Automated, AI-powered software and applications already help learners around the world gain knowledge without the help of another human. In fewer cases, developers have deployed similar technology into hardware that can move, watch, listen, record, and interact. Humanoid robot teachers have been developed and deployed in classrooms on limited bases around the world. For some, this is cause for excitement. Others are more cautious, or even deeply concerned. Two researchers from Durham University, Professors Douglas Newton and Lynn Newton, recently published an article describing the current uses and concerns of humanoid robot teachers, along with a proposed code of practice.
The Newtons describe robot as a “machine that senses, thinks, and acts and, when this is without external control, it is described as autonomous.” Furthermore, “Humanoid robots are intended to look and behave somewhat like people and they usually have some means of communicating with them.”
Humanoid robot teachers have already been put to work with the intention of addressing a wide range of issues.
Humanoid Robot Teachers Teaching Second Languages
Language learning is often rote and requires more repetition and practice than the average human teacher can stand. But robots can repeat almost ad infinitum. Robots have already been programmed to carry on conversations with second language learners, play games, and conduct lessons.
Some researchers have found a secondary benefit here. Most language teachers do not step outside the teacher-student dynamic when helping learners practice. Some believe, however, that robots can flow more seamlessly between instructor and peer.
Autism Spectrum Disorder and Severe Anxiety
Some learners find the classroom to be such a stressful, alien environment, that learning is nigh impossible. Specifically learners with more severe cases of autism spectrum disorder (ASD) and social anxiety experience this.
But more and more researchers have found humanoid robot teachers have the ability to act as a kind of social intermediary for these learners. They can help introduce them to a social environment that avoids the ‘baptism by fire’ alternative.
Learning by Teaching
Finally, humanoid robot teachers can also serve as captive humanoid robot learners. The pedagogy of ‘learning by teaching,’ where a learner is asked to master a concept themselves and then deliver that knowledge to others, has been well-established.
While social noise and classroom interference can detract from the power of this technique, robots are mostly immune to these distractions and can serve as an especially impartial audience.
The Risks: Data Security, and Algorithmic Decision-Making
Outside of familiar science fiction horror scenarios, humanoid robot teachers also present numerous risks to educational systems and individual learners. Perhaps the most evident one is that of data security and privacy.
To personalize learning and teach a learner without a human teacher, robots need to collect and store a huge amount of non-anonymized data. While we already trust third parties like banks and governments to store more sensitive data than that generated in classrooms, no entity has proven itself to be be immune from data breaches. What’s more, cyber attacks targeting education systems are on the rise.
Algorithmic Bias and Digital Redlining
Autonomous humanoid robot teachers are only as good as their design. In our limited use of AI to date, we have already witnessed numerous instances in which the bias of a bot’s designers or their training processes translate negatively into the bot’s decision-making. Many education stakeholders, such as Macomb Community College Professor Chris Gilliard, continue to investigate this issue.
Again, humanoid robot teachers are not widely used. But the Professors Newton condense another concern with their use that is more speculative. Humans, especially young humans, are highly imitative. Imitation is a primary learning and development technique. Some worry that if kids start spending more and more time around robots, they’re going to start acting more and more like robots.
Some researchers have already begun to investigate this possibility, and they tend to draw comparisons between how children understand robots with how they understand animals. But this work remains largely preliminary.
For these reasons, Newton and Newton suggest a 10-point code of practice surrounding using humanoid robot teachers in the classroom. This can be found detailed in their article, which is linked above and below.
They conclude, writing, “Teacher identity, or what it means to be a teacher, is an evolving complex collection of personal roles, behavioral norms, and social and cultural expectations.”
Read the full paper here.
Featured Image: Franck V., Unsplash.