Latest from WIRED
WIRED
20 hours ago
- Health
- WIRED
I Tried Hear.com's At-Home Test for New Prescription Hearing Aids. Here's How It Works
You no longer need to leave the house to be fitted with prescription hearing aids. All products featured on WIRED are independently selected by our editors. However, we may receive compensation from retailers and/or from purchases of products through these links. Putting aside their exorbitant cost, the trouble with prescription hearing aids is the prescription. Find a doctor. Get an appointment. Sit in a waiting room. Suffer through an hour of testing before getting the hard sell on a pair of hearing aids. Why, the indignity of it all is even worse than going deaf. Telehealth offers a glimmer of a solution, but as everyone who's tried to show off a concerning mole to a doctor via a smartphone camera knows: Video technology can only get you so far. Audiological testing requires significant one-on-one interaction and plenty of technology to back that up. Every audiogram I've sat for has been inside a sophisticated, soundproof chamber. company that sells prescription hearing aids online and via traditional retail channels —has a workaround. It puts all the technology needed to perform a full-on hearing test in a box and mails it to you along with a set of prescription hearing aids, then walks you through the whole thing with an audiologist via a video call. I've taken virtual hearing test twice—this is what a tele-hearing exam looks like, should you decide to take one. It's All Online The process begins with a simple discussion. If you've never had hearing aids before, this is probably a good first step; veterans will likely skip it, especially if they know what model hearing aids they want (or how much they want to spend). A salesperson will discuss your personal impressions of your hearing loss, any hearing aid devices you have tried, prior audiological testing, and your budget (along with any insurance you have). From there, they will suggest the best hearing aids (from their product line) to meet your needs, though you can also advocate for a specific product if you've already identified one. Next, you'll set a time for your testing and hearing aid fitting, and a few days later, a box will arrive on your doorstep. Open it up, and the collection of hardware and the tangle of wires connecting it all may seem daunting, but rest assured, with even a modicum of technical knowledge, it all goes together quickly, and professionals can guide you through any confusion over the phone. Despite the significant amount of gear inside, the box is about as organized as it can get. On top, you'll find a Microsoft Surface tablet that has been stripped down to do just one thing: serve as the conduit for a video-based audiology session. After plugging the Surface into wall power, you'll find a USB breakout box to connect to the Surface's USB port. Assuming all the other devices are still properly plugged into that USB breakout box (which is far from assured; it's good to check everything is snug), you are effectively good to go. The Surface has no other functionality besides initiating the call with the audiology professional, and at the scheduled time you tap a button on the screen to get the session underway. After greeting you, the audiologist will guide you through a series of tests. First is a physical exam of your ears, which is made possible thanks to the included USB-powered otoscope, which you probably know better as the lighted tool with the black, conical tip a doctor uses to look into your ear canal. There's no doctor to guide the tool by hand, so the audiologist will tell you how to position it in each ear so they can get a good look at your eardrum. And yes, you, too, can see what the doctor sees right on the Surface's screen. Spoiler: There is more wax inside your ear canal than you think. Can You Hear Me Now? Assuming you have no physical damage that would preclude hearing aid use, it's on to the traditional hearing exams. This series of tests will consume the bulk of the appointment time, and most involve wearing a pair of oversize, noise-isolating headsets like professional musicians use. The tests were very familiar to me, starting with tones played at varying frequencies and volume levels in each ear, during which you tap the Surface screen whenever you hear something. The test is then repeated with a bone-conduction appliance that attaches to your forehead with a Velcro band for another look at the way you process sound. Lastly, a third test measures your ability to distinguish various consonant and vowel sounds— sh vs. ch , for example—which is performed by having you repeat the words you think you hear back to the audiologist multiple times. From all of this data, an audiogram is synthesized and presented on the screen of the Surface. Quality-wise, my audiogram was nearly identical to the last professional audiogram I received in 2023 in an office setting. audiogram was even more detailed, because it tested my hearing at nine different frequencies, one more than the in-person visit. After going over the audiogram with me to discuss the particulars of my hearing loss, audiologist moved on to the hearing aid fitting. You'll have purchased the hearing aids you're about to put on already and will find them in the box, inside a charger that notes they are fully charged and ready to use. If for some reason these aids aren't appropriate for the hearing loss measured in the testing, will recommend an alternative at that time, but most customers should have a set in hand that will meet their needs, based on the previous intake call. The aids are programmed remotely and wirelessly via another device connected to the USB breakout box, and I was directed to put the hearing aids on and then put them back in the charging case repeatedly so we could tweak the sound profile to my liking. Again, this is a standard process identical to how things would be done in a doctor's office. The doctor will then train a hearing aid novice on how to insert, charge, and clean the aids and go over any lingering paperwork that needs to be dealt with. sells its own hearing aids under the Horizon brand, which are private-labeled Signia hearing aids in the Signia IX series. There's not an easy way to see exactly which models offers on its website, but they range from the Horizon 1IX ($1,975) to the Horizon 7IX ($4,950), the latter of which is what I was sent to try. Monthly payment plans are available for everything, and all of its hearing aids come with a 45-day money-back guarantee. When the hour-long session is up, you'll be directed to keep all the testing hardware for a while until you are certain the aids work for you. A follow-up call a few weeks later will check on your progress (there's a learning curve with all hearing aids) and, once you're satisfied, you'll be instructed to return the testing equipment via a prepaid shipping label in the box. I'll need to continue testing for longer, but so far, the Horizon hearing aids work great. Even better is the way you go about getting them fitted.
WIRED
21 hours ago
- Politics
- WIRED
How Democrats Are Meeting (and Missing) the Moment
Jun 19, 2025 7:00 AM After the shootings of lawmakers in Minnesota, the weekend's "No Kings" protests, and elected officials being placed in handcuffs, Democrats are searching for the path forward. Photo-illustration: WIRED Staff; Getty Images 'Well,' a senior Democratic strategist tells me, 'my wife and I are having a fight about me going back to the field in the midterms.' 'I would be lying if I said these aren't conversations I've been having with my family every day,' shared a Democratic candidate in a high-profile battleground state midterms race. In a matter of days, the truly unthinkable occurred: On Thursday last week, Senator Alex Padilla of California was forcibly removed from a news conference and handcuffed by officials wearing FBI-identifying clothing when he tried to question Homeland Security Secretary Kristi Noem on immigration raids in Los Angeles. On Saturday, Melissa Hortman, a Democratic Minnesota state representative, and her husband, Mark Hortman, were shot and killed. State senator John Hoffman and his wife Yvette Hoffman were also shot. On Tuesday, New York City Comptroller and Democratic mayoral candidate Brad Lander was violently arrested after accompanying an immigrant from a court appointment. Not to mention the fact that an estimated 4 to 6 million people attended 'No Kings' protests this past weekend against President Donald Trump's administration. How should Democrats be approaching this chaotic and, at times, violent moment? From lawmakers and candidates all the way down to the staff who fuel their operations, a series of events over the past week have made national Democrats reconsider their strategy. Following the shootings of the state lawmakers in Minnesota, a weekend of nationwide protests, and Democratic lawmakers in handcuffs, Democrats are trying to figure out what meeting the Trump moment really means for an out-of-power coalition that's already under fire from constituents for ostensibly rolling over. And for a number of lawmakers and strategists, that answer isn't so simple. 'I think that the hard reality is, a lot of Democrats and a lot of people say, like, why aren't we doing something?' David Axelrod, former President Barack Obama's chief strategist from his 2008 and 2012 campaigns, tells WIRED. 'Well, there are a limited number of ways in a democracy to do something, short of revolution—which I'm not advocating. You win elections, you go to court.' He says Padilla was in the right but that Democrats should exercise caution and not play into the White House's hands by rushing to get arrested or associating themselves too closely with protests that are unpredictable and prone to producing images of disorder. (Padilla's office did not return a request for comment.) Some laws of political gravity, Axelrod argues, still apply: 'As a general rule, don't do what your opponent wants.' The issue isn't just optics though, but expectation. 'I think the last thing Dems want to do is look unhinged or out of control,' a GOP strategist tells me. Right-wing media, certainly, has had a field day: a Fox News prime time segment from Jesse Watters on Tuesday night painted Democratic protesters as hysterical and ineffective. But Democratic voters might just want a strong, 'unhinged' response. 'Democrats are crying out for [a] fight,' the senior Democratic strategist tells me. 'They are crying out for someone who seems like they're fighting.' Democrats have been deeply pessimistic about the future of their own party, according to polling, despite a recent rebound for the party brand in the generic ballot against Republicans, where they now lead by 2 points, their best performance since last August, according to YouGov's latest survey. Pennsylvania state representative Malcolm Kenyatta—a newly elected vice chair of the Democratic National Committee since David Hogg stepped down—tells WIRED optics should not be the primary consideration when it comes to whether lawmakers with oversight abilities should consider risking the chance of an arrest. 'Listen, I think we have a responsibility to do our fuckin' job,' says Kenyatta, the grandson of civil rights leader Muhammad Kenyatta. 'And what Senator Padilla was doing was his job.' House Speaker Hakeem Jeffries of New York has often used the baseball analogy of not swinging at every pitch to explain how he thinks Democrats should respond to President Trump's deluge of executive actions and outlandish statements. For emerging leaders in the party like Kenyatta, the state of play has moved beyond that point. 'Things don't become less important because a lot of them are happening,' Kenyatta says. 'I don't think we have the luxury of ignoring any of the things he does that are going to make life worse for people.' Lander, who was detained outside of an immigration court hearing in New York City on Tuesday, took that exact type of risk. For the senior Democratic strategist fighting with his wife—who requested anonymity to avoid entangling other clients, and their spouse—Lander showed fellow Democrats how it should be done. (A representative for Lander did not return a request for comment.) 'Yes, this is exactly what people should be doing,' they say. 'There are people who are going to say this whole thing is gimmicky, and I get that. Put some fucking skin in the game. We've gotta draw attention to it somehow—look at all of those cameras they had to arrest him in front of.' But there are also very real security considerations. The battleground state candidate, who also requested anonymity to discuss sensitive security conversations, said the increased security posture lawmakers are dealing with has only made it harder to be as up close with voters, as regularly, in as many places. 'It's this ugly reality where you have to look out for your safety. It is going to rob people of access to their elected officials and candidates,' they say. This has happened already: Representative Hillary Scholten of Michigan postponed a Monday town hall in Muskegon after her name ended up on the alleged Minnesota shooter's list of lawmakers to target. (Scholten's office did not return a request for comment.) Even so, the Democratic candidate wants their constituents to know they're not alone, and they have every right to be angry. "I am as angry as you are about our government and our elected officials,' they say. 'I'm doing the hard thing. I'm putting my name on a ballot. I'm being vulnerable. I'm asking for people to support me, which is a fuck-ton harder than buying a gun and going and threatening people." The Chatroom Who stands out to you as a lawmaker or candidate meeting the moment? Leave a comment on the site or send your thoughts to mail@ WIRED Reads Want more? Subscribe now for unlimited access to WIRED. What Else We're Reading 🔗 Inside the Clashes Between Trump and Gabbard: Tulsi Gabbard is in a bind after more than two decades of pushing for an end to forever wars, and a rogue 3-minute video she recently released didn't help. (Politico) 🔗 Comptroller Brad Lander Arrested by Masked Federal Agents While Escorting Immigrant from Court: A comprehensive write-up of how Brad Lander's arrest went down, from an excellent local outlet in New York. (Hell Gate) 🔗 Bid to Protect Lawmakers' Data Gains Momentum: The Minnesota shooter's alleged use of data brokers has revived a bill in Congress that would allow lawmakers to remove personal data from the internet. (Semafor) The Download Our flagship show Uncanny Valley dives into some of the strange and disturbing hallucinations from AI chatbots around the Los Angeles protests—the same ones that would come up blank when users asked for information about the 2024 election. Listen now. Thanks again for subscribing. You can find me on Bluesky or on Signal at Leak2Lahut.26.
WIRED
21 hours ago
- WIRED
7 Ways to Limit Your Endless Doomscrolling
Jun 19, 2025 7:00 AM It's easier than ever to step away from the smartphone with these tips. Photograph:If you actually measure the amount of time you spend scrolling through algorithm-led social media feeds each day, it will probably add up to more hours and minutes than you'd guess. Time can quickly slip away while you're checking up on friends, celebrities, the news, and the viral memes of the day. The term doomscrolling has sprung up to describe this behavior, which, let's face it, isn't usually the most edifying or the best for our mood. It's no surprise that the word began to gain traction at the start of the Covid-19 pandemic, a period of lockdowns and social isolation that we're still feeling the effects of. There are ways to fight back against doomscrolling though, some of which are built into your phone and can be enabled with a couple of taps. See how much time and mental well-being they can help you reclaim. (The Android instructions below apply to Pixel phones running Android 15 or later. If you're using something different, you should be able to find similar features and settings, though they might not be in exactly the same place). Turn Off Notifications News and social media alerts can drag you into apps when you weren't even thinking about checking them. Disabling notifications can keep you from getting sucked in. You can shut off notifications altogether or just quieten them. On Android, head to Settings, then tap Notifications > App notifications. On iOS, from Settings choose Notifications to see an app list. Delete Apps For a more extreme approach, you can delete these apps from your mobile devices entirely, which stops you from checking them at all. On Android, you can drag an app from the app drawer up to the Uninstall button at the top of the screen. On iOS, tap and hold on an app in the app library, then choose Delete App from the pop-up menu. Deleting apps on an iPhone Courtesy of David Nield This may just lead you to check these apps more often on the web—but you can always delete your accounts entirely. This is usually not too difficult to do: In the case of X, for example, head to the website and log in, then click More > Settings and privacy > Your account > Deactivate your account and follow the instructions. Put Your Phone in Another Room You could consider putting your phone in a different room during parts of the day—especially at bedtime, to avoid doomscrolling when you should be sleeping. Note that several studies have shown that just having a phone in the same room with you can have an impact on your attention and focus, even if you don't actually pick it up and use it. Set Screen Time Limits Modern phones come with built-in tools that allow you to limit your screen time, so you can put restrictions on how long you use specific apps and how long you're on your phone overall. Of course, you're able to disable these limits as easily as you can enable them—it's your phone after all—but you can use them in combination with your own willpower to change your habits. If you're on Android, you can open Settings and choose Digital Wellbeing and Parental Controls, then App limits, to put some restrictions in place. Over on iOS, from Settings tap Screen Time, then App & Website Activity. On both platforms, you get a pretty detailed breakdown of how you've been spending your time on your phone. Scroll Something Else If you must spend time on your phone (what else are you going to do on the subway platform?), you can replace social media and news with apps like Calm for mindful meditation; Blinklist for reading summaries of books, podcasts, and talks; and Kindle for ebook reading, which will also be synced to your actual Kindle, if you have one. Tell Your Phone It's Bedtime Enabling a bedtime schedule on Android Courtesy of David Nield Both Android and iOS now let you set a specific bedtime on your phone, the idea being that it helps you wind down properly by steering you away from opening apps and doomscrolling into the early hours. These bedtime modes put limits on incoming notifications, dim the screen, and give you the option of using a more simplified phone interface late at night. If you're on Android, open the Clock app and switch to the Bedtime tab: You can set start and end times for the mode and access settings such as turning the screen gray at a specific hour. On iPhone, head to the main iOS Settings screen, then tap Focus, then Sleep. The next screen will show all the available options and timers. Get a Reminder to Stop This is iPhone only, but Adam Davidson over at How-To Geek has a neat idea: Create an iOS shortcut to generate a snarky message from ChatGPT or Claude AI that warns about the harm of doomscrolling, and have it go off every time you open certain apps. The message can be shown on screen or read aloud.
WIRED
21 hours ago
- Entertainment
- WIRED
Dosa Divas Is a ‘Spicy' New Game About Fighting Capitalism With Food
As protestors faced off against ICE agents and law enforcement in Los Angeles, game developers gathered just a few miles away for Summer Game Fest. They'd come to the annual show in early June to show off their games and make professional connections. The team at Outerloop Games was no different. But the studio, led and staffed by brown and Black developers, many of whom are immigrants themselves, had extra precautions to consider: 'In case of an ICE raid,' says game director Chandana Ekanayake of the team's plan at the time, 'we're gonna stay together.' Whether that meant at the event itself, or even dinner after, the team kept track of each other for the entire event. Outerloop, creator of Thirsty Suitors and Falcon Age, were at the event to show off their newest project, a 'spicy' narrative turn-based RPG called Dosa Divas due out in early 2026. As the Trump administration carries out mass deportations, targets visas, disappears migrants into a foreign megaprison, and tweets 'ASMR' videos of people being deported in chains from the official White House account, Outerloop is making games that find ways to connect people to different cultures through food. 'That's the most accepted version of ourselves or of culture, is food,' Ekanayake says. 'People are definitely willing to try food before they're willing to accept the people that make it.' Outerloop's games have always featured food in some capacity, and Dosa Divas only ramps up its importance. Named for dosas, savory crepes popular in South India, the game stars two sisters fighting against an evil fast food empire, literally; characters in its turn-based combat have unique abilities associated with different 'flavor profiles,' like sweet, spicy, or sour. To attack effectively, you need to match moves to the enemy's flavor craving. Dosa Divas centers around three sisters—Samara, Amani, and fast food entrepreneur, Lina—the last of which has become estranged after their family restaurant closes down. Lina's quick cuisine has become so widespread that nobody really cooks anymore. Ekanayake says that part of the game's story is about reconciliation and reconnection, especially through food. Collecting ingredients and cooking plays an important role in the game. By sharing meals with villagers, players can help repair communities and build their own reputation.
WIRED
a day ago
- Business
- WIRED
How Much Energy Does AI Use? The People Who Know Aren't Saying
Jun 19, 2025 6:00 AM A growing body of research attempts to put a number on energy use and AI—even as the companies behind the most popular models keep their carbon emissions a secret. Photograph: Bloomberg/Getty Images 'People are often curious about how much energy a ChatGPT query uses,' Sam Altman, the CEO of OpenAI, wrote in an aside in a long blog post last week. The average query, Altman wrote, uses 0.34 watt-hours of energy: 'About what an oven would use in a little over one second, or a high-efficiency lightbulb would use in a couple of minutes.' For a company with 800 million weekly active users (and growing), the question of how much energy all these searches are using is becoming an increasingly pressing one. But experts say Altman's figure doesn't mean much without much more public context from OpenAI about how it arrived at this calculation—including the definition of what an 'average' query is, whether or not it includes image generation, and whether or not Altman is including additional energy use, like from training AI models and cooling OpenAI's servers. As a result, Sasha Luccioni, the climate lead at AI company Hugging Face, doesn't put too much stock in Altman's number. 'He could have pulled that out of his ass,' she says. (OpenAI did not respond to a request for more information about how it arrived at this number.) As AI takes over our lives, it's also promising to transform our energy systems, supercharging carbon emissions right as we're trying to fight climate change. Now, a new and growing body of research is attempting to put hard numbers on just how much carbon we're actually emitting with all of our AI use. This effort is complicated by the fact that major players like OpenAi disclose little environmental information. An analysis submitted for peer review this week by Luccioni and three other authors looks at the need for more environmental transparency in AI models. In Luccioni's new analysis, she and her colleagues use data from OpenRouter, a leaderboard of large language model (LLM) traffic, to find that 84 percent of LLM use in May 2025 was for models with zero environmental disclosure. That means that consumers are overwhelmingly choosing models with completely unknown environmental impacts. 'It blows my mind that you can buy a car and know how many miles per gallon it consumes, yet we use all these AI tools every day and we have absolutely no efficiency metrics, emissions factors, nothing,' Luccioni says. 'It's not mandated, it's not regulatory. Given where we are with the climate crisis, it should be top of the agenda for regulators everywhere.' As a result of this lack of transparency, Luccioni says, the public is being exposed to estimates that make no sense but which are taken as gospel. You may have heard, for instance, that the average ChatGPT request takes 10 times as much energy as the average Google search. Luccioni and her colleagues track down this claim to a public remark that John Hennessy, the chairman of Alphabet, the parent company of Google, made in 2023. A claim made by a board member from one company (Google) about the product of another company to which he has no relation (OpenAI) is tenuous at best—yet, Luccioni's analysis finds, this figure has been repeated again and again in press and policy reports. (As I was writing this piece, I got a pitch with this exact statistic.) 'People have taken an off-the-cuff remark and turned it into an actual statistic that's informing policy and the way people look at these things,' Luccioni says. 'The real core issue is that we have no numbers. So even the back-of-the-napkin calculations that people can find, they tend to take them as the gold standard, but that's not the case.' One way to try and take a peek behind the curtain for more accurate information is to work with open source models. Some tech giants, including OpenAI and Anthropic, keep their models proprietary—meaning outside researchers can't independently verify their energy use. But other companies make some parts of their models publicly available, allowing researchers to more accurately gauge their emissions. A study published Thursday in the journal Frontiers of Communication evaluated 14 open-source large language models, including two Meta Llama models and three DeepSeek models, and found that some used as much as 50 percent more energy than other models in the dataset responding to prompts from the researchers. The 1,000 benchmark prompts submitted to the LLMs included questions on topics such as high school history and philosophy; half of the questions were formatted as multiple choice, with only one-word answers available, while half were submitted as open prompts, allowing for a freer format and longer answers. Reasoning models, the researchers found, generated far more thinking tokens—measures of internal reasoning generated in the model while producing its answer, which are a hallmark of more energy use—than more concise models. These models, perhaps unsurprisingly, were also more accurate with complex topics. (They also had trouble with brevity: During the multiple choice phase, for instance, the more complex models would often return answers with multiple tokens, despite explicit instructions to only answer from the range of options provided.) Maximilian Dauner, a PhD student at the Munich University of Applied Sciences and the study's lead author, says he hopes AI use will evolve to think about how to more efficiently use less-energy-intensive models for different queries. He envisions a process where smaller, simpler questions are automatically directed to less-energy-intensive models that will still provide accurate answers. 'Even smaller models can achieve really good results on simpler tasks, and don't have that huge amount of CO 2 emitted during the process,' he says. Some tech companies already do this. Google and Microsoft have previously told WIRED that their search features use smaller models when possible, which can also mean faster responses for users. But generally, model providers have done little to nudge users toward using less energy. How quickly a model answers a question, for instance, has a big impact on its energy use—but that's not explained when AI products are presented to users, says Noman Bashir, the Computing & Climate Impact Fellow at MIT's Climate and Sustainability Consortium. 'The goal is to provide all of this inference the quickest way possible so that you don't leave their platform,' he says. 'If ChatGPT suddenly starts giving you a response after five minutes, you will go to some other tool that is giving you an immediate response.' However, there's a myriad of other considerations to take into account when calculating the energy use of complex AI queries, because it's not just theoretical—the conditions under which queries are actually run out in the real world matter. Bashir points out that physical hardware makes a difference when calculating emissions. Dauner ran his experiments on an Nvidia A100 GPU, but Nvidia's H100 GPU—which was specially designed for AI workloads, and which, according to the company, is becoming increasingly popular—is much more energy-intensive. Physical infrastructure also makes a difference when talking about emissions. Large data centers need cooling systems, light, and networking equipment, which all add on more energy; they often run in diurnal cycles, taking a break at night when queries are lower. They are also hooked up to different types of grids—ones overwhelmingly powered by fossil fuels, versus those powered by renewables—depending on their locations. Bashir compares studies that look at emissions from AI queries without factoring in data center needs to lifting up a car, hitting the gas, and counting revolutions of a wheel as a way of doing a fuel-efficiency test. 'You're not taking into account the fact that this wheel has to carry the car and the passenger,' he says. Perhaps most crucially for our understanding of AI's emissions, open source models like the ones Dauner used in his study represent a fraction of the AI models used by consumers today. Training a model and updating deployed models takes a massive amount of energy—figures that many big companies keep secret. It's unclear, for example, whether the light bulb statistic about ChatGPT from OpenAI's Altman takes into account all the energy used to train the models powering the chatbot. Without more disclosure, the public is simply missing much of the information needed to start understanding just how much this technology is impacting the planet. 'If I had a magic wand, I would make it mandatory for any company putting an AI system into production, anywhere, around the world, in any application, to disclose carbon numbers,' Luccioni says. Paresh Dave contributed reporting.
