What will come after the smartphone? That’s the riddle that, right now, so many tech startuppers and big companies alike are trying to crack.
If we’d like to be realistic, we all know the answer is probably “yet another smartphone, but smarter, faster, more powerful, with lots of AI in it”.
Unfortunately, the rise of stochastic parrots, also known as large language models, has convinced a new ilk of young tech entrepreneurs that they can and will be the next Steve Jobs. They believe that by repackaging intangible generative AI models into hardware shells and sprinkling some magic marketing powder on top, they could be singlehandedly responsible for the new invention of the century.
The rise of AI gadgets: solutions in search of a problem
Here are 7 much discussed gadgets with artificial intelligence on board: supposed to be the next big thing in tech, so far most of them just make us say “this could have been an app”.
View Article details
- Andrea Nepori
- 01 October 2024
It’s a novel mix of naivete and disingenuousness, even by Silicon Valley standards. The most fascinating aspect of this new wave of AI hardware is that the startups making them all try to convince us of the need for a new single-purpose device to break away from our smartphones. As if we didn't like our smartphones! Or maybe these new-generation startuppers are too young to notice that the fundamental and utter advantage of the smartphone is the ability to do so many things in one. Including running apps that would be more effective than any of these devices in achieving their fundamental functions.
A lot of hope and a lot of money were poured into Humane's AI Pin project. The device, which was supposed to be the poster child of the post-smartphone era, turned out to be a barely baked device marred by usability and reliability issues. Reviewers destroyed it, sales lagged, and the promising startup is now already discussing a potential sale as it battles for its life.
Introduced at CES 2024, the Rabbit R1 attracted much attention for its low price and boastful claims about its intelligent functionalities. As it hit the market, it quickly became clear that the device wasn't more than a cheap hardware shell for third-party models such as chatGPT, which had severe security and privacy issues. Moreover, the device often didn't work as advertised. We wonder if the company will have enough runway for a second version of the product.
Brilliant's Frame Glasses promised to succeed where Google Glass failed, thanks to the addition of generative AI. That is, unfortunately, not the case.
The current version of the Frame glasses only offers a monochromatic text-based eye-sight display and has to connect with your smartphone anyway for all the AI heavy-lifting.
If there were a prize for the most tone-deaf, late-stage Silicon Valley of all AI Gadgets, Friend would win it. The device is a pendant that can listen to conversations, process their content through an app, and spit out LLM-generated platitudes about the topic at hand to motivate, scold, comment, or add an opinion. Yes, it sounds like a gadget from an episode of Black Mirror.
Plaud has created two portable recorders that can transcribe and summarize every recording through GPT 4o and Claude 3.5 Opus. We're sure the devices will work as advertised, but they do leave us with a question: Why would anyone want to burden their pockets with just another gadget that offers nothing more than many other smartphone apps?
Terra è un dispositivo tascabile che mira a sostituire gli smartphone per la navigazione mentre si cammina. L’idea è di sfruttare la potenza dell'intelligenza artificiale generativa (chatGPT) e le API di GooglePlaces per guidare gli utenti attraverso una passeggiata in modo più organico e meno invasivo. Di tutti i dispositivi elencati in questa galleria, Terra è l'unico che ha almeno una giustificazione progettuale e una filosofia di design alla base.
So far, the Rayban Meta Glasses are not exactly a flop—quite the opposite. Nonetheless, the way their AI features work (or not if you're using them in an unsupported region) still leaves a lot to be desired in terms of precision, privacy, and actual applications. They're also a privacy nightmare, but that seems to have become a minor concern for many younger users.