Key Takeaways
- Rushed products like Rabbit R1 and Humane AI Pin highlight the danger of prioritizing speed over quality in tech innovation.
- Fisker Ocean’s software issues show the importance of creating user-friendly experiences in new products from the start.
- Companies must focus on quality over speed to avoid releasing half-baked tech that disappoints consumers and hinders innovation.
April 2024 was the month of AI hardware, and it did not go well at all. Both the Humane AI Pin and Rabbit R1 were recently released as AI hardware devices promising to change the way we interact with the world. As you can probably infer from the title of this post, both devices failed to live up to their ambitious claims. While first-gen products aren’t usually the best, there’s a much deeper problem going on here.
We are observing a problem that has been plaguing the game industry for a while. Developers will release a game in a buggy, unfinished state for some reason, and then try to fix the mess in a couple of months (if we’re lucky). As a passionate follower of both the tech and gaming industries, this is a trend I’ve hated for a while now. Unfortunately, it’s only gotten worse in 2024. Today, we’ll be taking a look at some products that highlight this issue, and what it means for the tech industry as a whole.
Related
The Rabbit R1 proves why dedicated AI needs to be something different
Dedicated AI devices will need to be something special to deserve a spot in your pocket
Fisker Ocean
An EV gone horribly wrong
Fisker Inc. is a relatively new American company that was preparing to take on Tesla with their all-electric crossover SUV, the Fisker Ocean. It has a unique exterior design, comfortable and spacious seats, with a fantastic driving experience. It even has solar panels on the roof that can help add some extra miles to the range.
Unfortunately, the car had several software issues that turned it into a buggy mess. People had issues with the infotainment system, malfunctioning drive modes, Bluetooth connectivity problems, and random issues with the key fob. It also lacked a glove compartment on the passenger side, unclear labeling on certain controls, and no automatic brake hold.
All of these user experience issues led to many annoyances, even though the car is fundamentally great. Fortunately, Fisker fixed most of these issues with the release of Ocean OS 2.0. This software fixes most of the problems with the Ocean, but it was too late. In March 2024, Fisker paused the production of the Ocean due to financial difficulties. Word on the street is that they are on the edge of bankruptcy.
Production delays and management issues make it hard for a startup company like Fisker to survive. This means that if you bought a Fisker Ocean, you’re likely not going to receive future software updates that can fix current issues. It’s a terrible and sad situation all around.
Humane AI Pin
The definition of overhyped
We knew the Humane AI Pin was not going to end up in our favorite products of April 2024 list, but no one expected it to be this bad. The goal of this wearable AI hardware is to reduce reliance on smartphones by giving you access to essential functions through large language models (LLMs) like GPT-4. You can ask questions, make calls, take notes, and communicate with it naturally by using your voice alone.
While the concept is great on paper, the AI pin struggles to keep up with its biggest competitor, the smartphone. This wearable AI is slow, inconsistent, has terrible battery life, overheating problems, and an extremely high price of $700. To top it all off, you can’t actually interact with the AI if you don’t pay the $24/month subscription fee. That’s right, you need to pay a monthly fee to use your new $700 device, or else it turns into a paperweight.
All that for a product which is not nearly half as good as your phone. In almost every situation, using your phone is going to be quicker and easier. While LLMs are great at natural language processing, they still tend to hallucinate and confidently give you wrong answers. For the price, it’s a device that no one should be buying.
Rabbit R1
Great idea, terrible execution
Much like the Humane AI Pin, the Rabbit R1 is also half-baked. So much so that our editor-in-chief Rich Woods gave the device an embarrassing 2/10 in his review. The R1 runs on a Large Action Model (LAM). It is supposed to take voice input and interact with apps the way that you would yourself. For example, it can use an API to order food from DoorDash, play a song on Spotify, request a ride from Uber, or generate images with Midjourney.
The idea is that the LAM will learn and improve the way it interacts with these services over time. It’s a great idea in theory, because it could save you a lot of time if you never have to download an app or use one yourself. Unfortunately, it doesn’t have enough training data to be accurate. As such, it will often play the wrong song, deliver your food to the wrong address, or even mess up your Uber ride details. When it is unreliable, it wastes time; instead of saving you headaches, it creates them.
On top of that, the scroll wheel navigation is clunky, you can’t set an alarm on it, and it just doesn’t have any value over your smartphone. It very much feels like a product that was rushed to market.
So, what does this all mean?
We see signs of a bigger problem
The reason these products are bad is simple: they were rushed to market in a state that was not truly ready for release. Fisker wants to compete with Tesla, but they became so caught up with that idea that, while they made a great car, they failed to provide a comfortable and frustration-free experience. Similarly, both Rabbit and Humane want to grab a piece of the AI pie before Apple and Google eventually go ahead and incorporate the same enabled functions into iOS and Android.
Even the Apple Vision Pro is half-baked. While you can safely expect Apple to improve its product in the near future, it’s still strange that they released a $3500 device that doesn’t have much to offer right now.
I’m all for companies experimenting and trying to create fun hardware. Unfortunately, half-baked tech is hard to review, and it leaves a bad impression on us as customers. A lot of people are already skeptical of stuff like virtual reality, AI, and electric cars. Products like the ones discussed in this post confirm those doubts, and hence result in people being less interested in the newest tech innovations.
Rushing products to market is a quick race to the bottom. It discourages long-term investment in research and development, and hinders genuine innovation in tech. This also ends up slowing down market adoption due to skepticism and poor performance.
Looking ahead
I realize that this write-up is a bit dire, but this is a situation that we need to be aware of. Companies need to focus on quality over speed in order to create products that are polished and user-friendly before release. There should also be a clear roadmap for improvement and a focus on core functionality. We as consumers also need to vote with our wallets and not buy tech based on the promise of future improvements. Fortunately, there is still enough good technology out there to outweigh the bad. We do our best to highlight this here at XDA, so we appreciate you reading along as we truly examine the latest & greatest the industry has to offer.
