The quality of the output from a chatbot will depend on the quality of its database. It has no inherent way to tell whether the examples it finds are correct, so the output will be a blend of "truth" and "lies" (lies includes out of date info, mistakes, typos, and malicious posts). If you are lucky, the "truth" will predominate and you will get a good answer. But if you ask about a situation where many of the existing posts are wrong, confused or contradictory, the result will be poor. Until AI learns to grade its sources and add a "quality of sources" indicator to its output, a chatbot will not provide you with reliable data, and you follow its advice at your own risk.
Chatbots can give simple, correct examples. OTOH, they can also produce rubbish. Neither they nor you can tell before trying it which side any particular output is on.
Chatbots can give simple, correct examples. OTOH, they can also produce rubbish. Neither they nor you can tell before trying it which side any particular output is on.
Statistics: Posted by davidcoton — Sat Jan 04, 2025 6:10 pm