My name is Larry Liu.
I am enamored by the magic that is human sentimentality.
I am a graduate from Stanford University where I studied Psychology and Computer Science, which basically means that I have two sheets of paper, one that licenses me to get unreasonably excited about science and technology breakthroughs and another that licenses me to get unreasonably excited about human sentimentality and emotionality. This is great because I get unreasonably excited about pretty much everything: robots and romantic comedies, corporate social responsibility and competitive Tetris, puns and penguins and poor alliteration, and Oxford commas and organizational behavior. I love all sorts of music, but especially acoustic harmonies or country music.
Empathy absolutely fascinates me. In particular, I want to explore how empathy can be automated. Even though empathy is a fundamentally interpersonal thing, I believe it's also a fundamentally quantitative thing in that someone who is good at empathy can keep more things in mind about the individual they're trying to empathize with than someone who is poor at empathy -- how they were raised, their ethnic heritage, and traumatic life experiences, just to list a few. With technology improving at such a rapid pace today, we are able to analyze data in historically unparalleled quantities – specifically, massive quantities that allow us to start rigorously examining more intangible, abstract concepts like emotionality, value, fulfillment, and sentimentality in ways we've never been able to before.
To illustrate this point more concretely, consider this hypothetical example. Imagine someone – let's call him Nick – has a really high tolerance for hot sauce or other spicy foods. I also find that Nick is extremely resilient to setbacks or obstacles in his life. Intuitively, the hot sauce might not seem at all to shed light on what Nick's personality is like. However, it would not be a wild stretch of the imagination to believe that people who can handle spicy foods – suggesting that they've had the strength of will to repeatedly subject themselves to painfully spicy cuisines to the point that they've built that tolerance – also have the mental fortitude to deal with setbacks in their life. Knowing this correlation, even without being able to infer causation, allows us to better empathize with Nick and other enthusiasts of spicy foods.
While that example might seem absurd and trite, it still captures what I believe computing will be able to achieve in the future. Perhaps such an "empathy machine" will never be as accurate as the best humans at empathy, but even a moderately good empathy machine could vastly improve the quality of life of many people in so many ways. For instance, you could increase the reach of therapy programs for emotional disorders to rural and underprivileged demographics through online, self-report diagnosis. You could help people find lifelong romantic partners, or at least narrow the canonical sea of fish to a mere pond. You could better improve corporate recruiting procedures by identifying candidates who would better fit your company culture. You could reduce tensions or conflict between ethnic or religious groups. Call me idealistic, but in my eyes improved empathy, both human and autonomous alike, would bring us one step closer to world peace.