Wikipedia Pauses an Experiment That Showed Users AI-Generated Summaries at The Top of Some Articles, Following an Editor Backlash.
Technology
40
Beiträge
30
Kommentatoren
1
Aufrufe
-
Articles already have a summary at the top due to the page format, why was AI shoved into the process?
Because AI
-
Hey everyone, this is Olga, the product manager for the summary feature again. Thank you all for engaging so deeply with this discussion and sharing your thoughts so far.
Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March. As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further. With that in mind, we’d like to take a step back so we have more time to talk through things properly. We’re still in the very early stages of thinking about a feature like this, so this is actually a really good time for us to discuss here.
A few important things to start with:
- Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such.
- We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.
- With all this in mind, we’ll pause the launch of the experiment so that we can focus on this discussion first and determine next steps together.
We’ve also started putting together some context around the main points brought up through the conversation so far, and will follow-up with that in separate messages so we can discuss further.
How about not putting AI into something that should be entirely human controlled?
-
How about not putting AI into something that should be entirely human controlled?
Yeah as more organizations implement LLMs Wikipedia has the opportunity to become more reliable and authoritative. Don't mess that opportunity up with "AI."
-
Hey everyone, this is Olga, the product manager for the summary feature again. Thank you all for engaging so deeply with this discussion and sharing your thoughts so far.
Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March. As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further. With that in mind, we’d like to take a step back so we have more time to talk through things properly. We’re still in the very early stages of thinking about a feature like this, so this is actually a really good time for us to discuss here.
A few important things to start with:
- Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such.
- We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.
- With all this in mind, we’ll pause the launch of the experiment so that we can focus on this discussion first and determine next steps together.
We’ve also started putting together some context around the main points brought up through the conversation so far, and will follow-up with that in separate messages so we can discuss further.
So they:
- Didn't ask editors/users
- noticed loud and overwhelmingly negative feedback
- "paused" the program
They still don't get it. There's very little practical use for LLMs in general, and certainly not in scholastic spaces. The content is all user-generated anyway, so what's even the point? It's not saving them any money.
Also it seems like a giant waste of resources for a company that constantly runs giant banners asking for money and claiming to basically be on there verge of closing up every time you visit their site.
-
How about not putting AI into something that should be entirely human controlled?
These days, most companies that work with web based products are under pressure from upper management to "use AI", as there's a fear of missing out if they don't. Now, management doesn't necessarily have any idea what they should use it for, so they leave that to product managers and such. They don't have any idea, either, and so they look at what features others have built and find a way to adapt one or more of those to fit their own products.
Slap on back, job well done, clueless upper management happy, even though money and time have been spent and the revenue remains the same.
-
Hey everyone, this is Olga, the product manager for the summary feature again. Thank you all for engaging so deeply with this discussion and sharing your thoughts so far.
Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March. As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further. With that in mind, we’d like to take a step back so we have more time to talk through things properly. We’re still in the very early stages of thinking about a feature like this, so this is actually a really good time for us to discuss here.
A few important things to start with:
- Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such.
- We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.
- With all this in mind, we’ll pause the launch of the experiment so that we can focus on this discussion first and determine next steps together.
We’ve also started putting together some context around the main points brought up through the conversation so far, and will follow-up with that in separate messages so we can discuss further.
Summarization is one of the things LLMs are pretty good at. Same for the other thing where Wikipedia talked about auto-generating the "simple article" variants that are normally managed by hand to dumb down content.
But if they're pushing these tools, they need to be pushed as handy tools for editors to consider leveraging, not forced behavior for end users.
-
So they:
- Didn't ask editors/users
- noticed loud and overwhelmingly negative feedback
- "paused" the program
They still don't get it. There's very little practical use for LLMs in general, and certainly not in scholastic spaces. The content is all user-generated anyway, so what's even the point? It's not saving them any money.
Also it seems like a giant waste of resources for a company that constantly runs giant banners asking for money and claiming to basically be on there verge of closing up every time you visit their site.
If her list were straight talk:
- Were gonna make up shit
- But don’t worry we’ll manually label it what could go wrong
- Dang no one was fooled let’s figure out a different way to pollute everything with alternative facts
-
Summarization is one of the things LLMs are pretty good at. Same for the other thing where Wikipedia talked about auto-generating the "simple article" variants that are normally managed by hand to dumb down content.
But if they're pushing these tools, they need to be pushed as handy tools for editors to consider leveraging, not forced behavior for end users.
If we need summaries, let's let a human being write the summaries. We are already experts at writing. We love doing it.
-
Hey everyone, this is Olga, the product manager for the summary feature again. Thank you all for engaging so deeply with this discussion and sharing your thoughts so far.
Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March. As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further. With that in mind, we’d like to take a step back so we have more time to talk through things properly. We’re still in the very early stages of thinking about a feature like this, so this is actually a really good time for us to discuss here.
A few important things to start with:
- Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such.
- We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.
- With all this in mind, we’ll pause the launch of the experiment so that we can focus on this discussion first and determine next steps together.
We’ve also started putting together some context around the main points brought up through the conversation so far, and will follow-up with that in separate messages so we can discuss further.
Is there a way for us to complain to wikipedia about this? I contribute money every year, and I will 100% stop if they're stomping more LLM-slop down my throat.
Edit:
You can contribute to the discussion in the link, and you can email them at addresses found here: https://wikimediafoundation.org/about/contact/ -
Hey everyone, this is Olga, the product manager for the summary feature again. Thank you all for engaging so deeply with this discussion and sharing your thoughts so far.
Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March. As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further. With that in mind, we’d like to take a step back so we have more time to talk through things properly. We’re still in the very early stages of thinking about a feature like this, so this is actually a really good time for us to discuss here.
A few important things to start with:
- Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such.
- We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.
- With all this in mind, we’ll pause the launch of the experiment so that we can focus on this discussion first and determine next steps together.
We’ve also started putting together some context around the main points brought up through the conversation so far, and will follow-up with that in separate messages so we can discuss further.
I passionately hate the corpo speech she's using. This fake list of "things she's done wrong but now she'll do them right, pinky promise!!" whilst completely ignoring the actual reason for the pushback they've received (which boils down to "fuck your AI, keep it out") is typical management behavior after they were caught trying to screw over the workers in some way.
We're going to screw you over one way or the other, we just should have communicated it better!
Basically this.
-
Summarization is one of the things LLMs are pretty good at. Same for the other thing where Wikipedia talked about auto-generating the "simple article" variants that are normally managed by hand to dumb down content.
But if they're pushing these tools, they need to be pushed as handy tools for editors to consider leveraging, not forced behavior for end users.
Summaries that look good are something LLMs can do, but not summaries that actually have a higher ratio of important/unimportant than the source, nor ones that keep things accurate. That last one is super mandatory on something like an encyclopedia.
-
Articles already have a summary at the top due to the page format, why was AI shoved into the process?
Grok please ELI5 this comment so i can understand it
-
These days, most companies that work with web based products are under pressure from upper management to "use AI", as there's a fear of missing out if they don't. Now, management doesn't necessarily have any idea what they should use it for, so they leave that to product managers and such. They don't have any idea, either, and so they look at what features others have built and find a way to adapt one or more of those to fit their own products.
Slap on back, job well done, clueless upper management happy, even though money and time have been spent and the revenue remains the same.
Wikipedia can create a market niche by stating the authenticity of their content being 100% human. Some of the stupid upper management types understand being unique as a marketing strategy.
-
If her list were straight talk:
- Were gonna make up shit
- But don’t worry we’ll manually label it what could go wrong
- Dang no one was fooled let’s figure out a different way to pollute everything with alternative facts
Your last point states it all. Rather than being a source of truth, it is now meant to bend the truth. 2 plus 2 no longer equals 4.
-
Hey everyone, this is Olga, the product manager for the summary feature again. Thank you all for engaging so deeply with this discussion and sharing your thoughts so far.
Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March. As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further. With that in mind, we’d like to take a step back so we have more time to talk through things properly. We’re still in the very early stages of thinking about a feature like this, so this is actually a really good time for us to discuss here.
A few important things to start with:
- Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such.
- We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.
- With all this in mind, we’ll pause the launch of the experiment so that we can focus on this discussion first and determine next steps together.
We’ve also started putting together some context around the main points brought up through the conversation so far, and will follow-up with that in separate messages so we can discuss further.
I can't wait until this "put LLMs in everything" phase is over.
-
Summaries that look good are something LLMs can do, but not summaries that actually have a higher ratio of important/unimportant than the source, nor ones that keep things accurate. That last one is super mandatory on something like an encyclopedia.
The only application I've kind of liked so far has been the one on Amazon that summarizes the content of the reviews. Seems relatively accurate in general.
-
Hey everyone, this is Olga, the product manager for the summary feature again. Thank you all for engaging so deeply with this discussion and sharing your thoughts so far.
Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March. As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further. With that in mind, we’d like to take a step back so we have more time to talk through things properly. We’re still in the very early stages of thinking about a feature like this, so this is actually a really good time for us to discuss here.
A few important things to start with:
- Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such.
- We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.
- With all this in mind, we’ll pause the launch of the experiment so that we can focus on this discussion first and determine next steps together.
We’ve also started putting together some context around the main points brought up through the conversation so far, and will follow-up with that in separate messages so we can discuss further.
I canceled my recurring over this about a week ago, explaining that this was the reason. One of their people sent me a lengthy response that I appreciated. Still going to wait a year before I reinstate it, hopefully they fully move on from this idea by then. It sounded a lot like this though, kinda wishy washy.
-
How about not putting AI into something that should be entirely human controlled?
The sad truth is that AI empowers the malicious to create a bigger impact on workload and standards than is scalable with humans alone. An AI running triage on article changes that flags or reports changes which need more input would be ideal. But threat mitigation and integrity preservation don't really seem to be high on their priorities.
-
These days, most companies that work with web based products are under pressure from upper management to "use AI", as there's a fear of missing out if they don't. Now, management doesn't necessarily have any idea what they should use it for, so they leave that to product managers and such. They don't have any idea, either, and so they look at what features others have built and find a way to adapt one or more of those to fit their own products.
Slap on back, job well done, clueless upper management happy, even though money and time have been spent and the revenue remains the same.
I've already posted this a few times, but Ed Zitron wrote a long article about what he calls "Business Idiots". Basically, people in decision making positions who are out of touch with their users and their products. They make bad decisions, and that's a big factor in why everything kind of sucks now.
https://www.wheresyoured.at/the-era-of-the-business-idiot/ (it's long)
I think a lot of us have this illusion that higher ranking people are smarter, more visionary, or whatever. But I think no. I think a lot of people are just kind of stupid, surrounded by other stupid people, cushioned from real, personal, consequences. On top of that, for many enterprises, the incentives don't line up with the users. At least wikipedia isn't profit driven, but you can probably think of some things you've used that got more annoying with updates. Like google putting more ads up top, or any website that does a redesign that yields more ad space, worse navigation.
-
Summarization is one of the things LLMs are pretty good at. Same for the other thing where Wikipedia talked about auto-generating the "simple article" variants that are normally managed by hand to dumb down content.
But if they're pushing these tools, they need to be pushed as handy tools for editors to consider leveraging, not forced behavior for end users.
not forced behavior for end users.
This is what I'm constantly criticizing. It's fine to have more options, but they should be options and not mandatory.
No, having to scroll past an AI summary for every fucking article is not an 'option.' Having the option to hide it forever (or even better, opt-in), now that's a real option.
I'd really love to see the opt-in/opt-out data for AI. I guarantee businesses aren't including the option or recording data because they know it will show people don't want it, and they have to follow the data!
