I’ve been talking and talking about emotions and hinting that AI lack it. Rambling about that emotions are essential, because logic can’t choose.
Here’s a flip-flop I didn’t expect while thinking about emotional AI. What if the answer isn’t AI with emotions? What if humans and AI are already one complete system?
If you look closely, emotional weighting already exists. It’s just not inside the model.
When we use AI today, we form a hybrid system:
- AI handles logic, pattern recognition, synthesis;
- Humans provide values, emotional weighting and prioritization;
Together, it works! AI evaluates options Humans assign value
From this perspective, many AI “limitations” are boundaries of responsibility.
So what the problem then?! Why all of this thinking was needed in a first place?
Because problem still appears when humans aren’t in the loop: scale, availability, emotional exhaustion. So, how do we design systems where humans and AI complement each other, without exhausting the humans in the process?