Raw LLM Responses

>"It's just autocomplete on steroids, no real intelligence" wrong. >It's going to replace entire industries" possible. >"It doesn't actually understand anything" define "understand".Though Possible. >"It can write better code than most programmers". possible >"It has no consciousness, just pattern matching". wrong. oversimplification. >"It's passing medical boards and bar exams". true >Which one is it? answers are as above. >it's sophisticated enough to threaten millions of jobs, true >it's just fancy predictive text that doesn't really "get" anything. likely wrong >It can't be both. wrong >Here's my theory: We keep flip-flopping because admitting the truth is uncomfortable for different reasons: possible. >If it's actually intelligent: We have to face that we might not be as special as we thought. agreed >If it's just advanced autocomplete: We have to face that maybe a lot of "skilled" work is more mechanical than we want to admit. agreed >The real question isn't "Is ChatGPT conscious?" or "Will it take my job?". Wrong. Those are real questions. >The real question is: What does it say about us that we can't tell the difference?. True. that is also a real question. >Maybe the issue isn't what ChatGPT is. Disagree.what chatGPT is part of the issues. >Maybe it's what we thought intelligence and consciousness were in the first place. Agree. It's one of the issues.