Will We Ever See A Post-Smartphone Future?
AI "personal assistants" such as Humane & Rabbit threaten to replace the smartphone. But it looks like threats are all we'll see.
Welcome to another edition of Always Learning.
I’d like to apologize for my recent hiatus. I started the year very strongly with releases, publishing once a week. However, as tech and business news has slowed down, topics have been hard to come by. I have also been working on some personal goals in that time. That being said, I’m back and will continue to write great pieces for all of you as often as I can. Weekly or biweekly at most.
A quick announcement. We recently hit 300 subscribers! I want to thank each and every one of you for making that possible. If you’re not subscribed, please do, it’s free and it really helps. Feel free to share with a friend as well.
Enjoy the piece.
It has been an eventful few weeks in the world of consumer tech products. Humane and Rabbit, two ambitious startups, have both recently launched the first wave of AI enabled personal hardware products - namely the Humane AI Pin and the Rabbit R1. Both products have been received poorly, coming under fire for various reasons, some valid, some unreasonable. Today, we’ll go over why a transition into a post mobile world may be more difficult than we all expect.
Humane AI Pin
The Humane AI pin is a wearable personal AI-enabled device from Humane. The company was founded in 2018, by former high level Apple engineers. The founder, Iman Chaudhri, is credited as an inventor on several Apple patents, including a touch screen. The product, which is their first, released in March to an onslaught of moderate to negative reviews. Majority of them highlighting slow response times, poor battery life and the $700 price tag! (with a monthly $24 subscription fee).
Despite this, the device still represents a new paradigm in consumer hardware. Through the power of an on-cloud large language model (LLM), users are able to make calls, texts, order Ubers, make food delivery requests, take high definition photo and video, all at the command of their voice, without having to flip through a screen and open various apps to complete each of those tasks. The product is capable of this while being the size of a few poker chips stacked together. Its too early to tell, but this may indeed be the future.
Rabbit R1
The Rabbit R1 is another ambitious AI-enabled product that aims to simplify the mobile computing experience through the use of its proprietary large action model. The device is about the size of a pack of sticky notes, with a 2.9 inch screen, a scroll wheel and a camera. It promises a similar feature set to its Humane cousin, being able to play music, order food and rides, and even give explanations to various things the camera is pointed at. An all-in AI assistant.
However, the execution leaves a lot to be desired. Various reviewers have noted that the device fails at even its most simple promised tasks, such as ordering food and describing items. The battery life is reportedly woeful and most of the features are apparently yet to come. A $200 skeleton of an AI assistant.
Why did Humane & Rabbit take this approach?
The two devices were clearly rushed out by their respective companies. Humane and Rabbit favoured capturing consumer mindshare and gaining quick customer feedback over slowly perfecting their products until they were ready. While it seems naive, this strategy was forced on them as a direct result of the current mobile landscape, which is highlighted by the lock-in and entrenchment that the major players - Apple and Google - have.
“Why isn’t it a mobile app?”
As I covered in “App Store & Play Store: Apple and Google's Playgrounds for Impunity”, the two tech giants have a dominant monopoly on the mobile market. Here is a key excerpt from that piece.
When Apple began to allow developers to sell apps and in-app-purchases on their stores, they demanded a 30% commission on all sales. Google also settled on 30% for their Play Store. At the time, not much was thought of this figure, as the software was still nascent and it was unclear how valuable that distribution would be. Today, it has become a major point of contention, especially because it affects how much developers such as Netflix and Fortnite can charge consumers for these apps. So much so that various companies have taken the matter to court.
The Apple App Store and Google Play Store are dominant distribution platforms for mobile applications that act as complements to both mobile phones and mobile operating systems. Notably, Apple and Google are in the business of both. A great mobile app such as Uber would convince many a customer to get an iPhone, or a phone with an Android operating system. Thus the complementary nature. We don't think about this much now - 20 years on- but this is primarily the key selling point of our phones - what we can do with them.
Apple and Google benefit from being feudal landowners of the digital app market greatly. This changes however, when an app threatens to undermine the core competency of the app platform provider. When this happens, they typically either ban the app from their stores or limit functionality greatly. Beeper, a recent app that enabled Android users to seamlessly use iMessage, is the most recent casualty of this, as user accounts were banned forcing the app to shut down. Best not poke the bear, huh?
Mobile apps have to walk a tight rope - providing a great product without undermining the platform provider. Sometimes, however, the nature of the product makes this impossible. Imagine Humane and Rabbit were apps; and you could order an Uber off them in a single sentence, something that is currently impossible to do on Apple’s native Siri or on any Android assistant. Two things would likely happen; first; the Apple/Google might ban the app entirely. Second, Apple/Google would put their enormous resources to use to replicate the feature and put it on their platforms by the next software update. All of a sudden, as an app maker, not only do you have no distribution, you have no unique product either.
As summed up above by
, this leaves the startups with no choice but to develop their own hardware, where they will be free from indiscriminate wrath of trillion dollar platform providers.“Why are the products so unfinished?”
It’s no secret that Humane and Rabbit adopted a ship first, iterate later strategy. The unfinished nature of the products and the sheer number of features “coming soon” let us know that the startups were determined to launch and answer questions later. Why is this the case, wouldn’t it be better strategy to launch a complete finished product? Let’s analyse.
Developments in AI have been moving at rapid pace since ChatGPT launched 18 months ago. New tools and models launch every few months with increased capabilities. These 2 products are a beneficiary of this rapid development as they employ large language models (LLM’s) to enable their different features. Notably, the tech giants Apple and Google were relatively slow to the party. This comes as no surprise, as big companies have various checkmarks to cross before they can release products into the world. They simply cannot move with the reckless abandon that start-ups can.
However, nearly two years on, the giant incumbents are slowly releasing products to market that take advantage of these AI breakthroughs. In about a week’s time, on May 10th, Apple will have its worldwide developer conference (WWDC) where they are expected to announce new AI enabled features. Shortly after that, on May 14th, Google will take the stage for their Google IO conference to potentially announce the same.
Humane and Rabbit were no doubt determined to get their products out there and capture customer mindshare before they inevitably got drowned out by what Apple and Google would announce.
On the other hand, being a version 1.0 product in a new product category means that it is quite hard for the companies to pinpoint what the killer use case will be for these personal AI assistants. Nobody quite knows yet. A quick development and release cycle such as this that minimizes time and other resources allows the companies to gain quick consumer feedback and use that to get closer to what the final finished product will become in the next iteration.
This of course, comes at the expense of consumers. It would be a shame for anyone to spend their hard earned dollars on unfinished products, regardless of how promising future iterations of them may be. That said, we live in the most informed times in human history. If one would like to get feedback on a product before purchasing, they are free to do so for free all over the internet. There will no doubt be early adopters who buy into the products anyway just to get a glimpse of what the future may or may not hold.
I am all for innovation. History has shown us that human beings only progress when people try things. I will forever be an advocate of trying. All the same, some paths are more difficult than others, while some are outright impossible. I’ll be keen to see whether AI assistants will be the former, or the latter.
What these products have in common is at their current levels they are like fancy designer cases for existing technology. They seem to have a unique form factor which is (perhaps) necessary but unfortunately not sufficient to truly disrupt the smartphone. It's like buying an expensive watch that can also tell time – a feature your smartphone in your pocket already offers.
Solid read! However I’d say these are genuinely bad products and shouldn’t have been shipped. We can’t expect users to pay for unfinished hardware. A good question I always like to ask is what genuine problem does this product solve? And if it does, does it present a desirable solution ?