This spring, the hype around artificial intelligence (AI) seemed like it was starting to hit reality. Several start-ups launched widely publicized hardware with a new generation of voice assistants, generating answers using technology like that behind OpenAI’s ChatGPT. In particular, buzzy devices like the Humane Ai Pin offered travelers an intriguing pitch: What if AI assistants could help you with contextual information while you’re traveling—not just travel planning but actual on-the-ground adventuring?
Imagine, for instance, you’re walking down an unfamiliar street; the Humane Ai Pin could use its onboard camera to identify where you are and offer context like a tour guide. Pop into a restaurant in a country where you don’t speak the language and it could explain each dish on the menu to you, then translate dietary needs to your server in near real time. It sounds compelling, but then the pin launched with reports of glitchy hardware and commonly inaccurate answers to questions—and a lot of tech reviewers looked back to their phones with greater appreciation than before.
“People are trying to be that first iPhone for this next-generation category,” explains Julian Chokkattu, the reviews editor at Wired, whom I spoke to while he was in the middle of testing several of these new AI devices. His recent reviews concluded these gadgets often have serious usability issues (like overheating in the case of the Humane Ai Pin), have frustratingly limited functionality at launch, and are prone to provide incorrect information. “It’s going to be some time before anyone really figures out what is going to stick and what’s gonna work,” he concluded.
That’s an apt assessment of AI writ large. The term AI itself has become an unclear and imprecise marketing buzzword, which doesn’t really help everyday travelers understand the various technologies grouped under that umbrella. For example, AI is often used to describe automated recommendation systems, which are not new; they have for years been used in travel-booking platforms such as Airbnb, where its machine-learning algorithms influence things like search rankings and personalize your experience. The kinds of AI getting all the hype right now—and the technology powering the core features in these gadgets—are the same large language models behind popular AI chatbots. And there’s little you can really do with these devices you can’t already do with ChatGPT on your current phone. That makes the options in this new wave of hardware seem more like novelties than reliable travel assistants.
Thus far, the lesson we can draw from the many poorly received AI gadget launches? The likeliest future of travel, at least in the short term, is happening on devices like those already in your pocket. Here are some key takeaways from the current state of AI gadgets that highlight why phones are likely to remain the dominant platform.
1. The flopping Humane Ai Pin is on ice, literally
“Will anyone want to actually wear the pin? Or will it go the way of previously hyped devices such as Google Glass?” is the question I posed about the Humane pin in Condé Nast Traveler’s 2024 trend report, observing how the next phase of AI hype will push beyond trip-planning chatbots to real-world applications, with all the ethical and privacy concerns that raises. With slow sales and negative consensus among product critics, the answer to the question seems to be the latter.
Humane, one of the most widely promoted young AI start-ups, had grand ambitions with its cellular-and-GPS-enabled Ai Pin. It promised a clip-on virtual assistant the size of a bulky brooch that could accompany you anywhere it could connect to mobile data towers. (Though it was not marketed strictly as a travel device, the potential use cases for travelers, with its translation functionality, was immediately apparent.) The Humane pin had audio capabilities, a camera, and a unique laser projector, but no screen. It would be controlled through voice commands, using large language models to answer user requests. Among its bona fides: Humane even counted Sam Altman, OpenAI’s chief executive, among its investors.
But when it finally shipped units out this spring, reviewers almost universally dismissed it; Chokkattu at Wired rated it 4/10. Inaccurate or fabricated answers were a problem, of course, but the pin was also plagued with fundamental hardware issues like poor battery performance and a tendency to overheat. Humane executives apparently would put the device on ice packs to keep it from overheating when showing prospective investors, The New York Times reported in June amid news that Humane was in talks to sell the company.
What you can actually use on your travels: Google Translate
If you want to translate a menu, there’s already an app for that: Google Translate has long let you snap a photo. It’s also a good-enough translator if you’re really in a pinch, as you can speak what you want to say and have the app translate that out loud or in writing. The app recently added 110 languages, though speakers of some of those newly added languages, such as Venda and Swati in South Africa, have reported inaccuracies. (If a Star Trek–style universal translator is really your dream, you could also invest in already-established standalone translation devices like the business traveler’s favorite, Vasco Translator V4, which includes 108 languages and lifetime SIM access built-in.)
2. The Rabbit R1 is the walkie-talkie of AI
The latest device aspiring to be the so-called “iPhone of AI” is the pocket-size Rabbit R1. With its own camera and powered by Perplexity’s language model, the R1 promised similar travel benefits to the Humane Ai Pin. With its retro design by cult Swedish electronics company Teenage Engineering and its push-to-talk button, it looks more like the walkie-talkie of AI. Push the press-to-talk button to get the onboard system to, in theory, automate stuff you would normally have to do on your phone in your existing apps. Want a ride? Press the button, tell the R1 where you want to go, and it would order a car to your current location. But when the first devices shipped, the reality wasn’t so smooth.
The R1 had strikingly few app integrations at launch (limited to Uber, Yelp, DoorDash, automatic image generator Midjourney, and streaming apps Spotify and Apple Music). Thus, many early adopters have found its current capabilities frustratingly limited, making it about as useful as a walkie-talkie in most situations. For example, as Chokkattu wrote in Wired, the R1 often failed to do even simple tasks, like ordering a rideshare or getting suggestions for nearby coffee shops when you’re in a new-to-you neighborhood. Instead, Chokkattu found himself pulling out his phone to manually do those things in his trusted apps.
What you can actually use on your travels: Siri and Google Assistant
As clumsy as the current state of on-device voice assistants are, Siri’s Apple Maps integration, for instance, can help you find places nearby if you have cellular connectivity. Say “Show me nearby coffee shops that are open nearby,” and Siri will display a list and highlight when they’re closing soon. Google Assistant works similarly on Android devices, which can also help you call an Uber or Lyft rideshare using your voice.
3. The Ray-Bans by Meta are misidentifying zoo animals
When Meta launched its smart glasses in collaboration with Ray-Ban last October, they were primarily content-creation accessories: fast ways to snap pics for Instagram or capture a quick video for your story. But an update in April added a voice assistant to make more use of the onboard camera. Say “Hey, Meta, look and…” and the device will take a picture and then try to identify the subject, speaking to you through its open-ear speakers. But as clever as the glasses can appear, they’re still prone to confidently spouting inaccurate information, with reviewers reporting that the glasses would often misidentify animals in the zoo—referring to a monkey as a giraffe, for instance. (In other cases, the glasses helped reviewers do genuinely useful things, like make sense of a London Underground map using only a photo.)
But unlike the Humane Ai Pin or Rabbit R1, the Meta Ray-Bans are not standalone. They’re intended to pair with your phone, which seems to be a more honest pitch concerning what devices like this can currently achieve on their own: They’re better considered as novel toys to tinker with rather than any serious replacement for your phone.
What you can actually use on your travels: ChatGPT or, well, a local guide
You don’t need to go to the absurd level of wearing a voice-assistant-enabled camera on your spectacles to get context about what you’re looking at. If you really want to quiz a chatbot to see if it’ll tell you the truth, all the major ones from ChatGPT to Gemini have mobile apps where you can ask questions about photos taken with your phone. In late July, ChatGPT began rolling out its advanced voice capabilities.
Better yet, if you’d really like to get to know a destination, why not try something a bit more analog? Simply talk to a few locals or even connect with an expert guide.
In sum, the near future looks like it’s in our pockets already
Few new gadgets are perfect from the start. “There are growing pains with any type of first-gen product,” acknowledges Chokkattu of Wired. But if we’re paying for them, we should expect them to work, he says. Considering that many of these new devices remain plagued with glitches, he’s found the near future still looks a lot like the mobile devices we already own—whether they’re made by Apple or Samsung, or even the Motorola Hot Pink Razr, which is back with a Google Gemini infusion. “Overall, I just don’t think we’re anywhere near a future where the phone itself is going away.”
This is where Apple’s long-awaited announcements at its Worldwide Developers Conference on June 10 outlined a less splashy future, including an upgraded Siri, which will now be able to link up with ChatGPT due to a new partnership with OpenAI, and a host of automations and software features soon coming to iPhones. Perhaps, then, the most likely AI wearables are really your Siri-enabled AirPods and your smartwatch, which seem geared to get better at handling more complex prompts, such as live translation or giving you real-time context as you’re walking around an unfamiliar city. (Along similar lines, Google unveiled a series of announcements in May to bring its own competing AI features to mobile devices.)
How accurate and private will these new features prove to be—and how will they change familiar mobile interactions, which seem sure to evolve? Those are pressing questions. But when we want answers, we’re almost certain to start with our phones, our indefatigable travel companions that are already in our pockets—for better or for worse.
