Lisa Feldman Barrett discusses the flaws in the use of artificial intelligence to accurately capture emotions from images. These emotional depictions are just a single snapshot in time. Human emotions are a collection of inputs and not one facial expression can give meaning to all of the variations of the expression of the emotion.
Artificial Intelligence currently has a western world bias on what emotions look like and their interpretation. However there is not a one to one expression and meaning correlation. For example just because someone is smiling doesn’t mean they are happy, it could actually be a reaction to something that is unpleasant, but it is up to the individual to know the truth.
What is interesting to me is the way that science tries to strip things down into cold hard facts. Even during this presentation, just talking about emotions felt slightly mechanical and data centric. I don’t think facial expressions is the best way to analyze the emotions, and I did have a memory come up where people were isolating the facial muscles used to make expressions and correlate it to the emotions that were expressed subconsciously as microexpressions (Blink Malcom Gladwell, page 206 section 3. The Naked Face). This was in support of thin slicing and using body knowledge to determine things faster than conscious awareness.
I see the need to use something different when it comes to emotional recognition. Perhaps it is more of a frequency and vibrational measurement vice a visual image.
David Hawkins is another researcher and author that measured the frequency of emotions and determined that there was a range that can be depicted as a scale of consciousness. This scale ranks the emotions from low to high. The data was collected from people that were expressing emotions and measured using frequency in Hz. Emotions are energy in motion and they are an inner world exploration much of the time.
Emotions are comprised of various components that are related to thoughts and sensations in the body. They also rely on attaching meaning to the feeling, thought, or situation. Just because you experience one facial expression during the emotion, doesn’t mean that you have to whole story from that one moment. An individual’s history, mindset, self regulation, and comfort with emotions all play a part in the expression and impact the emotions have on a single person.
I am a believer that thoughts create emotions which were touched on briefly in the presentation. The thinking that is happening creates the feeling regarding the memory or experience. This is also why it is possible to rewrite these memories and shift the energetic frequency of them from a lower vibration to a higher vibration because we can chang ethe way we think about something. This is the work that is happening in regards to quantum physics and even Dr. Joe Dispenza who tests brain wave signals during meditations focusing on emotions. Thoughts become things in the words of Mike Dooley who has written such books as Manifesting Change and Infinite Possibilities. This concept isn’t new, yet it kind of seems like it isn’t well integrated into the argument. Emotions have a level of complexity that will require a different kind of study to really be able to read and categorize people’s outward expressions with any kind of accuracy.
My hope is that facial expression artificial intelligence will not be the sole identifier in learning about and exploring emotional expression and their interpretations. We already get this wrong so often because there is an exchange and a dialog required to truly understand what a person is experiencing internally. There are inherent biases in the assignment of meaning to the expressions. We as a collective are starting to bring attention to feelings and allow them to be expressed after generations of shutting down and stifled expression. There is a great potential of mislabeling that would lead to stunting emotional expression if caution is not utilized in this area of research. People would not feel acknowledged and validated if the AI was wrong, why trust it.
Leave a Reply
You must be logged in to post a comment.