You are viewing a single comment's thread:
Many of ChatGPT's biases are well documented. Many other biases no doubt remain undetected as-of-yet, and even more are likely undetectable.
Such biases are by far the greatest threat these tools pose to humanity, imho. The biases are ingrained in a dramatic fashion; they are inseparable from the tool. The vast majority of users will remain completely unaware that the results are subject to such biases.
Look at the level of media manipulation of the masses we have nowadays and imagine how that manipulation will exponentially grow with the help of ai...
View more