Key insights of today’s newsletter:
Recent product launches of the Humane AI Pin and the Rabbit R1 have been deeply disappointing.
These disappointments expose a gap between how AI products are being marketed and hyped up vs. what they’re actually capable of.
It’s part of a bigger trend that is rooted in Silicon Valley’s obsession with attracting investors and generating media attention at the expense of delivering real value to customers.
↓ Go deeper (8 min read)
There is no shortage of vision in Silicon Valley. They probably understand it better than anywhere else in the world. You have to turn people into believers. You got to sell them on a dream.
Generative AI has made many of these dreams seem more attainable than ever before. Yet, a big gap remains between the promises made and what the technology can actually deliver.
This is best illustrated by the recent product launches of the Humane AI Pin and the Rabbit R1. The first was described by Marcus Brownlee in a scathing review as the worst product he has ever reviewed. The latter secured 100,000 pre-orders based on a demo and an Apple-style presentation at CES 2024. But now that the first devices are being shipped, it appears to be lacking most of the features that were initially advertised.
It begs the question: when does a good sales pitch turn into a bait-and-switch? And where is the line between aggressive marketing and an outright misleading demo?
Meet your AI-assistant-in-a-box
The Rabbit R1 and Humane’s AI pin are both attempts at creating something I like to call the AI-assistant-in-a-box: physical objects with a virtual assistant inside. I won’t be reviewing either device in detail, but it’s worth highlighting how both fall short in their own way.
The Rabbit R1 is being marketed as a funky, voice-first pocket companion priced at 199 dollar. The big promise of this device is that it can do stuff for you, like navigating your apps, powered by a so-called large action model (LAM). All you have to do is ask.
It’s actually quite cleverly set up, but the technology appears to be in its infancy. The company is keen on selling the idea that the AI assistant can seamlessly interact with hundreds of apps on your phone, but now the device is being shipped there are only 4 available connections: Spotify, Doordash, Uber, and Midjourney. And going off early reviews none of these seem to work reliably (alongside a host of other issues that have been called out).
The Humane AI Pin takes a different shot at the same problem. This device can be best be described as a digital brooch. It has a built-in camera, laser projector, and to activate it you can tap the device with two fingers and proceed to ask it a question.
As for the big promises, I’ll quote a line from co-founders Imran Chaudhri and Bethany Bongiorno:
“Humane’s AI Pin is the embodiment of our vision to integrate AI into the fabric of daily life, enhancing our capabilities without overshadowing our humanity.”
Right. So, what you’re actually buying is this: the AI Pin is an AI-assistant-in-a-box that can’t be paired with your smartphone. It’s a stand-alone device that starts at 699 dollars and you pay a monthly subscription of 24 dollars on top of that.
A non-exhaustive list of shortcomings include: it’s slow to respond (to the point that it’s quicker to take your phone out of your pocket and google it yourself), can’t do basic things like setting a timer, heats up while using (so much so that it crashes), and because it is a standalone device, it has its own phone number and you can’t access anything that’s on your current device, like your text messages, emails etc. Suffice to say, this thing does not live up to its promises nor does it justify its premium price.
Oh, and did I mentioned the hallucinations already? Both the AI Pin and the Rabbit R1 hallucinate a lot, confidently answering questions wrong, as we’ve gotten used to from LLMs.
Everything is upside-down
All things considered, these AI devices show what’s possible if only the technology were better. Right now, though, these products feel wholly unfinished, yet, they are being shipped and marketed as if they are.
‘Underpromise, overdeliver’ used to be a respected business adage. But that’s not how things are done in Silicon Valley. In Silicon Valley, everything is upside-down. You pitch and then you build. You ask for money first and only then you figure out your product-market fit.
Which reminds me of this skit:
Don’t get me wrong, great things come from Silicon Valley, but its obsession with the next big thing has created a culture that is driven by the need to attract investment and generating as much media attention as possible. It’s pushing companies to market their innovations aggressively even if it means cutting corners or bending the truth. Move fast and break things. Overpromise, underdeliver.
Funny enough, after Marcus’ brutal review of the Humane AI Pin, some people found it to be overly negative and asked whether he, considering his reach, should have gone easy on them. In a response video, he asks his viewers:
“Do bad reviews kill companies or do bad products kill companies?”
If you ask me, we don’t need less scrutiny, but more. Remember kids! If something looks too good to be true, it usually is.
Join the conversation 🗣
Leave a comment or like this article if it resonated with you.
Get in touch 📥
Shoot me an email at jurgen@cdisglobal.com.
I commented on a substack post in January about Rabbit based on the demo. I have not used the product, so all my observations may not be accurate.
It will have to overcome several challenges:
1. Voice User Interface. I am sure many people may like it, but not everyone. Siri, Alexa, and others have failed with Voice UI, and even Humane AI Pin is struggling.
2. Another device to carry (what is stopping someone from building a master app on iOS and Android to replace this over time, even if this device has some hardware/software advantage today?)
3. The price is very low compared to Humane AI Pin, but without any subscription model, they have to find ways to make money beyond selling more and more devices as all these AI calls and building/maintaining connectivity to all these services will become costly)
4. I read somewhere the quality of the speaker is not great, but hopefully, it will have a good quality with the headphones. At $200, you will expect compromises. What else has been compromised to keep the cost low?
5. Do not know if they will provide a way to build connectivity to other services so that whatever they do not want or cannot do, the customers/developers can create. Is there a way for developers to make money?
6. I am personally not a big fan of using another device to configure something, as generally, it expands to doing more things on the other device to make it work over time.
7. The screen feels small but may be good enough.
Jurgen, thank you as always for these analyses!
On the topic of AI-driven physical products in my opinion there has been an over focus on manufacturers and products, driven by technological developments, and a lesser focus on consumer behavior.
They were not exposed to what needs or desires this product could fill, to explain what it was capable of, to introduce it in the right way in their consumption basket. These are new product categories in people's minds, and perhaps not enough thought was given to creating an appealing category for the people who would buy it.
As much as they certainly target people who are 'technology enthusiasts', the attempts have not been judicious and careful, even in positioning and communication.