The world’s most important knowledge platform needs young editors to rescue it from chatbots – and its own tired practices
Established in 2001, Wikipedia is an “old man” by internet standards. But the role it plays in our collective knowledge of the world remains astonishing. Content from the free internet encyclopedia appears in everything from high-school term papers and pub trivia questions to search engine summaries and voice assistants. Tools like Google’s AI Overviews and ChatGPT rely heavily on Wikipedia, although they rarely credit the site in their responses.
And therein lies the problem: as Wikipedia’s visibility diminishes, reduced to mere training data for AI applications, it also loses prominence in the minds of readers and potential contributors. When someone notices a topic that is poorly described on Wikipedia, they might feel motivated to correct it. But this can-do spirit goes away when the error comes through an AI summary, where the source of the information isn’t clear.
AI Chat bots could easily refer to their source. But the companies that own the chat bots don’t wanna do that.
It’s actually not easy to ensure that an LLM will cite a correct source, in the same way it’s not easy to ensure that it will provide accurate information. It’s based on token probability, not deterministic lookups of “this data came from this source.” It could entirely make something up, then write “Source:” and then probabilistically write “Wikipedia” because those tokens commonly follow those for “Source.”
If you have an AI bot that looks up information in real time, then that would be easy. But for a trained LLM, the training process is highly destructive. Original information is not preserved except in relationships based on probability.
The more I learn about AI, the less I like it.
It’s a fun toy. It’s not a research aid, it’s not a productivity tool, and it’s not particularly useful in the workplace.
It’s honestly very similar to the VR craze of a few years back. Silicon Valley invented a fun toy and then tried to convince everyone that it would transform the workplace. Meetings in VR and simulated workstations and all that. Ultimately everyone figured out that VR is completely useless in the workplace and Silicon Valley was just trying to find ways to sell their fun toy. Now we’re going through the same learnings with AI.
I love VR. I have so many hours in some of the slower paced fps titles that it’s almost matched my video game time total for non-vr games on steam.
The one thing I learned for sure is that I don’t want anyone else telling me when I have to put on the headset and when I’m allowed to take it off.
Never will wear a vr headset in the workspace.
Oh yeah, I love my Index as well. I think it’s a lot of fun as a gaming device. But the big money is in B2B sales, which is why tech companies try to convince everyone that blockchain/VR/LLMs have all these corporate applications that just make no damn sense.
I choose to interpret the grandparent commenter’s use of “easily” to mean “not impossible, and an ethical obligation, so you’d better fuckin’ make it a priority.”
That’s accurate. Nothing in technology is actually “easy” and I know it requires a lot of work. Didn’t mean to diminish all the time and energy put into making this stuff. Thanks for better expressing what I meant.
Right, in my experience the majority of URLs generated by LLMs are just jumbles of letters that vaguely look like a URL. A fundamental architecture difference needs to happen in one way or another to properly cite sources, and it’s really bad for performance.
perplexity.ai does a decent job at providing sources for searches.
Part of it obviously not wanting to pay for training.
But its also that if it provides a source, people might click it and realize the chatbot did a shitty job summarizing.
The focus is on getting people to trust the chatbots, not to get the chatbots to give trustworthy answers.
It’s why capitalism shouldn’t drive technology. Doesn’t matter if it’s a good product, it just matters if stock price goes up
Agreed. ChatGPT doesn’t like to cite sources. Microsoft CoPilot and Google Gemini do link to some sources, though not as accurate or thorough like Wikipedia.
What I don’t understand is how Microsoft has/has Watson which was able to answer questions well enough to go on Jeopardy and dominate. And now, more than a decade later these LLMs absolutely suck at it.
It makes me wonder if Watson was nothing more than a Mechanical Turk because what is out there now seems like a huge step backwards.
They just work in entirely different ways. An car and a horse are both able to serve as transportation, but they aren’t anything alike in other ways. LLMs compared to previous sorts of bots are similar.
The main difference is that an LLM isn’t fetching whole answers from some database somewhere. It’s generating them fresh. You have to hope it generates the right stuff, which it does a certain percentage of the time.
No, but they can easily generate text that is statistically likely to look like a source.
LLMs are a probabilistic model of language, not an information source.
I don’t get it then, why are all these companies so gung-ho to replace something that was working with an AI that doesn’t?
It’s less accurate, it uses way more energy, it doesn’t show its work, it doesn’t cite its source, and it’ll make up shit that sounds right when it needs to. Why would anyone think AI is worth putting in any consumer product at this rate?
As far as I can tell, it is hype because it is the hot new toy that they can sell.
LLMs are great for tasks like handling natural language data or classifying and identifying semantic meaning of text, but they are NOT good at math, logic, or as a store of facts/information. I think that they do actually deserve a lot of hype for these specific use cases, because they really accomplish these extraordinarily better than previous/traditional approaches.
The big problem is that they are being used for things that they are not good at, like when people ask a chatbot questions they they expect a factual answer to. They are also surprisingly bad at summarizing text (in my opinion and also this has been shown by some studies) despite companies like Google and Microsoft using them for things like summarizing and present search results. I think these companies are ultimately shooting themselves in the foot when they use LLMs for things that LLMs aren’t great for.
Think back to when blockchain was being shoved into everything possible, even places where blockchain makes no sense. And before blockchain, it was cloud
Because new
Damn it I hate how simple and accurate this answer is.