These days, I can’t go anywhere without seeing a smartphone or device enhanced by AI. It appears to be a catch-all buzzword companies use to increase the value of a product. I’m pretty sure it’s featured in air fryer I got for Christmas — after all, who doesn’t want AI-augmented bacon? It reminds me of companies using terms like “blockchain” or “internet” to help bolster projects in years past. If you had a floundering stock in the 90s, adding “internet” to your business plan increased your valuation overnight. A few years ago, if you had a game or idea for a service, implementing blockchain technology gave your venture credibility — not so much anymore.
But unlike previous hype tools, AI has significant consumer product potential, so I would hate to see it devolve into a cheap tagline slapped on everything. And with companies like Samsung andGoogle looking to bring it to every productfrom smartphones to the world’s most popular search engine, it seems like a tool that’s here to stay — so long as companies can figure out what they’re doing with it.
AI is overwhelming
I can’t tell you how often I’ve sat at the ChatGPT prompt without knowing what to type. It seems I pay $20 a month to ask what it would look like if my cat were rooting for my beloved Miami Dolphins (editor’s note: Go Bills). I know it’s capable of much more, but it’s almost too open-ended. Where am I even supposed to start?
To convince me that AI is the future of smartphones, companies need to focus the broad capabilities of AI into prompt-less enhancements of apps and features. I don’t want to sit there and figure out where AI can help the most; I want developers and manufacturers to determine which features of their devices function better with AI.
Google does it best
With thePixel 8 Pro, Google displays what AI could (and should) look like in smartphones. The company’sGemini AImodel is scalable, allowing it to run on multiple device classes. Google uses Gemini Nano for smartphones starting with its latest flagship, promising productivity enhancements and improved Assistant with Bard capabilities with computing handled on the phone — no cloud access required.
Currently,Gemini is only used in two apps, but the company promises more integrations going forward. Even with a small sample size, we get an idea of the enhancements AI can bring to software without needing user expertise. Summarize in Recorder is a fantastic tool, giving users the ability to boil long interviews or lectures into the salient points — without the need for an internet connection. Even if you aren’t sleeping during class, AI is a useful backup, picking up information and points you may miss during the broader presentation.
Google wants AI to augment software so that the user doesn’t have to know what they want done to get the best result.
Smart replies in Gboard might initially seem a little underwhelming, but they yield better contextual responses in apps like WhatsApp. Gemini will follow the conversation and adapt to tone and topic, framing responses in your own voice — your friends won’t feel they are getting a response from a robot, and you can keep your eyes on the road.
Video Boostand Night Sight video are not Gemini-specific, but these features highlight Google’s uses of AI throughout its computational models. Lighting and color adjustments, stabilization help, and noise reduction are aided by AI, providing crisp, clear video that would otherwise be limited by the raw data coming through the camera. Some features, like Best Take, are slightly more controversial, but Google’s overall direction is clear — the company wants AI to augment software so that the user doesn’t have to know what they want done to get the best result.
And while I don’t agree with Google’s limitations on Gemini features — the Pixel 8 is powered by a Tensor G3 the last I checked — I do agree with the direction of focusing AI to serve consumers better.
Not to be outdone
With theupcoming Galaxy S24 Ultra, Samsung aims to make its own AI splash. This follows theSnapdragon 8 Gen 3 launch, where Qualcomm displayed the AI enhancements the new chipset makes possible. Samsung plans to use AI modeling in its photo and video software, fixing lighting issues and removing any graininess — stop me if you’ve heard of these features before on another device. Much like Google, Samsung’s approach combines server-side and on-device computing.
I’m not saying Samsung is trying to copy Google; all cars have steering wheels, but the battle lines where companies attempt to sell us on the benefits of AI are clearly being drawn, and Samsung is intent on joining the fight with the S24 Ultra. I hope that Samsung doesn’t find artificial (groan) ways of charging more for a device based on AI features. I cringe at the thought of seeing the S24 Plus nerfed in the same manner as the Pixel 8, with features held back despite more than enough horsepower available. One thing can be sure: if companies can find a way to squeeze a few extra dollars out of us, they will.
Will it all work?
If AI doesn’t take hold on smartphones, it won’t be for a lack of trying, as Google and Samsung appear set to invest significant amounts to make it work. While I don’t doubt AI’s capabilities, I doubt companies' ability to market and implement the technology properly, leaving me relatively unconvinced in the short term — but I’m enjoying my cat in a Dan Marino jersey in the meantime.
Google Pixel 8 Pro
The Google Pixel 8 Pro with the Tensor G3 leads the way in AI with several useful features. Gemini Nano allows on-device AI computing for Summarize in Record and contextual responses on Gboard, with several new features promised in future updates. It also sports the Super Actua display, giving the Pixel 8 Pro the best brightness on a Pixel device ever.
