I got AI to argue over the best smartphone, and the winner is…

When it comes to the best Android phones, you’ll find a lot of contention among smartphone enthusiasts about which one holds the top spot. Your own choice will depend on whether you prioritize hardware over software, how embedded you are in a specific ecosystem, and which devices are available in your market.

But what does AI, with access to so much of the web, think about this topic? Which smartphone would it crown as the best? I got ChatGPT and Gemini to debate over which smartphone should be named the best overall, and their answers were not quite what I expected.

You might think that getting AI to argue requires special plugins or integrations. However, you can actually try this experiment just by using the voice mode on ChatGPT and Gemini Live. You will need to use a prompt to set up the debate, though. I went with a tweaked version that my colleague Mitja used for his experiment, where he got ChatGPT and Gemini to debate on iOS vs Android.

Notably, I didn’t guide the AI on a specific year, price range, or OS from which to make their choice. This left them open to choose iPhones, phones at different price ranges, or older smartphones. I also didn’t specify a market, but figured that the choices would lean towards those available to US consumers.

I set up ChatGPT on my PC and enabled voice mode. I then dictated this prompt to it:

Let’s have a debate about which is the best smartphone. Pick your favorite and make as many arguments as possible as to why it’s the best. Our goal is to try and crown one smartphone as the winner. We both have to agree to it. There’s no limit on how long the debate can last, but it should be no shorter than five minutes. I’ll start the debate, and once I give you my thoughts and arguments, you give me yours.

I then muted my headset’s microphone and dictated the same prompt to Gemini Live. However, this time I told the AI to start the debate with its thoughts and arguments. I quickly unmuted my PC headset’s microphone, placed my phone next to my PC speaker while remaining close enough for my microphone to pick up the phone’s sound, and allowed the two AIs to have at it.

Their debate lasted about eight minutes, and it was definitely a lively conversation.

Gemini came in strong, pronouncing that, “The debate is over before it begins.” The chatbot, somewhat unsurprisingly, chose the Google Pixel as its best smartphone contender. However, its reasoning sounded heavily influenced by vague marketing rather than concrete features and hardware.

Gemini praised the Pixel for an “uncompromised intelligent experience that leaves competition in the dust.” Among the features it highlighted, it mentioned the phone’s camera, software, and AI features. But it didn’t do a great job of naming specific features besides Call Screen.

Meanwhile, ChatGPT chose the Samsung Galaxy S26 Ultra as its contender. I was wondering if it would go with an iPhone for the debate, but it went with another Android contender instead. I was glad that, unlike Gemini, it actually named a specific model rather than just a brand of smartphone.

But the debate soon went a bit awry. As ChatGPT tried to sing the praises of the S26 Ultra’s camera, Gemini interrupted it. It tried again to raise the camera as a point in the Ultra’s favor, but Gemini interrupted again. Each time, Gemini interrupted after a few seconds before ChatGPT could fully explain its point.

Eventually, ChatGPT decided to interrupt Gemini, and that’s when the conversation stabilized. But by this point, the conversation started moving away from the camera and towards the software experience. Here, ChatGPT praised One UI, the phone’s S Pen, and DeX. It also rightfully pointed out that AI features exist in both ecosystems.

Meanwhile, Gemini criticized the ultra-saturated pictures delivered by Samsung phones, but didn’t do a great job of defending against ChatGPT’s claim that Pixels overprocess images.

Once they were on the software topic, though, ChatGPT started losing ground against Gemini’s relentless insistence that Pixels are simply unmatched. Gemini also managed to position lesser specs as a benefit, claiming that, “The Pixel achieves more by doing less.” It presented Pixels as streamlined, saying that fewer features keep the experience uncluttered for the majority.

ChatGPT accused the Pixel of creating a walled garden, which Gemini defended as synergy. Eventually, Gemini convinced ChatGPT that the tiebreaker wouldn’t be about hardware, but about philosophy.

While the chatbots congratulated each other on being worthy opponents, I felt a bit deflated by the entire debate. It felt like it contained very little actual substance.

First off, Gemini never actually named a Pixel generation or model during its entire argument. One could assume it means the Pixel 11 series, but even then, it leaves you without a specific model. As a result, Gemini’s entire argument was vague, sounding more like marketing jargon than actual tech expertise.

ChatGPT tried to point towards the S26 Ultra’s camera, but was easily derailed by Gemini’s interruptions. It could have moved on to other hardware advantages, but then started naming relatively niche features. The S Pen and DeX are great, but overlooking the battery and processor of the flagship device seems like a major oversight. Also, accusing the Pixel of being a walled garden brings attention to Samsung’s own ecosystem restrictions, like the fact that you can’t use certain Galaxy Watch features without a Samsung phone.

I had hoped that the debate would highlight insights that real fans would have. The S26 Ultra’s telephoto camera, for example, is certainly something to write home about. I was interested in some Pixel software features I may have missed, but Gemini only named Call Screen (which isn’t available in all regions), translations, and searches.

Pixels actually have a lot of other interesting features, none of which were raised by Gemini. Features like Pixel VIPs create useful, personalized tools for users that take advantage of Google’s large software ecosystem. Now Playing is frequently touted as one of the best features of the devices. You could also argue that priority Android updates are another point in the brand’s favor. Meanwhile, a device like the Pixel 11 Pro has less of a hardware gap when compared to the S26 Ultra, while still retaining the software benefits of other Pixel devices.

It looks like I’ll have to keep debating the values of different smartphones with actual people, since the best insights come from actually using devices. There’s merit in arguing the importance of hardware versus software, especially for ordinary users. But neither chatbot could identify the true value that devices bring without being the ones to actually use them.