Engineers at Heriot-Watt University and Alana AI have unveiled a groundbreaking robot receptionist powered by OpenAI’s GPT-3.5. The project, housed at the UK National Robotarium, aims to explore the cutting-edge capabilities of large language models (LLMs) in the field of human-machine interaction.
An Unprecedented Experiment
The robot is not merely a text-based conversational agent; it features a humanoid face complete with moving eyebrows and mouth. This distinct setup takes the interactivity to another level by incorporating non-verbal cues, making it the first of its kind to marry a subject-specific knowledge base with verbal and non-verbal communication abilities. To minimise the risk of disseminating incorrect information, engineers have developed a dedicated database by scrapping the robotarium’s website. The robot taps into this database to answer visitor queries more accurately.
The Inner Workings
When engaged in conversation, the robot initially converts the received speech into text, which explains the slight pause observed during interactions. This latency could be attributed to the time needed for speech-to-text conversion and the subsequent natural language understanding processes. Once decoded, the robot utilises GPT-3.5 to generate an appropriate response, which is then vocalised. Simultaneously, the robot mimics lifelike facial movements and expressions, providing a more realistic interaction.
A Step Forward, But Not Quite There Yet
Despite its groundbreaking nature, the execution isn’t flawless. The robot’s conversational tone might tick the boxes for politeness and contextual understanding, but the delivery remains unambiguously robotic, akin to conventional text-to-speech systems found in smartphones. However, it’s important to note that these are early days for AI-powered robotic receptionists, and we can anticipate continual improvements.
The Future of Robotic Human Interaction
This project serves as a tantalising glimpse into the rapidly evolving domain of AI and robotics. While the uncanny appearance of the robot may evoke mixed feelings, its existence heralds an era where the utility of LLMs like GPT-3.5 extends beyond text-based applications into the realms of complex, humanlike interactions. As technology advances, we can only expect these robotic interfaces to become more nuanced, and potentially indistinguishable from human receptionists. For now, the future looks both exciting and, perhaps for some, a tad unsettling.
The robot receptionist at the UK National Robotarium represents a bold step into uncharted territory. While it is not perfect, it’s an impressive application of emerging technologies, and a hint of what’s yet to come in the world of AI and human-machine relations.
.thumbnailWrapper
width:6.62rem !important;
.alsoReadTitleImage
min-width: 81px !important;
min-height: 81px !important;
.alsoReadMainTitleText
font-size: 14px !important;
line-height: 20px !important;
.alsoReadHeadText
font-size: 24px !important;
line-height: 20px !important;
