Key Takeaways:
- The widespread adoption of AI and robotics will depend on their ability to integrate seamlessly into human environments. As highlighted by Meta’s continued investment in AR/VR (selling 2 million Ray-Bans since October 2023) and Amazon’s humanoid robot training, the emphasis is shifting from “what can tech do?” to “how should it present itself?”.
- While drones offer efficiency, companies like Amazon are investing in humanoid delivery robots (e.g., Digit from Agility Robotics), not for functional superiority, but for human acceptability. This suggests that the design of AI in human spaces prioritises relatability and trust over pure optimisation.
- A “fork in the road” exists for technology’s relationship with humans: one path leads to discreet, physiology-integrated tech (rings, patches), while the other embraces human-mimicking AI companions (robots with posture, vocal assistants). The core design challenge is evolving from hardware capability to managing intimacy and human tolerance.
Wearable tech has spent the last decade promising to change our lives, and occasionally delivering – and portable tech (i.e. our phones and smart watches / rings) has certainly became a big category in its own right. But what do people actually want from a wearable – AI or otherwise? And how does that choice reflect how we feel about robotics and the way technology appears in our day-to-day lives in general?
Is the dream to have devices that disappear into the background, offering convenience without intrusion? Or do users want expressive, even communicative, machines that feel like companions rather than tools? Apple reinvigorated the wrist as a site of technological engagement. Meta covets ocular real estate. Johnny Ive is sketching an AI something. Google keeps circling back to glasses. While a plethora of companies like BioBeats and Lingo think the future is a different breed of device, wearables that are not worn on the body but fused with it, embedded in the skin itself, designed to monitor health with minimal footprint.
Plenty of companies have visions for where wearables are going. Far fewer seem to agree on why people would actually want them there. Look to the Rabbit R1 and Humane AI Pin for examples of wearables that promised a lot, but moved the needle comparatively little. While in these cases the failure was down to pure capabilities (it’s pretty clear that people want AI devices that can actually do things, especially if they come with a monthly subscription attached), the near-term horizon for how technology appears on our bodies and around us, in the world, is going to be determined by a much more complicated set of variables.
As a case in point: last week Meta’s CTO declared that 2025 would be a “pivotal year” for AR and VR. And maybe it will! The devices are lighter. The software has become smoother, but the same friction remains: do people actually want to wear tech all day? Do they want it on their faces, against their skin, in their ears? Or do they just want it to be nearby, waiting for a cue? Irrespective of how cost-prohibitive crystal lenses are for the creation of AR glasses, are glasses actually the form factor the world is going to settle on?
Health monitoring devices offer the clearest use case to wearable burden balance. They are clinically orientated, materially unobtrusive, and socially silent. A lot of the platforms in the space aren’t chasing mass adoption and multi-million user counts either; they’re solving real, clinical problems through quiet design. These wearables aren’t wearables in the cultural sense. They’re body infrastructure.
But the cultural wearables, the ones trying to redefine lifestyle, fashion and social interaction, still haven’t landed. Not properly, although Meta’s Ray-Bans selling 2 million units since October 2023 shows adoption is growing. The issue remains that while AI gadgets are impressive, they’re still socially awkward.
Against this backdrop, a parallel narrative is unfolding that flips the paradigm entirely. Here, the technological object no longer seeks to be worn by the human, but rather to wear the human form itself. Specifically: humanoid delivery bots and other avenues where technology has the potential to appear in our lives in human-compatible forms, within human-defined structures.
Amazon is reportedly nearing completion of a dedicated training facility for humanoid robots. According to insiders cited in the report, published this week, the facility is a purpose built space designed to test how robots like Digit (from Agility Robotics) or models from Unitree might exit electric delivery vans and walk parcels to a customer’s door. Why take this approach rather than drones or some other form factor? Because cities are built for human motion. Because homes have stairs and doorknobs. But also because a robot that looks like a person is perhaps less alarming than a robot that looks like a solution. Design here isn’t about function. It’s about acceptability.
Walmart, by contrast, is still investing in drone delivery (though Amazon is still exploring that route too). This past week, they expanded the service to five Southern American cities. If you live in Houston or Tampa, your next purchase could descend from the sky like a gift from the cloud gods. It’s both fast and efficient. But here’s the tension: if drones are faster and cheaper, why bother with humanoid bots at all? Outside of scale (which you might be able to make work, economically, with more drones) why do we keep insisting the future needs faces, voices, and human-coded means of ambulation?
The simple answer is that tech doesn’t just need to function. Performance gets it in the door but being liked might be what keeps it there. A flying robot might be optimal from a logistical point of view, but a humanoid robot might be less frictional to invite inside your home – especially if that home already contains one or more AI devices.
All of this is to say that it feels like we’re at a fork in the road, not just for wearables, but of wearers. One track leads to discreet, frictionless tech that becomes part of our physiology, be that rings, watches or a simple patch – all things that communicate quietly, often in the background. The other leads to tech that mirrors us – AI assistants with charm, robots with posture, glasses that watch as much as they display. Things that talk back. One wants to be discreet, while the other wants to be relatable.
It isn’t just an aesthetic split, it’s psychological. A product that wears us back by mimicking our form, voice, and movement, demands something different than a product that tries not to be seen at all. It isn’t just performing a function, it’s performing for us.
The humanoid delivery robot is the perfect example. There is functionally no reason that a delivery robot needs to look like a human in shape, but it does, because walking up to your front door with a distinctly human gait feels more trustworthy than something rolling up like a box on wheels.
So what do people actually want from wearable tech?
Maybe the better question is what kind of relationship they want with technology at all. Do they want tools that behave like extensions of the body? Or AI companions that behave like extensions of the self? A health patch doesn’t need a personality, but a face on your wrist might. A walking robot might not need a voice, but people might trust it more if it sounds like a valet.
As AI, robots, and wearables converge, the real design challenge isn’t hardware. It’s intimacy. Tech that needs to move through human space without unsettling humans.
That’s the shared theme across Meta’s new ambitions, Amazon’s humanoid experiments, and the rise of discreet clinical wearables. Whether it’s something we wear, or something that wears us, the future of technology will likely need to be tolerated before it can be adopted at scale. We’re entering an uncanny age where the question has moved from what can tech do for us, to howit should present itself while doing it.