On January 17, Samsung held its traditional new product presentation ceremony, Unpacked. But this year’s edition had a different spice. Samsung touted the event as the arrival of Galaxy AI.
The new phones of the Galaxy line, the company’s main one, were revealed. But the technical specs have hardly garnered as much attention as the artificial intelligence features the company has emphasized so much.
Among the possibilities are the simultaneous translation of messages and calls, summaries of notes and the editing of text to different tones (more casual or more professional, for example). The interesting thing is that some of the features run natively on the phone itself.
The announcements weren’t exactly a surprise. In times of ChatGPT and the race to develop more sophisticated language models, it was only a matter of time before generative artificial intelligence appeared on smartphones. Samsung just jumped ahead, at least in the sense of riding the hype of the moment.
And this movement is also happening in other segments. Companies are scrambling to figure out where to insert artificial intelligence tools, and how to make them useful to the consumer.
Artificial intelligence for everyone
The week before Samsung Unpacked, another major tech event took place: the Consumer Electronics Show, or CES. The theme of the 2024 edition said it all: “AI for All”.
Do you want to make a good steak but don’t trust your cooking skills? The AI grid saves you. Want to identify birds more accurately? Binoculars with AI. Do you need to rock your child? The solution, of course, is the AI stroller.
All of this is to be expected at a CES. There are always devices that are too quaint or too ambitious, that may not even reach the end consumer. But the guideline seems clear: put AI in everything possible.
Despite the weird products and applications, there’s at least one aspect where the use of artificial intelligence tools makes a lot of sense: the use of large language models (LLMs) for natural language understanding.
This is the bet behind the integration of ChatGPT into Volkswagen’s voice assistant, for example. Some of the use cases involve requests to plot routes (“I want to go to place X”) or activate some feature of the vehicle, such as air conditioning and windows.
In the same logic, a personal home robot, such as the one presented by LG, needs to understand what its owners ask for. Hence the need for a system that understands natural language, and LLMs are there for that.
The potential of these language models is there, and now companies are looking for practical uses. Within this context, there are even those who promise totally innovative devices.
On Mobile and Beyond
One of the highlights of CES was the R1, an artificial intelligence gadget produced by Rabbit. CEO Jesse Lyu’s promise is that the device will use his apps for you.
The device would be able to understand an order made in natural language, such as “Ask me for a pizza,” and use an app to order the food. The user would only need to confirm the order on the R1 screen.
On the Rabbit portal, the user logs in to the applications they use. This is how R1 knows exactly where to order food, or make a purchase. Lyu calls this technology a model of action; the logic there is that it not only understands language, like an LLM, but is also able to perform objective actions.
The R1 arrives at a very friendly price for the American market, US$ 199.00, which certainly contributed to the success of sales. About ten thousand units have already been sold.
Despite the difference in design and price, the R1’s proposition is similar to that of the AI Pin, announced last year. Devices that execute natural language orders, eliminating the friction of navigating between multiple apps. Whether this poses a threat to the gadget normally used to run apps, the mobile phone, we need to wait and see.
However, smartphones can also have access to language models, either natively or through the cloud. And it’s unlikely that Samsung will be the only one to offer this experience to its users.
In cell phones and gadgets that promise to replace them, in cars, TVs, appliances: we will see artificial intelligence everywhere. It remains to be seen in which segments it will be confirmed as a practical evolution, and in which it will pass as another of the technological hypes that have been left behind.