 Scientists from Columbia University have revealed that current AI models have limitations when it comes to distinguishing nonsense from natural language inputs. One of the authors of the research paper said it was premature to allow these models to replace human decision making in areas like law, medicine and student evaluations. The good news is that all of this advances so quickly. So it's the bad news but it's also the good news because hopefully we can train these models a little bit better to start recognizing nuance and forming those language structures a little bit more accurately and in ways that just make sense. We've only really been using them for not even a year yet. You know chat GBT hasn't even been out for a year in the way that it is out now. So it's amazing that you know like you know how fast we've moved and okay yeah I'm gonna figure out how to trick this thing we're actually trying to fool something that hasn't even been out for a year yet.