Online: Trending Now

We are most accustomed to writing prompts and challenging Artificial Intelligence (AI) to give us answers, insights, sources, videoclips, and creativity. Emerging now is the walking, talking AI that engages us on our level.

We have front-row seats to the maturing of AI in higher education. There has been audio for a while, but now we are preparing for embodied AI (EAI). The race is on among multiple competitors to give AI a humanoid body that enables a wide variety of teaching capabilities as well as unique opportunities to advance research and engage the public. Images and early specifications of an assortment of humanoid robots under development can be viewed in this April 8, 2025 report in the Robozaps site.

Pricing and delivery dates vary, but in general the cost of an autonomous robot may be as low as a few thousand dollars or as high as $50,000 and even hundreds of thousands of dollars for specialized research models.  The earliest mass-produced embodied AI may begin coming out in the fourth quarter of this year and many more models in 2026. Earlier this year, Dean Fankhouser wrote in Robozaps “Humanoid robots in education are transforming the learning experience by acting as teaching assistants, engaging students, and providing personalized learning support.”

Leading the way in humanoid AI robots is a rapidly growing need for models designed for manufacturing and warehouse applications. As Edward Ludlow writes in Humanoid Robot Development Seen as Vital to the Future of Manufacturing:

Talking with me on Bloomberg Tech last week, Tesla Inc. investor Nancy Tengler put it plainly in the context of robots: “If you believe the BLS numbers, we had 450,000 manufacturing jobs that have gone unfilled. Robots are the solution…. Just as AI is writing the AI code, I think the robots will build the robots and we’ll see manufacturing plants use more and more robotics.” Activity from dozens of companies and investors willing to fund them signals that the effort to build humanoid robots for the workplace is pushing us forward irrespective of what human workers will do. That tension sets the stage for Tesla’s Optimus push in particular, as a front-runner to scale deployment of such bots in the real world. The company boldly claims it will deploy thousands of Optimus humanoids in its own facilities over the next year and produce millions of them in the next five.

At the Tesla Website, Optimus is described as “a general purpose, bi-pedal, autonomous humanoid robot capable of performing unsafe, repetitive or boring tasks. Achieving that end goal requires building the software stacks that enable balance, navigation, perception and interaction with the physical world.” The list of applications is much longer than the three listed at the site.

In higher education the list of specific applications is long. Realizing that humanoid EAI robots will serve not just one job function, but multiple job duties as needs arise, just like humans, the applications are seemingly endless. Here are a few:

• Embodied AI Teaching Assistants can directly engage students in one-to-one tutoring and advising sessions. With all of the capabilities of text and avatar-based AI tutors, embodied assistants can observe and respond appropriately to student frustration and uncover points of misunderstanding in ways that our current tools cannot. The vision capability of EAI enables it to perceive human behaviors and responses in ways that are not possible using computer-based tools.
• The study, “Designing embodied generative artificial intelligence in mixed reality for active learning in higher education” led by Andy Nguyen and published April 29, 2025 in Innovations and Education Teaching International observed:

Our analysis revealed that the embodied GenAI agent was perceived by participants as authentic and human-like, which positively influenced their engagement. Although participants were not very familiar with learning using this technology, they responded positively to the embodied agent. The findings suggest that despite limited prior experience, learners were receptive to the agent’s presence and interaction, highlighting its potential to support the learning process without adding an extra cognitive burden for technology familiarisation (Mayer & Fiorella, Citation2021)…. Notably, participants who perceived the GenAI as more anthropomorphic and engaging reported higher levels of immersion and willingness to engage actively with the content, indicating that these qualities may be instrumental in fostering sustained student interaction and learning in AI-driven environments.

• In fields such as Engineering and Chemistry, EAI assistants on campus can conduct set-ups and demonstrations of experiments using dangerous substances or other instances where it is unsafe to have the faculty member and students in the lab. The students can observe, question and further interact with the robotic assistant from an adjacent or distant location while the actual components are utilized, rather than observing recorded simulations.
• The addition of embodied AI can significantly improve the physical campus facilities and services. Reception and Information desks can be staffed 24 x 7 with a physical assistant that can ask and answer questions. These EAI robots can be imbued with super-vision, extra-human strength, multi-lingual capability, and programmed to be both supportive as well as to be a superior first responder to crises. The multi-tasking capability of these devices is not to be underestimated. They can serve the roles of campus security, receptionist, first-line supporter of those suffering from mental or physical distress, and more.
• Libraries have multiple potential uses for such workers. At the reference desk, embodied AI units can serve as encyclopedic knowledge sources, while at the same time collect and deliver books and media to the patron desiring to dig deeper. After hours tasks can be accomplished by the embodied AI agent who can multitask to most efficiently re-shelve books; respond to late night emails, texts and calls; and any other tasks that arise. It should be noted that these tireless workers require no time off, no health insurance, and none of the other benefits afforded traditional workers. Yet, they grow in intelligence each day and refine their physical skills at tasks through reinforced learning.
• Much like library assistants, embodied AI robots can operate 24/7 in research labs, collecting data, conducting experiments, and even performing follow-up interviews, significantly increasing productivity and potentially gathering more comprehensive data such as nonverbal cues than computer-based bots.

The list is endless. The economies are many, considering the 24 x 7 service, with brief re-charging interludes, that EAI can perform. Even in those re-charging periods, it would seem that the units could perform computer-based functions in text, video and audio. The race is on among manufacturers to fabricate market-ready models

Imagine that it may be as early as next year that you will be working, not just fingertip-to-keyboard, but rather shoulder-to-shoulder with EAI at your university. Are preparations underway at your institution to best utilize these capabilities? Who is leading the initiative to determine the most effective and efficient applications on your campus? [Generative AI tools were used in researching this column.]

 

This article was originally published in Inside Higher Ed.