The article does not show up without JS, so I'll answer the title. I agree. In practice, you'll likely need to double check what it says using a search engine

The article does not show up without JS, so I'll answer the title. I agree. In practice, you'll likely need to double check what it says using a search engine anyway. As long as the Chatbot does not provide the sources for whatever it states, I won't trust it. There's no reason to. Unless it produces pure logic that can be checked using reasoning. … as I have in real life from people unless they are expert in the area / can be trusted to varying degrees for some reason, or from anything on the internet, including Wikipedia. AI-based stuff hasn't changed this fundamental principle. ChatGPT is trained on Wikipedia (which is mostly fine, though you still need to check the sources!), plus random things where people could have stated random stuff, maybe jokingly, or could have been sarcastic or whatever. So the training set is already not trusty. And even on a trusty set, the process is so opaque that we don't know what plausible but false rehash it's capable of outputting. Actually we know that it produces what has been called "hallucinations", so… It's easy to get lazy about double checking the overly confident tone or the flawless writing of the chatbot. It's probably a good way to explore, but it can't be fully trusted yet. Even for exploring, I'm afraid of risking being influenced by reading a false piece of information written perfectly, even if I know I should double-check it. However, the day the chatbot provides the sources, it might change a lot of things.