The smartphone of the future is powered by AI. What else?
Three years ago, I was the ghostwriter of Thierry Geerts, the CEO of Google Belgium. Together we wrote Homo Digitalis, a book about the impact of technology on society, intended for the general public. I would visit Thierry for weeks to hear his vision and then turn it into text. One of the topics that often came up was the end of the smartphone.
Our smartphones, Thierry said, will “disappear behind the wallpaper.” As everything around us becomes smart, from speakers to cars, we will no longer need a glass disc in our hands.
In 2021, it sounded like a scenario that might happen in the distant future. Devices like the Google Nest reached a niche audience at the time, and intelligent assistants in cars had the mind of a toddler. Three years later, that is entirely different, courtesy of ChatGPT.
ChatGPT in a box
Indeed, thanks to that “text-completion at speed,” an entirely new category of devices is emerging: smart assistants that fit in your pocket. Consider, for example, the Ai Pin from Humane, the startup of former Apple employees Imran Chaudhri and Bethany Bongiorno, which raised more than $200 million. In November, they presented a small device that hangs from your clothing and controls you with your voice. You ask aloud what you want to know; then you get the answer whispered to you. The Ai Pin also has a projector, which allows you to project messages on your hand.
Another such device is the Rabbit R1, one of the eye-catchers of CES 2024. The gem is reminiscent of a smartphone, though you no longer interact with individual applications. You talk to it, and then you can do anything you usually do with your smartphone, from booking a cab to listening to music. Jesse Lyu, the founder of Rabbit, says he doesn’t want to replace the smartphone. Or at least not anytime soon. In two days, pre-orders of the first 10,000 units went entirely out the door.
Those devices often feel like ChatGPT in a box — and in many cases, that’s because they simply communicate with OpenAI’s software via an API. At that company, Jony Ive, the man who gave us the design of several Apple products, is thinking about how OpenAI can make its own hardware product. SoftBank is said to have already released 1 billion for the development.
A solution without a problem
Each of these is a device with a solution in search of its associated problem. There seems to be no concrete need for it for now, but somewhere, they do feel like “a future.” In light of this, it is interesting to examine what smartphone manufacturers are doing. How should they position themselves from now on?
The new Galaxy devices Samsung presented at Unpacked seem to answer that question. Where such a presentation used to revolve primarily around screen brightness, battery life and ruggedness of the device, it is now increasingly about the software. This year, it is specifically about a magical two-letter word: AI. In the chip market, for a long time, Samsung seemed to miss a boat when it came to AI chips, but in terms of software, it is going much further than other smartphone makers.
Indeed, Samsung is bringing many features that we have seen emerge on desktop computers in recent months to our pockets. Think of software to improve texts (which I usually use Grammarly for, as English is not my mothertongue) or to automatically complement images (autofill from Adobe Photoshop) to algorithms that can write out and organize texts. A whole group of companies are being sidelined because Samsung now embeds their applications directly into the operating system.
Better photo editing, translation and summaries
Last Monday, I already got to test the new devices extensively. Because the Galaxy S24 Ultra feels very similar to my S23 Ultra, and the camera also seems to be just as good, I focused mainly on the new applications. These three stood out for me.
The photo on the left is an actual image taken with my Galaxy S23. With that same smartphone, I tried using the traditional magic eraser to remove the two pads, which produced a clearly edited image. With the S24, I not only succeeded a lot better, but I could conjure up a non-existent cushion in no time. The result is a photo that “could have been true” — the metadata does indicate that generative AI was used to create it.
Another interesting feature are the automatic translations when calling. When you call in your own language, the person on the other end of the line will automatically hear your message in, say, Korean or German, and vice versa. Currently, Dutch is not yet supported, but I think that is only a matter of time. The technology is reminiscent of Live Translate on the Google Pixels and Earbuds, but it feels more intuitive because it’s directly integrated into the calling software. I made about five calls, and they all went off without a hitch.
The third feature is one that we at my creative agency Hartstikke already use ChatGPT often for: structuring and summarizing texts. Put your smartphone on the table during a meeting, and the audio is automatically transcribed. The software knows who is speaking and can afterwards create a transcript according to the template you set. Afterwards, you can ask it to summarize it.
The verdict
With the new features, both Samsung and Google, which put similar things in the new Pixels, show that even apart from the commands via voice technology in the Rabbit R1 or Humane’s pin, there are a lot of AI applications you want in your pocket. It feels like a first step in the right direction, though they remain primarily the first experiments in a quest far from over.
