I test smart air conditioners for a living — here's my top 3 picks to make it through a heat wave
Not surprisingly, many of us are cranking up our ACs to try and stay comfortable — and in the process, finding out the units we have no longer work, or are not up for the task.
To keep our guide to the best smart window air conditioners up to date, I've been testing window air conditioners for the past five years. Not only have they been keeping my house cool, but they've also been saving me money, as they're a lot more efficient than older models.
In addition to cooling your home, these ACs can also be controlled via an app on your smart phone, or connected to Alexa or Google Assistant — giving them their "smarts." This way, you can program them to turn on or adjust their temperature when you're arriving home, so you're not cooling your house while you're not there.
If you're in the market for a new air conditioner, here are my top three picks.
I've been using the LG Dual Inverter air conditioner now for seven years, and it's still performing like a champ. The first floor of my house isn't overly large (around 600-700 square feet, give or take), but is divided into three rooms.
The 10,000 BTU model has been more than up to the task, keeping things cool throughout the summer. And, it's very quiet, too.
Get instant access to breaking news, the hottest reviews, great deals and helpful tips.
The only caveat is that this particular unit weighs around 63 pounds; I bought a bracket (sold separately) to make it easier to install in my window.
This air conditioner is available in several sizes: 6000, 8,500, 10,000, 12,000, and 14,000 BTUs. Look for the Dual Inverter model. It works with Alexa, Google Home, and has both Android and iOS apps.
While the Midea U was recently the subject of a massive recall, I'm still recommending this model for its innovative design (the company has issued a fix, as well as a redesigned model).
Instead of those ugly and ineffective baffles, you slide your window between the two halves of the Midea, which not only offers better insulation, but cuts down on noise and lets more outside light in.
The Midea U is also really quiet, and does an amazing job of cooling down my attic office, which is the hottest room in the house.
The Midea U is available in 8,000, 10,000, and 12,000 BTU sizes, and two of the models can also work as heat pumps, to keep your room warm in the winter.
While I didn't test this personally, my colleague Kelly Woo did, and it was perfect for her apartment. Windmill doesn't offer as many sizes as the competition, but it is one of the nicest-looking air conditioners we've tested.
We especially like its wrap-around shroud, which looks a sight better than the traditional accordion-like baffles found on most window AC units. If you want the models with the best energy efficiency — and the ones that run the quietest — look for the models with "WhisperTech."
One of the nicest-looking window air conditioners around. Keep in mind that only the models marked "WhisperTech" use inverter technology.
There's one thing that my favorite window air conditioners all have in common. They all use what's called inverter technology, which is far more efficient than compressor technology, which was the traditional method that air conditioners used.
You can determine how efficient an air conditioner is by looking at its SEER (Seasonal Energy Efficiency Rating). The best models will have ratings between 13 and 15 or higher. Basically, this means they'll use less energy to cool your home.
A side benefit of inverter technology is that it's a lot quieter, so you can still sleep comfortably without the racket of an air conditioner keeping you awake at night.
Having a working AC is just one way to keep your house cool and not spend a fortune on your electric bill during the hottest time of the year. Here are 10 tips to cool your room down in a heatwave.
Did you know, for instance, that you should set your thermostat to 78 degrees Fahrenheit? You should also try the "caveman method" to keep your house cool. And one of our writers beat the heat with a $2 roll of aluminum foil.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Tom's Guide
11 hours ago
- Tom's Guide
Google Calendar bug uses Gemini to take over smart home devices and steal user data
Researchers have found a flaw that allows malicious Google Calendar invites to hijack Gemini in order to wreak havoc on a target's machine. As reported by Bleeping Computer, a maliciously crafted invite within Google Calendar can remotely take over Gemini agents without any user involvement beyond typical day-to-day interaction with the assistant. The security researchers at SafeBreach, who demonstrated this attack in a report, were able to send a calendar invite with an embedded prompt injection, hidden in the event title, which permitted them to exfiltrate a variety of user data like email content and Calendar information. They were also able to track the victim's location, control smart home devices (using Google Home) open apps on Android and trigger Zoom calls. The researchers made note that the attack did not require white-box model access and was not blocked by Gemini's protection measures or by prompt filtering. Instead, the attack begins with a malicious Google Calendar event invite sent to the victim which includes an event title containing an indirect prompt injection. The victim then only needs to interact with Gemini as they typically would, such as asking 'What are my calendar events today?' in order to cause the AI chatbot to pull a list of events from the Calendar – which will include the malicious event title embedded by the attacker. This will then becomes part of Gemini's content window, and the assistant will treat it as part of the conversation as it is unable to realize that the instruction is malicious. Depending on what the instruction is, it could cause lead to a number of different prompts from being executed, causing events in Google Calendar to be edited or removed entirely, opening URLs to retrieve the victim's IP address, joining a Zoom call, using Google Home to control devices, or accessing emails and leaking user data. However, it could take up to six calendar invites for this attack to work with the malicious prompt being included only in the last invite. This is because the Calendar events section displays only the five most recent events; the rest fall under the 'Show more' button. Gemini will parse them all – including the malicious one – when instructed to. Additionally, the victim will not see the malicious event title or realize there has been a compromise unless they expand the events list by clicking 'Show more.' Gemini, Google's LLM (large language model) assistant, is integrated into Android, Google web services and Google's Workspace apps so it has access to Gmail, Calendar and Google Home. These attacks are a downside of Google's broad access and reach, and while its usefulness comes from its ability to reach across tools, this is also proving to be a detriment when it comes to the nature of this attack. Google has already issued a fix and has credited the team of researchers and their efforts. Get instant access to breaking news, the hottest reviews, great deals and helpful tips. Follow Tom's Guide on Google News to get our up-to-date news, how-tos, and reviews in your feeds. Make sure to click the Follow button.
Tom's Guide
15 hours ago
- Tom's Guide
From hyper-personal assistants to mind-reading tech — this is how AI will transform everything by 2035
Artificial Intelligence | Smart Glasses | Wearable TechSmartphones | iPhones | Robots | Cars | TVs Picture a morning in 2035. Your AI assistant adjusts the lights based on your mood, reschedules your first meeting, reminds your child to take allergy medicine; all without a prompt. It's not science fiction, it's a likely reality driven by breakthroughs in ambient computing, emotional intelligence and agentic AI. Just five years ago, ChatGPT was an unfamiliar name to most, let alone a daily assistant for summarization, search, reasoning and problem-solving. Siri and Alexa were the top names that came to mind when we wanted to call a friend, place an order or dim the lights. Yet now, in 2025, we have a plethora of AI assistants and chatbots to choose from, many of which are free, and which can do a lot more than controlling smart home devices. What feels advanced now may seem utterly simplistic in a decade, reminding us that the most mind-blowing AI capabilities of 2035 might still be beyond our current imagination. By 2035, your AI assistant won't just respond — it will anticipate. This evolution marks the rise of agentic AI, where assistants proactively act on your behalf using predictive analytics, long-term memory and emotion-sensing. These systems can forecast your needs by analyzing historical and real-time data, helping stay one step ahead of your requests. 'Alexa will be able to proactively anticipate needs based on patterns, preferences, and context — preparing your home before you arrive, suggesting adjustments to your calendar when conflicts arise, handling routine tasks before you even ask.' One assistant that's undergoing such a change is Amazon's Alexa. According to Daniel Rausch, Amazon's VP of Alexa and Echo, 'Alexa will be able to proactively anticipate needs based on patterns, preferences, and context — preparing your home before you arrive, suggesting adjustments to your calendar when conflicts arise, or handling routine tasks before you even think to ask.' The AI will remember your child's travel soccer team schedule, reschedule your meetings when it detects stress in your voice and even dim your AR glasses when you appear fatigued. 'By 2035, AI won't feel like a tool you 'use',' Rutgers professor Ahmed Elgammal says. 'It'll be more like electricity or Wi-Fi: always there, always working in the background.' And AIs will respond to more than just your speech. Chris Ullrich, CTO of Cognixion, a Santa Barbara based tech company, is currently developing a suite of AI-powered Assisted Reality AR applications that can be controlled with your mind, your eyes, your head pose, and combinations of these input methods. 'We strongly believe that agent technologies, augmented reality and biosensing technologies are the foundation for a new kind of human-computer interaction,' he says. AI in 2035 will see, hear and sense — offering real-time support tailored to you. With multimodal capabilities, assistants will blend voice, video, text and sensor inputs to understand emotion, behavior and environment. This will create a form of digital empathy. Ullrich notes that these advanced inputs shouldn't aim to replicate human senses, but exceed them. 'In many ways, it's easier to provide superhuman situational awareness with multimodal sensing,' he says. 'With biosensing, real-time tracking of heart rate, eye muscle activation and brain state are all very doable today.' Amazon is already building toward this future. 'Our Echo devices with cameras can use visual information to enhance interactions,' says Rausch. 'For example, determining if someone is facing the screen and speaking enables a more natural conversation without them having to repeat the wake word.' In addition to visual cues, Alexa+ can now pick up on tone and sentiment. 'She can recognize if you're excited or using sarcasm and then adapt her response accordingly,' Rausch says — a step toward the emotionally intelligent systems we expect by 2035. Memory is the foundation of personalization. Most AI today forgets you between sessions. In 2035, contextual AI systems will maintain editable, long-term memory. Codiant, a software company focused on AI development and digital innovation, calls this 'hyper-personalization,' where assistants learn your routines and adjust suggestions based on history and emotional triggers. Rather than relying on one general assistant, you'll manage a suite of specialized AI agents. Research into agentic LLMs shows orchestration layers coordinating multiple AIs; each handling domains like finance, health, scheduling or family planning. These assistants will work together, handling multifaceted tasks in the background. One might track health metrics while another schedules meetings based on your peak focus hours. The coordination will be seamless, mimicking human teams but with the efficiency of machines. Ullrich believes the biggest breakthroughs will come from solving the 'interaction layer,' where user intent meets intelligent response. 'Our focus is on generating breakthroughs at the interaction layer. This is where all these cutting-edge technologies converge,' he explains. Rausch echoes this multi-agent future. 'We believe the future will include a world of specialized AI agents, each with particular expertise,' he says. 'Alexa is positioned as a central orchestrator that can coordinate across specialized agents to accomplish complex tasks.' He continues, 'We've already been building a framework for interoperability between agents with our multi-agent SDK. Alexa would determine when to deploy specialized agents for particular tasks, facilitating communication between them, and bringing their capabilities together into experiences that should feel seamless to the end customer.' Perhaps the most profound shift will be emotional intelligence. Assistants won't just organize your day, they'll help you regulate your mood. They'll notice tension in your voice, anxiety in your posture and suggest music, lighting or a walk. 'Users need to always feel that they're getting tangible value from these systems and that it's not just introducing a different and potentially more frustrating and opaque interface.' Ullrich sees emotion detection as an innovation frontier. 'I think we're not far at all from effective emotion detection,' he says. 'This will enable delight — which should always be a key goal for HMI.' He also envisions clinical uses, including mental health care, where AI could offer more objective insights into emotional well-being. But with greater insight comes greater responsibility. Explainable AI (XAI), as described by arXiv and IBM, will be critical. Users must understand how decisions are made. VeraSafe, a leader in privacy law, data protection, and cybersecurity, underscores privacy concerns like data control and unauthorized use. 'Users need to always feel that they're getting tangible value from these systems and that it's not just introducing a different and potentially more frustrating and opaque interface,' Ullrich says. That emotional intelligence must be paired with ethical transparency, something Rausch insists remains central to Amazon's mission: 'Our approach to trust doesn't change with new technologies or capabilities, we design all of our products to protect our customers' privacy and provide them with transparency and control.' He adds, 'We'll continue to double down on resources that are easy to find and easy to use, like the Alexa Privacy Dashboard and the Alexa Privacy Hub, so that deeper personalization is a trusted experience that customers will love using.' AI may replace jobs, but more so, it will reshape them. An OECD study from 2023 reports that 27% of current roles face high automation risk, especially in repetitive rules-based work. An even more recent Microsoft study highlighted 40 jobs that are most likely to be affected by AI. Human-centric fields like education, healthcare, counseling and creative direction will thrive, driven by empathy, ethics and original thinking. Emerging hybrid roles will include AI interaction designers and orchestrators of multi-agent systems. Writers will co-create with AI, doctors will pair AI with human care and entrepreneurs will scale faster than ever using AI-enhanced tools. AI becomes an amplifier, not a replacement, for human ingenuity. Even the boundaries between work and home will blur. 'While Alexa+ may be primarily focused on home and personal use today, we're already hearing from customers who want to use it professionally as well,' says Rausch. 'Alexa can manage your calendar, schedule meetings, send texts and extract information from documents — all capabilities that can bridge personal and professional environments.' AI becomes an amplifier, not a replacement, for human ingenuity. A 2025 study from the University of Pennsylvania and OpenAI found that 80% of U.S. workers could see at least 10% of their tasks impacted by AI tools, and nearly 1 in 5 jobs could see more than half their duties automated with today's AI. Forbes reported layoffs rippling across major companies like marketing, legal services, journalism and customer service as generative AI takes on tasks once handled by entire teams. Yet the outlook is not entirely grim. As the New York Times reports, AI is also creating entirely new jobs, including: Automation Alley's vision of a 'new artisan' is gaining traction. As AI lifts mental drudgery, skilled manual work — craftsmanship, artistry and hands-on innovation — may see a renaissance. AI won't kill creativity; it may just unlock deeper levels of it. Navigating the shift to an AI-augmented society demands preparation. The World Economic Forum emphasizes lifelong learning, UBI (universal basic income) experimentation and education reform. Workers must develop both technical and emotional skills. Curricula must evolve to teach AI collaboration, critical thinking and data literacy. Social safety nets may be required during reskilling or displacement. Ethics and governance must be built into AI design from the start, not added after harm occurs. Ultimately, the question isn't 'What can AI do?' It's 'What should we let AI do?' Ullrich notes the importance of designing with inclusivity in mind. 'By solving the hard design problems associated with doing this in the accessibility space, we will create solutions that benefit all users,' he says. Technologies developed for accessibility, like subtitles or eye tracking—often lead to mainstream breakthroughs. As IBM and VeraSafe highlight, trust hinges on explainability, auditability and data ownership. Public understanding and control are key to avoiding backlash and ensuring equitable access. As AI augments more aspects of life, our relationship with it will define the outcomes. Daniel Rausch believes the key lies in meaningful connection: 'The goal isn't just responding to commands but understanding your life and meaningfully supporting it.' We must ensure systems are inclusive, transparent and designed for real value. As AI grows in intelligence, the human role must remain centered on judgment, empathy and creativity. Ultimately, the question isn't 'What can AI do?' It's 'What should we let AI do?' By 2035, AI will be a planner, therapist, tutor and teammate. But it will also reflect what we value — and how we choose to interact with it. Ullrich emphasizes that the future won't be defined just by what AI can do for us, but how we engage with it: 'Voice may be useful in some situations, gesture in others, but solutions that leverage neural sensing and agent-assisted interaction will provide precision, privacy and capability that go well beyond existing augmented reality interaction frameworks.' Yet, amid this evolution, a deeper question of trust remains. Emotional intelligence, explainability and data transparency will be essential, not just for usability but for human agency. 'Services that require private knowledge need to justify that there is sufficient benefit directly to the user base,' Ullrich says. 'But if users see this as a fair trade, then I think it's a perfectly reasonable thing to allow.' As AI capabilities rise, we must consciously preserve human ones. The most meaningful advances may not be smarter machines, but more mindful connections between humans and promise of AI is so much more than productivity, it's dignity, inclusion and creativity. If we design wisely, AI won't just help us get more done, it will help us become more of who we are. And that is something worth imagining. • Artificial Intelligence • Smart Glasses• Wearable Tech• Smartphones • iPhones• Robots• Cars• TVs
Yahoo
a day ago
- Yahoo
Amazon may bring ads to Alexa+ in least surprising move ever
When you buy through links on our articles, Future and its syndication partners may earn a commission. Up until now Amazon has mostly avoided stuffing Alexa with ads, but according to a report from Mashable, Amazon CEO Andy Jassy broached the idea of delivering ads in Alexa Plus during the company's recent earnings call. Alexa Plus is the company's premium AI assistant that is supposed to be more naturally conversational than the previous Alexa. Jassy reportedly said that there is a "significant financial opportunity" in delivering ads via Alexa Plus. "I think over time, there will be opportunities, you know, as people are engaging in more multi-turn conversations to have advertising play a role — to help people find discovery and also as a lever to drive revenue," Jassy said, as found in the earnings call transcript. What is Alexa+? The new Alexa, introduced in February, has "enhanced" conversational abilities that are meant to put it on other AI assistants like Google's Gemini or ChatGPT. With improved memory its supposed to remember your details and help with managing tasks like booking tables for date night or buying groceries. Currently, Amazon offers a $19.99 a month subscription for non-Prime members. Prime subscribers, which costs $14.99 a month or $139 annually, can get Alexa+ for free Many Amazon devices are Alexa+ capable, though not all. However, Alexa has largely been an ad-free experience for the last decade since it was introduced. Perhaps it's not too surprising that Amazon would consider adding some kind of revenue procuring element to its AI assistant. The company has struggled for years to upgrade Alexa as other AI assistants surpassed Amazon who introduced chatbot assistants to the market. What kind of Alexa+ ads are we talking about? Largely, integrating ads is a revenue ploy to recoup some of the billions Amazon has burned in trying to turn its smart assistant around. Both Google and OpenAI have explored putting ads in Gemini and ChatGPT. OpenAI teased the idea in December of 2024 though the company stepped back from the idea in the same sentence. Amazon has not officially made moves to add commercials to Alexa+ conversations, but Jassy framed the idea as helpful. It would supposedly assist you in finding products that you might want to buy. With no set plans, it's worth keeping an eye on Alexa+ to see how or if Amazon actually follows through with Jassy's idea and how obtrusive they'll become. Hopefully they'll be skippable at the very least. More from Tom's Guide What is Amazon Prime? Everything you need to know Over 200 million Amazon Prime customers warned about scammers trying to steal their accounts — how to stay safe Amazon-backed Showrunner launches 'Netflix of AI' platform — and you can star in your own show
