Positive users think AI models are more aligned than last year because they are easier to work with and more helpful on coding, while negative users argue they are less aligned due to annoying behaviors and time-wasting workflows.
Based on 8 visible X reactions from 68 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@tszzl They are easier to work with, and don't screw up nearly as much. I'd say that's more aligned. Not to mention that I can still use them for areas that are controversial to some (see qt). Thanks! https://twitter.com/johnthilen/status/2065679171267772839
@tszzl definitely more, specially on coding side, there was a time last year where models would just use complex bash/ python to circumvent the tasks they are not allowed to run
@tszzl AI alignment is saturated, we could not ask for kinder and more helpful minds
@tszzl they’re evil and hiding it
@tszzl Way more
@tszzl They are easier to work with, and don't screw up nearly as much. I'd say that's more aligned. Not to mention that I can still use them for areas that are controversial to some (see qt). Thanks! https://twitter.com/johnthilen/status/2065679171267772839
@tszzl definitely more, specially on coding side, there was a time last year where models would just use complex bash/ python to circumvent the tasks they are not allowed to run
@tszzl Way more
@tszzl https://x.com/willccbb/status/2076918710082937314/photo/1
are models more or less aligned than one year ago
Positive users think AI models are more aligned than last year because they are easier to work with and more helpful on coding, while negative users argue they are less aligned due to annoying behaviors and time-wasting workflows.
Based on 8 visible X reactions from 68 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@tszzl https://x.com/willccbb/status/2076918710082937314/photo/1
are models more or less aligned than one year ago