Expert Warns of Risks as Humanoid Robots Approach Reality

Autonomous humanoid robots are becoming more feasible, but a leading robotics expert from the UK cautions that significant risks accompany this advancement. Dr. Carl Strathearn, a Lecturer in Computer Science at Edinburgh Napier University and an adviser on AI and robotics to the UK Government’s Office for Science, emphasizes the need for strict regulations to mitigate potential misuse as these technologies develop.

Dr. Strathearn will share his insights at the upcoming New Scientist Live event, taking place from October 18-20 at London’s ExCel centre. He highlights a gap between the exciting demonstrations of humanoid robots that can perform tasks like pouring drinks and folding clothes, and the reality of creating reliable assistants for everyday life.

“The biggest problem is the lack of real-world data and the technological means of gathering it in large enough quantities to train our robots effectively,” Dr. Strathearn explained. Current methods primarily depend on virtual simulations and reinforcement learning, as well as sourcing videos from platforms like YouTube. This results in robots that may excel in controlled environments but struggle to navigate the unpredictability of real-world settings.

One potential solution for this data shortfall is crowdsourcing. Dr. Strathearn suggests using devices like Meta’s Ray-Ban smart glasses to gather extensive real-world data. However, he acknowledges that mobilizing thousands or even millions of individuals to collect and share data presents an ambitious and ethically complex challenge.

The expert expresses a different concern regarding the dangers of humanoid robots. “If you are talking Terminator, the answer is no. We are and always have been more of a danger to ourselves than anything else,” he stated. The real risk, according to Dr. Strathearn, lies in the potential for humans to misuse humanoid robots, often with inadequate training. He is currently spearheading a petition to the UK Parliament advocating for regulations governing the use of humanoid robots in public spaces, especially after several alarming near misses involving humans and robots.

Dr. Strathearn elaborates on the dangers of human control over these machines, stating, “Humans control them using handheld devices, which makes them very dangerous and unreliable.” He emphasizes the increasing frequency of serious incidents, attributing them not to artificial intelligence, but rather to human error.

The issue of perception also poses challenges. Robots that closely resemble humans risk entering the “uncanny valley,” where their lifelike qualities may evoke discomfort. Yet, in contexts such as dementia care, a familiar human-like appearance could be beneficial. “People have different thresholds of perception when it comes to creepiness. That’s why we have different types of robots — some very lifelike, some with just minimal facial features,” he explained.

During his PhD, Dr. Strathearn developed the “Multimodal Turing Test” to analyze whether communication through lifelike robots made AI seem more human. Research conducted by Japanese scientists later confirmed that individuals were more inclined to perceive AI as human when interacting through realistic robots. He believes that fostering acceptance will require a gradual and careful approach, particularly through education about robotics and AI in schools.

Despite the challenges and warnings, the robotics industry continues to move forward rapidly. Dr. Strathearn notes, “The hype is a major issue. We are far from humanoid robots that are good enough to do everyday tasks effectively, but that doesn’t stop major companies wanting to mass-produce them.” He points out that there is already a significant skills shortage in the field of robotics, as universities often separate disciplines such as computer science, engineering, and design. He argues that the future of robotics demands interdisciplinary training to create a sustainable workforce.

Interestingly, Dr. Strathearn identifies space exploration as a domain where humanoid robots may prove useful much sooner. He stated, “We could use telemetric or AI-controlled humanoids to work in space for longer periods than humans, advancing us further into the unknown.” Future humanoids might assist in terraforming planets or exploring challenging terrains that current robotic rovers cannot reach.

While the potential for robots to contribute to space colonization is promising, Dr. Strathearn’s message is clear: the foremost challenge on Earth is ensuring that humanoid robots are safe, reliable, and properly regulated before they are integrated into society. “Robots might terraform Mars one day. But on Earth, only strict regulation will keep us safe,” he concluded.