Also people need to understand that reading microexpressions is like using a polygraph aka ‘lie detector’ where there are common reasons for reactions but not everyone reacts to everything the same and one behavior or reaction doesn’t always mean the same cause.

Like it could be a great tool to help those who agree with how it is working, but since it will be developed by someone who reads way too much into microexpressions I expect every single implementation to feel like gaslighting for someone like me who understands myself.

The ai bullshit in the article sounds even worse!