't really have time to fix it quickly enough so it was a bit of a nonstarter.
@SpencerG I'll re-state what I said in Discord earlier: A lot of moderators have come forward and stated that they feel their feedback on the mod-only update about this decision was ignored or left unconsidered. Given the sentiment, was the decision to publicly announce this just one of trying to get broader community feedback, or a means of trying to kinda... Push this thing along?
My concern is that, if the feedback from mods wasn't considered at the mod-only phase, what guarantee do we have that the public feedback will be considered?
@SpencerG This is a problem with A51 in general, though. My two biggest problems have been (1) network fracturing and (2) no outside growth. A lot of sites proposed through A51 today are offshoots of existing sites. That's...not good for engagement metrics on existing sites as people can't spread their time everywhere. So the company saw an existing problem with A51 and, using their power, ignored it to push through their wish.
@Spevacus Yeah and that's totally fair. When I went through the feedback I tried to address the things that I definitely thought would come up, which is why some additional tweaks to it were added, but you are right to say that not everything was, and that was intentional. Mostly just so that we could see how the larger community responded as well, so that is likely on me for not framing that correctly with the mods.
4:25 PM
@ThomasOwens I'm not quite following your argument - it seems to me you're both arguing that this should have gone through A51, and that A51 is fatally flawed. I suppose if you want the proposal to die that makes sense, but it doesn't seem like it would make sense to Spencer or the company to take that route
I do have work now, but those are my biggest concerns. Thank you, @SpencerG, for taking this time. Having this kind of communication and transparency about other things earlier would have prevented a lot of this mess. I hope that it can be done earlier and often for other things. I'll check in later tonght.
@BryanKrauseisonstrike Area 51 is flawed and we, as users, have been saying that for a very long time. But it's been ignored. And now that people at the company see an opportunity, they short circuit the whole process and jam a new community into the network. It seems suspect, that's all. We don't have the power to do that. No one outside the company does.
I... created a tag for the new site (promptdesign) on MSE, but @SonictheAnonymousHedgehog (correctly) rolled it back because I accidentally failed to meet the naming guidelines for tags. That user said to wait until my tag was automatically removed, and then the correct one (prompt-design vs promptdesign) can be created. But, I've already made a tag wiki for the wrong one, so I don't think it will actually get deleted automatically.
We can't just create the right tag and merge it because the tag names are similar. What's the right way to fix this? Can staff members/mods bypass that message to make the new tag, and then do a tag merge?
@ThomasOwens Thank you for explaining - yes, that complaint makes sense to me
@cocomac if there are no posts, the tag will roomba, tag wiki or no
@ThomasOwens Right, which is why its also an experiment, "How do we build new sites and communities when the community may not be on the network?" For example, this Facebook group had just under half a million users in it when I first started looking for existing communities. Now it's just under a million. That is also not even addressing the known and well documented issues Area 51 has.
So we think they are out there and sending them to Area 51 via marketing just wasen't tenable.
4:29 PM
@Spevacus TIL - I (wrongly) thought tag wikis block it. Thank you!
When it comes to site incubation, I think most people agree Area 51 could be improved in a number of ways. So this really was an experiment to try something to see if it's a possible route. As I am sure anyone knows with any company there are lots of competing and important priorities all over the place.
So using this as an experiment really does have the possibility for us to get it on the docket, though I do want to specify this isn't alluding to some kind of pending initiative that is just around the corner.
One thing I think is worth noodling on: A new site incubation process would be neat. I think we're all in agreement that Area 51 isn't perfect, and something that's built into the global Meta (MSE) could be interesting. However, kicking this idea off with a Prompt Design site, especially when we're all agitated about AI at the moment, might not be the best way to trial this kind of functionality.
5
I feel as though the way that this site proposal navigates this new process would be... Artificially doomed to fail, or succeed, depending upon a variety of factors not inherent to the way we're used to founding a site.
@SpencerG Just saw your post on MSE. FYI: I haven't read it fully, but it looks like the text under What Would Stakeholders Be Doing? and also in How to Join the Community Stakeholders Group is somewhat broken... was that meant to be a bulleted list?
4:46 PM
@cocomac I threw this into a bulleted list
@Spevacus Thanks!
Thank you for clarifying; copying over here so it's all in one place. If it was in the works since February - why roll it out this quickly, on this short notice, with a timeline that looks more like a PR push for a conference than an organic community growth?

In this < 2 month runway, you will have to:
1. Iterate on what questions are/aren't acceptable
2. Curtail and limit the scope of certain topics, or find ways to tag things properly (if it's prompt help, even a minor version of GPT can have major differences in techniques)
4:57 PM
I would love to see investment in communities but personally, having a new Area51 after we just had to fight to keep the dumps... Involving more people without knowing where this is all going... when we just learned the smaller sites are a dead end... meh. There would need to be more than this for me to entertain being more involved.
5:25 PM
I just find the whole… stack rolling out the red carpet for this community while not doing so for others a bit disappointing. I get it, from a marketing perspective, that having more AI communities here for marketing to talk about will be useful for fooling more businesses into buying into teams, just disappointing. 🤷‍♂️ if it’s a community that has lasting value, it wouldn’t hurt for it to go through the normal process.
4
Sorry I have been in meetings and am currently in one, but I will circle back in a bit.
5:46 PM
Random suggestion: you (or other SE staff) could offer a badge for the Community Stakeholders, similar to this one for the Staging Ground
There's already a Beta badge, IIRC
also... it seems odd to ask users to publicly volunteer to participate using an answer on MSE, where there's a lot of negative sentiment already about the site here
Where else?
Twitter?
a chat room, where you won't get obliterated with downvotes
a signup form offsite
@KevinB You'll get obliterated with a super-starred response.
@KevinB Marketed how? At the conference? Seems like they want this stuff done before Prashanth holds his speech. I guess they don't worry so much about having an ongoing strike during the speech. I guess he can preach about replacing an angry mob with AI to seamlessly help the world's developers without the annoying distractions of the loud screamers.
6:02 PM
@Spevacus What makes you think this is a trial and not a one-off for this site?
No. just link to it in the MSE post. it'll receive the same "marketing" as the mse post in that form without the negatives.
They probably want to learn a thing or two from it to improve the process for other sites, but is there anything that indicates all future new sites will go through the same process?
Future new sites will be established when ChatGPT decides it needs a new playground.
@AndreasdetestsAIhype That's not how it works.
This process, if it became the norm, would mean all future sites would need to be in some way beneficial enough to the business ends of things to justify the CM involvement this requires.
6:05 PM
@Mast-onstrike Unless you missed the joke, I'm waiting for the punchline. ;)
@AndreasdetestsAIhype If that's a joke, I do not consider it funny.
@Mast-onstrike Ok.
6:23 PM
@cocomac Yeah its being considered. It's not a hard no, its just not something that would happen quickly if it does. Don't get me started on badges, though; been advocating for a custom badge tool for almost all of my two years. I think we will get one day though
@KevinB Uh, that is definitely not a thing. CMs are heavily involved here, because it's a new process, and we think it needs a CM to guide it. Mostly so it can make it to its destination and so we have an eyewitness viewpoint of it going completely off the rails.
So... This kind of hand-holding seems to me like it would be mandatory for all sites created in this way. How would a community approach SE about potentially creating a new site about their area of expertise? Area 51 offered a way for them to do this, even if it required a pre-existing SE account and knowledge of how it works.
This new process seems like it would need some form of direct communication with SE to start.
well, they'd buy a SOfT and build a community there, of course.
@Spevacus I mean the major contributing factor here is that this is company-led, though I don't know from any of the various speculations I have witnessed about a potential future site incubation process would require CM's be involved.
Though, it has been mentioned it might be more beneficial if they were to. It would honestly be something I am interested in discussing at some point if/when we get there.
Just broadly speaking about community management as a profession that the way SE spins up communities is not typical to how the profession typically starts one from scratch. Not to say that either method is superior, but there are likely principles from both to help that out.
@SpencerG Maybe this should be a comment, but: the MSE post states Early moderation of the site (Moderation by appointment, or CMs moderate [...]) and then These individuals are not granted any special tools or powers - would the Stakeholders have to be ♦ mods on the new site, or just "community mods" with normal user abilities, like the non-mod queues? (I'm considering asking to join, but I wouldn't be comfortable having a ♦ right now)
my interpretation is that, the stakeholders would discuss whether they needed some appointed mods to start things out
6:36 PM
@cocomac I can edit that piece in a bit. Stakeholders wouldn't get a diamond. We do want the stakeholder group to decide how the site should initially be moderated since it will launch with out them. Whether they just want CM's to appoint interested people from day 1 or if they would prefer the CM's do it till formal elections can be held.
@SpencerG That makes sense, thanks for clarifying!
Honestly, from my experience on Programming Language Implementation, I think a mod would certainly help some things. Closing is available to everyone, but some reviews need moderator handling early on (such as tag wiki/excerpt edits), and tag merging/synonymization. Relying on a CM for these tasks, who has other duties to attend to, isn't really great, even if it does get done eventually (Shoutout to Catija for her assistance :D)
Part of the reasoning for not deciding one way or another was because there have been times when in the past, it has been requested that we appoint mods on sites, and we stopped doing it. So we thought about just going the appointment route but ultimately thought it would be better to leave it up to the group to decide on.
 
1 hour later…
8:10 PM
I read the request for participation in the stakeholder group. I really hope that you're able to get the right people involved. Consider people who are skeptical of AI to help tamper any conversations or people with expertise in IP law (there are open course cases involving copyright and generative AI) or ethics, which are of utmost importance. Although I'm partially interested, I don't have energy to spend on this community when there are concerns for the network and my existing communities.
3

  last day (117 days later) »