Introducing human emotions to machines..
As artificial intelligence begins to seep into our social and consumer lives, we should "in theory" eliminate all the flaws that people make.
But the reality is very different, from Facebook algorithms that instigate aversion to attract more viewers, to facial recognition apps that don't recognize people of color and artificial intelligence. Often they do not help solve human problems.
But former Google researcher Alan Quinn, who has a psychology background and founded a research firm called Hume AI, says it could help make artificial intelligence's chaotic tasks more empathetic and human. By learning hundreds of thousands of facial and voice expressions from around the world, the Hume platform's AI can interact with how users actually feel and more precisely meet their emotional needs, Quinn says.
He hoped that the platform would eventually be integrated into consumer applications such as video and digital assistants.
A beta version of the platform will be released in March next year, more details to come later. "I know this fight is going to be long..."
Quinn said in an interview, but we need to start improving on that.
Not the first
The 31-year-old entrepreneur is not the first technologist to inject human emotions into the digital space. There is a movement on “ethical artificial intelligence” that aims to incorporate fairness and fairness into algorithms, which includes many of its members, such as the Mark Rothenberg Center for Artificial Intelligence Policy and Digital Policy, and new distributed research. on artificial intelligence. Anti-bias lab with AI ethicist Timen Gebro who worked with Google in this area.
Tement gipro was hired and fired by Google to openly criticize unethical AI.
There are many academic experts who take a strong public stance on the development of artificial intelligence to eliminate social prejudice.
However, what Quinn brings to the field is the high-level psychological research carried out with these moral goals. His previous work includes the study of emotional responses across cultures (eg, studies of similar responses to sad songs in the United States and China) and work on sound effects of various nuances.
Empathy Machine Learning Assistants
Hume's initiative forms an ethics committee with several emotional and ethical AI scientists, including Empathy Lab founder Daniel Kretek Cobb, "computation justice" expert Kartik Dinkar, and Quinn mentor and UCLA professor Angeles Dasher Keltner. I did. I worked on emotions with Pixar when creating the Inside Out comics.
Quinn said he raised 5 5 million from the studio "age ventures" with another subsequent tour. The funds will be directed to investigate how AI is formulated not only to process very quickly and see invisible patterns, but also to develop its understanding of humans, an approach Quinn has dubbed "empathic AI".
Features and cons
The idea of having more emotions may seem to contradict mainstream ideas about artificial intelligence, whose core power is often seen as making decisions without taking human feeling into account.
But many in the emotional computing community say that AI's inability to read people is what makes it dangerous, and makes it important that AI sees the human aspect in the humans it serves.
Of course, there's no guarantee that if AI can measure sentiment, it won't take advantage of it, especially if big tech companies seek to maximize their profits. Another challenge in developing AI emotions is how to avoid exploiting the emotions of human programmers, which can be biased.
Coyne's colleagues believe that the Hume model can avoid bias. "The Hume model is informative but unbiased," said Arjun Nagendran, co-founder of VR education company morchen. "Initiatives like Queen can play a role in creating racially biased AI," said Ben Schneiderman,
University of Maryland professor and artificial intelligence expert Ben Schneiderman. You can, but you can't: by 2030, AI will be used primarily for social good.
Quinn acknowledged the risks associated with providing more emotional data to fast-growing AI.
Comments
Post a Comment