🌐RTN: Humane's Pin & the 'goldilocks zone' of new hardware

+ RealTech Conference speakers finalised

Humane’s new AI Pin - computing revolution or consumer hardware flop?

Humane just released the Pin for $699. It’s an ambitious project by ex-Apple design engineers to produce a new personal computing form factor. It’s somewhere between a phone, wearable and an AI voice assistant.

The device works by pinning it to your clothing (video: 0:40), it can project video onto your hand to display information (2:51), is voice activated and can ambiently record video (6:57). The general idea is that it’s seamless, voice and gesture controlled and can provide an AI overlay to your interactions.

Humane’s Pin is part of a broader hardware trend of “ambient compute”, the integration of compute with our everyday. The biggest success of this trend, has arguably been Airpods and the ultimate extrapolation of this is something akin to brain-computer interfaces (eg Elon’s Neuralink).

Does new hardware have a ‘goldilocks zone’

All of this, got me thinking about hardware form factors. There’s a big prize if you succeed but it is really hard to build a new consumer hardware device that finds adoption. There’s a long list of failures from Palm Pilot, Google Glass, MagicLeap to Amazon’s Alexa.

With new form factors, you have to combine new technological utility, with device familiarity. Raymond Loewy, the father of industrial design, called this the MAYA principle:

For example, Google Glass, whilst there’s familiarity with glasses, there was no real marginal utility in strapping those things to your face. If you’re going to look stupid, you better make sure there’s a good reason to do so.

The iPhone’s initial breakthrough was two-fold; the ability to type on a screen-based keyboard (novel yet familiar) and a combination of web browser, phone and mp3 player (utility).

Steve Jobs presented his own version of ‘MAYA’ at the 2007 iPhone launch

In industrial robotics, we’ve seen a recent surge in humanoid robot development. A form factor that is both intuitive (it’s human-shaped duh) and might deliver high utility given its potential for generalisable applications in a world largely built for humans.

Humane’s AI pin, in my view, falls short on both familiarity and utility. The use of a pin is non-intuitive and complicated by taking clothing on/off. There are questions around privacy, how can you use the device without it displaying or blurting out your communication? The utility of voice assistants, as we saw with Amazon Alexa, is low. Voice computing has not been limited by a lack of availability (Siri?) but by issues of app discovery without a GUI and usefulness. Voice is a low information medium and largely restricts the user to checking weather, messages and playing music. Some other people, like Om Malik, are more bullish on Humane’s device.

---

RealTech Conference Update

We are nearly sold out for the event on November 22nd in London. We have added two awesome speakers to the line-up, more on them below.

If you’d like to join, please apply using this link.

Anmol Manohar, CEO at GreenJets

Anmol (left) with the IPM 5 engine

GreenJets is a UK-based startup who are pioneering quiet(er) electric propulsion for aircraft.

Aircraft noise is a big problem, it limits use of aircraft near densely populated areas, with planes regularly taking diversions to minimise noise pollution. This has been a limited factor not only to traditional passenger aircraft but to newer drones and eVTOLs.

GreenJets has built an engine which delivers an order of magnitude reduction in noise. This could open up the use of eVTOLs in city centres and the use of stealth drones within the battlefield.

Janis Vavere, CEO of Trace Space

Trace Space are bringing modern software design and AI to hardware requirements, in aerospace, automotive and defence.

In the typical V-model of complex product builds, you need to specify and break down what you want to build. You start by formally specifying your design (“How fast do I want the car to be able to drive?”). Then you break these needs down into requirements for subsystems and parts (“How much torque does the motor need to generate for that?”) Existing requirements management tools like Jama are outdated, Trace Space is the Figma to Jama’s Adobe Photoshop.

The company is San Francisco based and was co-founded by Janis (CEO), a B2B SaaS expert. Trace Space allows engineers to produce collaborative notebooks, by providing a simple UI with robust traceability, time-based & custom versioning, item-level attributes, and flexibility to adapt to workflows.

Top Stories

• AEC SaaS, Snaptrude, raises A round from Accel (Techcrunch)

• The potential of LLMs in robotics (RoboDK)

• Automakers invest in Niron, which is making rare-earth free EV components (Electrek)

• The Dutch tradition of knowledge behind ASML (Bismarck Analysis)

• Humanoid robot startup, 1x, shares how they collect data for training (1x)

• Maersk cuts 10,000 jobs as shipping demand falls (BBC)

• RIP 3D printing, the cart before the horse (3D Print)

• The UK passes the ‘Automated Vehicles Bill’ for AVs (Gov UK)

• GPT moment for robotics is near (TechCrunch)

🙏🏻 Slightly shorter edition this week as have been busy cooking a few things up! 👨🏼‍🍳