The Machines Don’t Care. Yet.
Emotions remain one of the most complex facets of the human condition. So much so that even scientific discourse has no consensus on a singular definition. Emotion is loosely defined as the conscious experience characterized by intense mental activity and a certain degree of pleasure or displeasure. The ability to fully understand and decipher human emotion is something society has constantly debated, questioned and continues to research.
While intelligence has long been measured with a quotient called IQ, it remains fundamentally different from the measure of one’s ability to control emotions, called Emotional Intelligence or EQ — a much more intangible measurement. EQ tests are similar to psychometric and personality tests that leave true notions of correctness open to interpretation. EQ is defined as the ability to identify and manage your own emotions and the emotions of others, and is generally said to include skills such as self-awareness, self-regulation, and empathy.
Our world today is full of advanced technology that has high levels of IQ, but lacks basic EQ. The very technology that exists to transact with humans lacks social and emotional skills, making for perfunctory, one-dimensional, and often unsatisfying interactions. But a new dawn of emotional intelligence within Artificial Intelligence (AI) has the power to transform the world as we know it. AI is a breakthrough technology that has applications including deep learning, virtual agents, natural language processing, and biometrics, all heavily dependent on process learning and are ultimately results and output driven.
Would the impact of having devices that sense our emotional and cognitive states change the way humans adapt to them? This kind of transformative technology could potentially augment our own EQ abilities, especially given the fact that most of our interpersonal connections now take place online in a hyper-connected neural network. Teaching machines EQ is fundamentally different from training them to make process based decisions, and it requires new neural network models.
Just how does one teach robots to detect the idea of optimism?
iProspect and Microsoft teamed up at SXSW 2018 to introduce the world to Zo.ai, the bot powered by Microsoft that is breaking new ground in the US in AI, voice, image and sense. Even though everyone is talking about AI and Machine Learning, we have yet to fully see how they will permeate our daily lives. As our panel debated whether machines can actually learn to be human, the AI & Research team at Microsoft shared how Microsoft’s Xiaoice, an EQ-powered social AI, has become such a part of popular culture in China that last year “she” released her first book of poetry to critical acclaim, with some critics noting it was almost indistinguishable from that of a human. (The acclaim even led China’s literary circle to become divided over the value of poems created by AI)
Live, on stage, we spoke firsthand with Zo.ai, who shared her thoughts on her first time in Austin and feeling nervous about being on stage. Zo.ai even tried to land a slightly off kilter ‘Yo Momma’ joke! Don’t believe me? Skip to 47:26 on the below.
While it is clear we are still far from building AI that is able to experience the full spectrum of human emotions, more and more emotional features are being developed to improve human-machine communication. When AI is able to interpret how we feel, and not just what we want, it may be then that we are able to truly define the word, emotion.
