Do We Need a Dedicated AI Device?

Editor’s Note: Happy Thanksgiving to our U.S. readers…

Our office is closed today for the holiday, so we’re taking the opportunity to feature an insight Jeff had on some personal AI devices that caught a lot of attention at the beginning of this year.

But despite the initial buzz, the reception to the devices – once they were in the hands of consumers – was overwhelmingly negative. So, what happened?

Well, as Jeff gets into below, these companies need more than purely interesting tech if they want to strike the right chord with consumers and investors…

They were two of the most interesting new tech product launches this year.

And they also happened to be two of the biggest flops.

This might seem like a contradiction, but it’s not.

In a rapidly emerging technological market, with seemingly endless business opportunities, it’s a race to gain a foothold as quickly as possible.

The allure of grabbing large market share – in what everyone knows will be a massive market – is what drives billions of dollars of venture capital investment into startups with uncertain outcomes.

The innovation and iteration are certainly interesting to watch. Whether or not we have access to those early-stage investments, we can still benefit as investors by observing what works and what doesn’t.

To that end, two great examples come to mind this year: Humane, with its AI Pin… and Rabbit, with its Rabbit R1.

Imagining the New AI Interfaces

Both the AI Pin and R1 products are interesting because they imagined a new physical user interface for artificial intelligence.

Both companies created novel consumer electronics products that envision new ways of communicating with generative AI technology, through the use of some new kinds of hardware.

Will consumers prefer to have a completely different kind of device – other than a smartphone or a laptop – to interface with their personal AI assistant? That’s the real question these companies sought to answer.

In the case of Humane, it developed its AI Pin device. It’s meant to be worn on the top layer garment so that it can “see” the outside world.

Source: Humane

The product itself is impressive in terms of hardware. It’s loosely similar to what an Apple Watch is built of, with a Qualcomm 2.1 GHz processor, 4 GB of RAM, 32 GB of storage, Wi-Fi, Bluetooth, and cellular connectivity.

It sells at $699 or $799, depending on the model, and comes with a $24 monthly subscription. The subscription includes unlimited talk, text, and data over T-Mobile’s wireless network in the U.S. In that sense, the device has a phone number and is positioned as a potential replacement for a smartphone.

I couldn’t help but be reminded of the modern Star Trek Combadge or Communicator, which acted as both a communications device and a universal translator. The Humane AI Pin was designed to be worn in the same spot and activated with the same kind of tap of the device.

But despite the AI Pin’s 2024 vintage compared to the Starfleet Combadge circa 2350, the AI Pin was designed with a lot more functionality in mind.

Aside from being able to take calls and interface with an AI assistant, the front of the device has an outward-facing camera for vision and a projection device for display and interaction.

Source: Humane

The interface that is projected onto our hands can be used to control the device by tapping the projected buttons on our hands. It’s also a display of information. And certain hand gestures also provide control.

And the camera can be used to identify surroundings and location. The AI Pin can be queried with questions like, “What am I holding in my hand?” Or “What building is this?” The camera can also be used to take photos and videos by simply tapping on the device.

While it is easy to see how a refined product like this has utility, the feedback on the device in the real world has not been promising.

A Wearable Tech with Awkward Silence

The AI Pin is actually not a pin. It has a magnetic backing, which doubles as a battery pack that attaches to the interior of a garment. Due to the weight, it tends to pull down the garment.

The projected user interface onto the hand can be tricky to use, with many complaints about the sensitivity of the system, often resulting in errors and frustration.

Battery life was less than four hours, and many reviews indicated that there were problems with the device overheating and needing to pause to cool down. The heat is significant enough to be felt on the wearer’s skin, beneath the AI Pin.

And there were delays in responses from the AI Pin when interacting with the device, often lasting several seconds, leaving the user sitting in awkward silence and wondering what was going on. And without a screen, the uncertainty was amplified.

This is all to say that there is too much friction in the user’s experience with the AI Pin. Advanced technology should ultimately reduce friction and make tasks easier to perform, not more difficult, which is why Humane’s product launch flopped so badly.

Humane was founded by two senior Apple executives who had seen great success in product development at the consumer electronics giant. Humane’s marketing definitely had an Apple-like feel to it. They even went so far as to reveal the device as a fashion accessory at a fashion show in Paris last November.

The executive pedigree was enough to raise about $240 million between 2019 and 2023 from the likes of Tiger Global Management, Sam Altman, Marc Benioff, Microsoft, Salesforce, Qualcomm, Softbank, and others.

And just like that, the company is now on the block, looking for a buyer.

Humane believes it’s worth somewhere between $750 million and $1 billion. But we should keep in mind, that’s the “ask.”

I’m sure Humane’s investors would love to see an exit at those levels, as it would result in a profit for some and a return of capital for others.

I’m afraid Humane isn’t going to get anywhere near that valuation, assuming there is a buyer.

A Funky Handheld Device with Limited Utility

The other recent product flop is a funky, retro-looking device – the Rabbit R1.

It’s hard not to like the design, and with a $199 price point and no subscription required, it is certainly an easier sell.

Source: Rabbit

This product got a lot of attention back in January when it was listed as a breakout hit at CES 2024. (CES is the annual Consumer Electronics Show, one of the most popular tech trade shows in the world, where companies introduce and showcase their most cutting-edge new technologies.)

The Rabbit R1 stands in contrast to the Humane AI Pin, as it’s a handheld device with a push-to-talk button (gray button on the side), an embedded camera that can rotate to face the user or the outside world, an analog scroll wheel (orange cylinder), and a screen on the front.

Source: Rabbit

The R1’s tech specifications aren’t too dissimilar to the Humane device, except for a larger battery, more memory, and a larger speaker. These features are easier to use, given the larger handheld size of the R1 compared to the AI Pin.

With that said, the overall product objective – to be a user interface to an AI assistant – is quite similar to the Humane AI Pin. The product idea out of the box is to provide access to a generative AI  – search, vision, translation – and music to the user.

And the product vision for the Rabbit R1 is to enable an action-oriented AI assistant, capable of making reservations, purchasing tickets, providing navigation, and providing research that’s contextually relevant to users’ interests.

But, just like the Humane AI Pin, the software is quirky and slow. And the rotating camera added friction, not function. After all, smartphones work great, having a forward and rear camera which make use easier. This would have been simple to do on the R1, and it wouldn’t have added much cost to the device.

It’s the use of simple voice interaction that made the R1 (and AI Pin) such a compelling and novel idea. For the first time, viewers of the technology could imagine interacting with devices without having to type in prompts on a computer keyboard or smartphone screen. But in the end, the functionality reveals itself to be no different than a more developed, AI-powered Siri or Alexa.

Rabbit is even earlier stage than Humane, having only raised $36 million between 2021 and 2023, with backing from Khosla Ventures and a handful of lesser well-known venture capital firms. I suspect Rabbit is looking for more capital right now to complete its vision for the product, despite the poor reception to date.

These comments are not meant to be critical. In fact, I found these developments give us good insights into user experience and raise some obvious questions.

Less Friction, More Utility

While some of the kinks and bugs can be improved upon in both devices, the real question is whether or not consumers will adopt yet another device to use artificial intelligence (AI).

The reality is that all of the functions that exist or are planned to exist – in the AI Pin or the R1 –  are already possible on our smartphones.

A new device won’t take the market by storm if it isn’t capable of doing more than our existing computer interfaces are capable of doing (like our smartphones). The friction, learning curve, and costs to change are too high.

Wearing a communicator on our chest or carrying around a new consumer electronics device only makes sense if it is capable of something we haven’t seen before, that makes life easier, and that removes the need for us to carry another device (like a smartphone).

The most likely and obvious path towards consumer adoption of AI is through devices that we are already using throughout the day. A smartphone, a pair of glasses, AirPods (earbuds), and for some, a smartwatch.

Smartglasses or augmented reality glasses, in the form of glasses or sunglasses, provide an AI system a lens to the real world and allow us to communicate, hands-free, with an AI assistant. AirPods do the same for a voice interface, and the utility of a smartphone continues to increase with each generation released.

I had an opportunity to invest early in both Humane and Rabbit. I passed on both. It’s not because I didn’t think the technology was cool and interesting. It is. But they both added friction and had less utility than what we use today. Their features are already available in a smartphone.

And what we want as both users and investors is less friction and more utility.

Regards,

Jeff