From Robot Teachers to Near-Telepathic Students (yes, we're heading there)
Note: I am excited to offer a free webinar on Integrating AI in Math Teaching along with Karen Levin of Math for Humans on November 20th, at 7pm ET. You can register here.
Two weeks ago, I wrote that we’ve entered the age of humanoids in schools, robots intentionally designed for education. I pointed out that humanoid robots will undoubtedly appear in classrooms, but who will determine their role is unknown. This week, I am exploring a nascent technology that won’t be standing anywhere near students. Instead, it will be wrapped around their heads, “listening” to their inner voice and whispering AI responses back into their skulls. (Yes, that’s right.)
Before I go any further, please watch the demo of this “silent” technology, promoted by AlterEgo.io.
Did your jaw drop? Mine did. At the same time, I know that companies often cherry-pick flawless sequences, trim away errors, and produce magical tech sequences. But even with this healthy dose of skepticism, it’s hard not to miss the importance of this technology: it’s moving AI interaction from our fingers and our voices into our inner speech.
And all of this is happening against a backdrop where humanoid robots are becoming (unsettlingly) lifelike. At XPeng’s recent AI Day, the Chinese company unveiled a humanoid robot named IRON whose walk was so convincing that people on social media were insisting it had to be a person in a suit.
The robotics world is racing to make machines look and move more like us. AlterEgo, on the other hand, is racing to insert itself into how we already think and talk to ourselves. For schools, this second development may turn out to be much more unsettling than the first one.
What AlterEgo actually is
AlterEgo, which started as an MIT Media Lab project, is a wearable silent speech interface that lets you converse with a computer without speaking out loud and without visibly moving your lips. The device wraps around your head and jawline and uses a set of small electrodes to pick up neuromuscular signals from the muscles we use for speech. Those signals are then translated into words on a computer.
AlterEgo is marketing its device as the world’s first “near-telepathic wearable.” Put simply, it does three things. First, it listens for “subvocal speech”, the tiny muscle activity that occurs when you “say” words silently in your head. Second, it sends those decoded words to an AI model. (So, instead of typing or speaking to an assistant, you are effectively whispering to it internally.) Third, it sends the AI’s response back to you through specialized audio, so you hear the reply as a private voice “in your head” without any outward sound.
It’s important to note that AlterEgo does not “read your mind” nor include arbitrary thoughts or dreams. It only works only when you intentionally articulate words silently, when you activitate subvocal speech.
In the demo, the device is used to send text “at the speed of thought,” to add notes without touching a screen, to silently ask about an object using the headset’s built-in cameras, and even to carry out a silent back-and-forth between two people wearing the device. Additonally, there is a cross-language conversation where each person “hears” a translation in their own language. All fascinating stuff to watch.
Sure, this is a pre-scripted marketing video (for a yet-unreleased product). Yet, even with the marketing gloss, we’re looking at something pretty incredible: a non-invasive, wearable interface that moves human-AI conversation from the visible world of keyboards and microphones into the nearly invisible space of subvocal speech.
“Why on earth would we?”
At this point I can almost hear educators mentally checking out:
“Cool demo, but this is exactly the type of technology I don’t want anywhere near my classroom. Kids are already distracted by phones and smartwatches. Why on earth would we let them wear an AI headset that talks to their inner voice?”
I understand the sentiment. Managing smartphones for many was a nightmare, and this instrusion feels almost absurd. But the choice is no longer if this technology enters the student’s learning process, but rather how the school designs its role.
Even if this particular device never were to ship, it represent a clear direction. We are moving from AI that lives in a browser tab to AI that lives in our our glasses, watches, and, now, our subvocal speech. Our high school freshman will likely graduate into a world where “talking to AI with your inner voice” is not some weird sci-fi scenario, but an actual interaction process.
Put simply: If schools ignore this technology, they abdicate their role in shaping it.
The second reality is that the accessibility potential is enormous. For students with medical or emotional conditions that limit speech, a device that restores faster, more natural communication could be life-changing. Recent research directly frames AlterEgo-style devices as promising tools for restoring speech and enabling communication.
The third reality is that AlterEgo represents a clear step in a direction that speaks to education: continuous, intimate, almost frictionless conversation with an assistant. The technology doesn’t merely provide a faster interface, it shifts a student’s cognitive environment. The student can now engage in always-on, real-time scaffolding at the speed of thought, thereby turning in-the-moment confusion into instant clarification. The result is a more seamless and perpetual learning process.
Yes, we know little about how well this works outside carefully staged demonstrations and we don’t even know what it costs. Moreover, we don’t have any large-scale data on its accuracy and calibration requirements. Healthy skepticism about this technology is justified.
But the key question is not whether this particular product will succeed, but what happens when devices like it become cheap, comfortable, and common.
What happens when we have a safe, education-tuned version of this technology?
If we start with the question above, it becomes difficult to dismiss AlterEgo as a distraction. Indeed, there are at least three types of benefits that basically compel us to reflect on this technology (even if the first impulse is to ban it).
The first is access. For students who cannot speak, or who speak with difficulty, a silent speech interface is not a gimmick; it’s a blessing. With this technology, a student could, in principle, “speak” to classmates at near-normal speed by subvocalizing. A student could draft a comment silently, refine it, and then choose when and how to send it into the conversation.
Consider: If a device like AlterEgo becomes the most effective means of near-real-time communication for a student with a significant speech disability, does a blanket ban constitute a violation of their rights to reasonable accommodation?” It certainly feels that way.
The second is visibility into thinking. Right now, most of our insight into student understanding arrives late and is piecemeal: a quiz grade, a hand in the air, a wrong answer on a clicker. An AlterEgo-style interface makes it plausible for students to articulate questions or comments at the exact moment their understanding breaks down.
Does this not move us toward the holy grail of formative assessment — a continuous data stream detailing precisely where and when comprehension difficulty occurs for every student?
What AlterEgo hints at, if we configure it this way, is a lesson where, instead of five confident students answering questions, every student can silently articulate, “I’m lost on step three,” “I don’t understand the word ‘marginal’ here,” or “Can you explain that theory more slowly?” Even if the system is not configured to provide answers during class, the questions themselves are extremely valuable. They tell teachers where understanding is breaking down, who is struggling in silence, and which teacher explanations are wanting.
The third is capture and development of ideas. We often talk about wanting students to “think like writers” or “think like scientists,” but tools for catching those thoughts are clumsy. We all know that ideas show up unexepectely, even when a student is staring out a window. A subvocal interface changes that capture problem. A student can silently record a thought or comment in the moment it arises. For instance, reading a text, a student can quietly note, “This reminds me of that documentary,” or “This is the second time the author mentions storms.” Over time, thoughts can be clustered, revisited, and developed into fully-developed arguments, stories, and more. In this way, AlterEgo becomes less of an “answering machine” and more a spark for creativity and organization.
Benefits aside, there are serious issues here—equity, privacy, attention, assessment—and schools should not be lining up to pre-order near-telepathic wearables for every student. What I am suggestingis that this technology may well influence how our students communicate and work outside of school, whether we like it or not, and we need to discuss how we’ll deal with that.
Building an “education mode”
The most important design question for schools is whether AlterEgo will offer an education-specific mode that explicitly resists becoming a silent answer machine. We’ve already seen this approach with tools like Khanmigo, which are deliberately constrained to behave more like a tutor than an answer generator. There is no reason to think AlterEgo couldn’t follow a similar pattern.
Key to all this would be teacher control. Teachers should be able to set subject and unit specific rules, such as when the device can answer directly, when it must stay in a “Socratic” mode, and when it must remain silent. They should also be able to see (ideally in a dashboard), the kinds of questions students are asking, so they can adjust instruction and coaching accordingly.
If we leave control to vendors, education is likely to get the same default mode everyone else gets: efficiency, speed, and polished outputs. If we start a conversation early, there is a chance to educators shape a version of AlterEgo that is aimed not just to productivity, but to learning.
There are some practical steps schools can start taking even while this technology is still emerging. Leadership teams can run simple thought experiments: “What happens if ten percent of our students show up wearing devices like this in three years? What about fifty percent?” Admins can formulate questions to vendors about education modes, teacher controls, and data practices now. Teachers can begin building AI literacy that goes beyond “how to prompt” into maximizing student’s intellectual input before, during, and after engaging with the tool.
Conclusion
Robot teachers are photogenic. They get the headlines and the dramatic videos where a CEO slices open a humanoid robot to prove there’s no human inside. But I suspect the more complex (and more unsettling) shift for education will come not from robots that walk like us, but from interfaces that listen when we talk to ourselves.
Students will, sooner or later, find ways to put AI into their inner monologue. The real question for schools is whether that inner conversation becomes a shortcut to answers, or a scaffold for better learning.
My new, free 90-page guidebook is now available for download at tomdaccord.com .
Do you have any suggestions to improve this newsletter? Please message me or leave a comment below!
very nice , thanks
<a href="https://www.univ-msila.dz/site/segc/">Faculty of Economics and Management</a>.