I’m reminded of time learning a new language, and how little it takes to completely change the meaning of something.
Imagine if you thought you understood 90% of what somebody said to you but you didn’t catch the word “not”; you would have exactly the wrong interpretation. Or what about skipping a modifier like “slightly” or “significantly”? (And don’t even get me started on sarcasm or other things that may not be detectable to all.) Google’s summaries, or any auto-summarization, risk inventing a new conclusion that was never part of the original.
This is a cornerstone to critical thinking as well. If you spend your time trusting one-sentence tweets and other shallow writing, are you also asking yourself what hasn’t been stated? For example, I could say: “Engineers quitting Google amid latest executive actions”; the reality might be “it was exactly two engineers”, “their reasons for quitting were unrelated”, etc. It is extremely easy to create misleading summaries.
translate.google.com somewhat prone to misnegation. Basically one word doesn't change the score that much, and there are things like different word order that overwhelm it.
Imagine if you thought you understood 90% of what somebody said to you but you didn’t catch the word “not”; you would have exactly the wrong interpretation. Or what about skipping a modifier like “slightly” or “significantly”? (And don’t even get me started on sarcasm or other things that may not be detectable to all.) Google’s summaries, or any auto-summarization, risk inventing a new conclusion that was never part of the original.
This is a cornerstone to critical thinking as well. If you spend your time trusting one-sentence tweets and other shallow writing, are you also asking yourself what hasn’t been stated? For example, I could say: “Engineers quitting Google amid latest executive actions”; the reality might be “it was exactly two engineers”, “their reasons for quitting were unrelated”, etc. It is extremely easy to create misleading summaries.