Developing Realistic Emotions in Humanoid Robots
Chinese scientists have made a significant breakthrough in the world of humanoid robots. They have developed a way to make humanoid robots express emotions more naturally and accurately. This new method was a collaborative effort between researchers from Hohai University and Changzhou University. Their creation involves a two-stage process that enables robots to perform more complex facial expressions. This innovation aims to improve how humans interact with and relate to these robots.
The new AI system can generate detailed examples of facial expressions. These examples are then learned by a specially designed robot with multiple degrees of freedom for facial movements. Consequently, humanoid robots can perform specific facial expressions when instructed, making interactions more lifelike. This method was showcased at the China Association of Science and Technology annual meeting and published in the IEEE Transactions on Robotics journal.
Humanoid robots typically have fewer motors in their faces compared to the numerous muscles in the human face. This limitation often prevents robots from displaying authentic expressions. To address this challenge, Chinese researchers, along with scientists from the University of Manchester and the University of Leicester, employed Action Units (AUs). AUs define individual muscle movements for facial expressions. These units help translate AI-generated expressions into motor commands for the robot’s face. As a result, the system makes use of physical constraints to refine and make expressions more realistic.
Additionally, the new two-stage process divides nine motors into 17 AUs. This enables richer expressions and smoother transitions through coordinated movements. Using this technique, robots can physically reproduce expressions generated in the first stage, preparing them for various applications such as nursing homes, kindergartens, and special education schools. According to Liu Xiaofeng, a professor at Hohai University, humanoid robots will not only assist humans in completing tasks but also bring more emotional value (source).
The Commercial Implications of AI and Generative AI Challenges
In contrast, while AI advancements in emotional expression are promising, not all sectors are experiencing seamless transitions. New research from Gartner suggests that at least 30% of generative AI businesses will fail by 2025. These findings were presented at Gartner’s Data & Analytics Summit in Sydney. The statistics highlight that early adopters of generative AI are grappling with escalating costs.
For example, deploying custom generative AI models can range from $5 million to $20 million. Such substantial financial requirements are posing a significant challenge for businesses. Designing a specific model, like fine-tuning a Llama model on industry data, can cost a business up to $6 million upfront with additional recurring costs. Setting up even simple features like document search through retrieval augmented generation (RAG) costs upwards of $750,000.
Gartner’s research indicates that businesses are struggling to justify investments due to these high costs. Moreover, executives are becoming impatient for quick returns on generative AI investments. Rita Sallam, Gartner’s distinguished vice president analyst, points out that although benefits might not be immediately evident, the potential remains high. Once business outcomes meet expectations, there’s potential to expand investments further, scaling usage across broader user bases or additional divisions (source).
Conclusion: The Future of Emotion in AI and Generative AI
In conclusion, advancements in AI have recently opened up new possibilities for humanoid robots. The ability to express realistic emotions significantly enhances interaction and application in various fields. Meanwhile, the commercial aspect of AI, particularly generative AI, faces its set of challenges. High costs and the need for substantial financial investment can be daunting. Nonetheless, with careful analysis and strategic investment, the potential benefits can be immense. Ultimately, both the emotional development in humanoid robots and the practical deployment of generative AI show promise for the future.
