More on News
As artificial intelligence (AI) becomes increasingly prevalent in various aspects of our lives, researchers are now focusing on developing social skills in AI systems. Intending to create more human-like interactions between machines and humans, experts believe that equipping AI with social skills will be crucial for its successful integration into society. From recognizing emotions to understanding social cues, the ability to navigate social situations will enable AI to communicate more effectively and enhance its usefulness.
Chinese researchers claim that although artificial intelligence (AI) is smart, it is limited by a lack of social abilities.
In a study published in the CAAI Artificial Intelligence Research, the researchers outlined the siloed subfields that make up Artificial Social Intelligence (ASI), including social perception, theory of mind (the recognition that people think from their perspectives), and social interaction.
“Artificial intelligence has changed our society and our daily life,” first author Lifeng Fan, from Beijing Institute for General Artificial Intelligence (BIGAI), said.
“What is the next important challenge for AI in the future? We argue that Artificial Social Intelligence (ASI) is the next big frontier,” Fan said.
Fan asserted that the discipline would be better positioned to grow by employing cognitive science and computational modeling to pinpoint the gap between AI systems and human social intelligence, as well as present problems and future directions.
Fan added ASI includes the ability to interpret subtle social cues like yawning or eye-rolling, understand the mental states of other entities such as beliefs and intentions, and work together towards a shared goal.
“ASI is distinct and challenging compared to our physical understanding of the work; it is highly context-dependent,” Fan said.
“Here, context could be as large as culture and common sense or as little as two friends’ shared experience. This unique challenge prohibits standard algorithms from tackling ASI problems in real-world environments, which are frequently complex, ambiguous, dynamic, stochastic, partially observable, and multi-agent.”
According to Fan, the most effective method is to adopt a comprehensive approach that imitates how people interact with each other and their environment. It is necessary to create an environment that allows for open-ended and interactive communication, as well as to take into account the incorporation of improved human-like biases into ASI models.
“To accelerate the future progress of ASI, we recommend taking a more holistic approach just as humans do, to utilize different learning methods such as lifelong learning, multi-task learning, one-/few-shot learning, meta-learning, etc.,” Fan said.
“We must define new problems, create new environments and datasets, set up new evaluation protocols, and build new computational models. The ultimate goal is to equip AI with high-level ASI and lift human well-being with the help of Artificial Social Intelligence.”