Emotion-reading algorithms cannot predict intentions via facial expressions. Though algorithms are increasingly being deployed in all facets of life, a new study has found that they fail basic tests as truth detectors. (1)
How could they not fail?
We cannot see it now, but algorithms will forever fail.
For even when they succeed, they will lack the main characteristic of human spirit: Knowledge of what they are doing. Even if they understand someone is sad, they will never really know what ‘sad’ is. And what is really sad is that we start forgetting ourselves what emotions are really about. Day by day, instead of the computers being able to think more like us, we tend to think more like them.
One day we will be happy that the algorithms we made have succeeded.
Not because they will have succeeded.
But because we will have failed.
And only then, will the algorithms detect true sadness for the first time…