They provide information --- some of which is random in nature and only casually reflective of any truth or reality.
And as this example illustrates, they are far from being trustworthy. Their main achievement is to consistently produce functionally acceptable grammar.
You’re asking a lossily compressed database with an imprecise and ambiguous query language interface about hard facts, you get a plausible reconstructed answer.
Work with the tool to get best results instead. You wouldn’t do csi style zoom enhance on a jpeg either.
So, when was it released? Did one of them get it right? Or are all readers about this article on LLM (non-)capabilities expected to be familiar with Cisco's product lines?
For someone enthusiastically using LLMs since GPT-3, the question gives off a strong vibe of not being a good question for a LLM. Is anyone still surprised by that? Doesn’t everyone quickly develop such intuition?
6 comments
[ 0.27 ms ] story [ 25.0 ms ] threadThey provide information --- some of which is random in nature and only casually reflective of any truth or reality.
And as this example illustrates, they are far from being trustworthy. Their main achievement is to consistently produce functionally acceptable grammar.
Work with the tool to get best results instead. You wouldn’t do csi style zoom enhance on a jpeg either.