Latest Tech News

Stay updated with the latest in technology, AI, cybersecurity, and more

Filtered by: mo Clear Filter

We Rewrote the Ghostty GTK Application

August 14, 2025 We just completed rewriting the Ghostty GTK application fully embracing the GObject type system from Zig and also verifying with Valgrind every step of the way. The result is a more feature rich, stable, and maintainable Ghostty on Linux and BSD. There are multiple interesting, technical topics from this process, but I want to focus in on two (1) interfacing with the GObject type system from Zig and (2) verifying a GTK application with Valgrind and reflecting on the memory issu

The 65 Best Movies on Disney+ Right Now (August 2025)

In the game known as the streaming wars, Disney+ came out swinging, bringing with it a massive library of movies and TV shows—with new ones being added all the time. Watched everything on Netflix? Disney+ has a seemingly endless selection of Marvel movies and plenty of Star Wars and Pixar fare too. Problem is, there’s so much stuff that it’s hard to know where to begin. WIRED is here to help. Below are our picks for the best films on Disney+ right now. For more viewing ideas, try our guides to

Netflix's KPop Demon Hunters is headed to theaters for two days

You'll soon be able to watch KPop Demon Hunters on the big screen. The Netflix animated film has become a global hit since it launched on the platform and has reportedly become the second most-watched movie on the service's history after Red Notice. Now, Netflix is holding a limited theatrical event, wherein fans can watch a sing-along version of the film. Aside from the movie itself being a hit, its songs have gained massive popularity, with the track Golden recently taking the number one spot

Starlink Users Will Now Have to Pay $5 to Pause Service

Starlink is ending a popular free feature that let customers pause service at any time for free. Now, you’ll have to pay $5 a month to enter what the company is calling “Standby Mode.” Subscribers who’d been using the pause feature began receiving emails yesterday notifying them that they’d have to opt in to the new Standby Mode by Sept. 13, or their service would be canceled. “We recently upgraded pause to include Standby Mode,” the company wrote on a support page. “Previously, the pause feat

Google releases pint-size Gemma open AI model

Big tech has spent the last few years creating ever-larger AI models, leveraging rack after rack of expensive GPUs to provide generative AI as a cloud service. But tiny AI matters, too. Google has announced a tiny version of its Gemma open model designed to run on local devices. Google says the new Gemma 3 270M can be tuned in a snap and maintains robust performance despite its small footprint. Google released its first Gemma 3 open models earlier this year, featuring between 1 billion and 27 b

Newly Discovered Fossils Reveal Unknown Humanlike Relative

Researchers have uncovered fossils belonging to a previously unknown ancient human relative. And they may have lived in the same time and place as the earliest-known members of the genus Homo, from which modern humans evolved, according to a new study. A team of archeologists working at Ethiopia’s Ledi-Geraru research project area unearthed a set of fossilized teeth that likely belonged to an unidentified species within the Australopithecus genus, known for having both human- and ape-like trait

The “Godfather of AI” Has a Bizarre Plan to Save Humanity From Evil AI

Geoffrey Hinton, the pioneering mind behind AI industry-transforming neural networks, who's often referred to as a "godfather of AI," says we need to infuse AI with "maternal instincts" to save humanity from rogue AI. Though his work on neural networks helped to usher in the large language models (LLMs) that dominate Silicon Valley today, these days, Hinton is known for being somewhat of an AI alarmist: he believes that there's a significant chance that superintelligent AI will wipe out humanki

Google unveils ultra-small and efficient open source AI model Gemma 3 270M that can run on smartphones

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now Google’s DeepMind AI research team has unveiled a new open source AI model today, Gemma 3 270M. As its name would suggest, this is a 270-million-parameter model — far smaller than the 70 billion or more parameters of many frontier LLMs (parameters being the number of internal settings governing the model’s behavior). While more parameters

Gemma 3 270M: Compact model for hyper-efficient AI

Today, we're adding a new, highly specialized tool to the Gemma 3 toolkit: Gemma 3 270M , a compact, 270-million parameter model designed from the ground up for task-specific fine-tuning with strong instruction-following and text structuring capabilities already trained in. The last few months have been an exciting time for the Gemma family of open models. We introduced Gemma 3 and Gemma 3 QAT , delivering state-of-the-art performance for single cloud and desktop accelerators. Then, we announce

The 15 Best Genre Movies of Summer 2025 (and Where to Watch Them)

A few short weeks remain in the 2025 summer movie season, and we have to say, it’s been a pretty great one. Lots of the big blockbusters really delivered. Lots of smaller films came out of nowhere to surprise us. And even a few streaming titles made their mark. Below, we’ve got our choices for the 15 best genre movies released during the summer of 2025, from the first weekend in May through the end of August. 15. Thunderbolts Spoiler: this is the only Marvel movie on this list. Not that we dis

Scientists Taught AI to Predict Nuclear Fusion Success—and It’s Actually Working

AI is giving a huge efficiency boost to one of the biggest nuclear fusion facilities in the world—but perhaps not in the way you think. In research published today in Science, scientists at Lawrence Livermore National Laboratory report how its newly developed deep learning model accurately predicted the results of a 2022 fusion experiment at the National Ignition Facility (NIF). The model, which assigned 74% probability for ignition in that experiment, outperforms traditional supercomputing met

What is Bluesky? Everything to know about the X competitor

Is the grass greener on the other side? We’re not sure, but the sky is most certainly bluer. It’s been over two years since Elon Musk purchased Twitter, now X, leading people to set up shop on alternative platforms. Mastodon, Post, Pebble (two of which have already shuttered operations) and Spill have been presented as potential replacements, but few aside from Meta’s Threads have achieved the speed of growth Bluesky has reached. As of February 2025, Bluesky has surpassed 30 million users. Its

Anthropic takes on OpenAI and Google with new Claude AI features designed for students and developers

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now Anthropic is launching new “learning modes” for its Claude AI assistant that transform the chatbot from an answer-dispensing tool into a teaching companion, as major technology companies race to capture the rapidly growing artificial intelligence education market while addressing mounting concerns that AI undermines genuine learning. The S

Anthropic brings Claude's learning mode to regular users and devs

This past spring, Anthropic introduced learning mode, a feature that changed Claude's interaction style. When enabled, the chatbot would, following a question, try to guide the user to their own solution, instead of providing them with an answer outright. Since its introduction in April, learning mode has only been available to Claude for Education users. Now, like OpenAI did with Study Mode, Anthropic is making the tool available to everyone. Starting today, Claude.ai users will find a new opt

Why LLMs can't really build software

One of the things I have spent a lot of time doing is interviewing software engineers. This is obviously a hard task, and I don’t claim to have a magic solution; but it’s given me some time to reflect on what effective software engineers actually do. When you watch someone who knows what they are doing, you'll see them looping over the following steps: Build a mental model of the requirements Write code that (hopefully?!) does that Build a mental model of what the code actually does Identify t

Stop using these ESR power banks that have been recalled for fire and explosion risks

is a senior reporter who’s been covering and reviewing the latest gadgets and tech since 2006, but has loved all things electronic since he was a kid. Posts from this author will be added to your daily email digest and your homepage feed. ESR has issued a recall for 33,000 HaloLock wireless power banks, in 6,000mAh and 10,000mAh versions, because their lithium-ion batteries can “overheat and ignite, posing fire and burn hazards to consumers.” The power banks were cheaper alternatives to Apple’

Topics: 000 banks esr model power

Consumer safety groups are demanding an FTC investigation into Grok’s ‘Spicy’ mode

is The Verge’s senior AI reporter. An AI beat reporter for more than five years, her work has also appeared in CNBC, MIT Technology Review, Wired UK, and other outlets. A new letter to the Federal Trade Commission, and US attorneys general for all 50 states and the District of Columbia, calls for an urgent investigation into Elon Musk’s Grok — especially its new “Imagine” tool for AI-generated image and video. The tool, released by xAI earlier this month, encourages users to create NSFW conten

Gemma 3 270M: The compact model for hyper-efficient AI

Today, we're adding a new, highly specialized tool to the Gemma 3 toolkit: Gemma 3 270M , a compact, 270-million parameter model designed from the ground up for task-specific fine-tuning with strong instruction-following and text structuring capabilities already trained in. The last few months have been an exciting time for the Gemma family of open models. We introduced Gemma 3 and Gemma 3 QAT , delivering state-of-the-art performance for single cloud and desktop accelerators. Then, we announce

The End of Wireless Dead Zones? Starlink Texting Now Available on AT&T, T-Mobile and Verizon Phones

After six months of beta testing, Starlink's partnership with T-Mobile officially launched on July 23. The direct-to-cell messaging service enables texting from anywhere in the US. It's not limited to T-Mobile customers, either -- AT&T and Verizon mobile customers can purchase it for $10 a month. T-Mobile says its goal is to "eliminate mobile dead zones for good" by way of 657 Starlink satellites that'll be used exclusively for cellphone service. The new satellite texting service represents a

Lovable projects $1B in ARR within next 12 months

In Brief Vibe coding startup Lovable aims to hit $1 billion in annual recurring revenue within the next 12 months, according to its CEO, Anton Osika. Speaking on Bloomberg TV on Thursday, Osika said the company grows by at least $8 million in ARR each month. In a blog post written this summer, the company said it passed $100 million in ARR just eight months after making its first $1 million. Osika told Bloomberg Thursday the company is projecting to reach $250 million in ARR by the end of this

Show HN: Zig-DbC – A design by contract library for Zig

Hi everyone, I've made an open-source library for using design by contract (DbC) principles in the Zig programming language. It's called Zig-DbC, and it currently provides the following features: - A simple API to define preconditions, postconditions, and invariants. - Contracts are active in `Debug`, `ReleaseSafe`, and `ReleaseSmall` modes to catch bugs early. - All checks are removed at compile time in `ReleaseFast` mode for zero performance cost. - An optional mode to handle partial sta

Is chain-of-thought AI reasoning a mirage?

Reading research papers and articles about chain-of-thought reasoning makes me frustrated. There are many interesting questions to ask about chain-of-thought: how accurately it reflects the actual process going on, why training it “from scratch” often produces chains that switch fluidly between multiple languages, and so on. However, people keep asking the least interesting question possible: whether chain-of-thought reasoning is “really” reasoning. Apple took up this question in their Illusio

Misunderstood “photophoresis” effect could loft metal sheets to exosphere

Most people would recognize the device in the image above, although they probably wouldn't know it by its formal name: the Crookes radiometer. As its name implies, placing the radiometer in light produces a measurable change: the blades start spinning. Unfortunately, many people misunderstand the physics of its operation (which we'll return to shortly). The actual forces that drive the blades to spin, called photophoresis, can act on a variety of structures as long as they're placed in a suffic

Buzzy AI startup Multiverse creates two of the smallest high-performing models ever

One of Europe’s most prominent AI startups has released two AI models that are so tiny, they have named them after a chicken’s brain and a fly’s brain. Multiverse Computing claims these are the world’s smallest models that are still high-performing and can handle chat, speech, and even reasoning in one case. These new tiny models are intended to be embedded into Internet of Things devices, as well as run locally on smartphones, tablets, and PCs. “We can compress the model so much that they ca

Computing’s Top 30: Zhihao “Zephyr” Yao

On a typical mobile device today, financial and medical apps nestled up next to everything from karaoke playlists to time-killing games like Fruit Ninja. How to secure data that matters in this diverse digital buffet is a challenge for many researchers. For Zhihao “Zephyr” Yao, it’s a challenge that fuels his life’s work and also led to an award-winning project. That project—which earned ACM MobiSys 2023’s Best Artifact Award—demonstrated that making systems less complex can actually enhance m

Why LLMs Can't Build Software

One of the things I have spent a lot of time doing is interviewing software engineers. This is obviously a hard task, and I don’t claim to have a magic solution; but it’s given me some time to reflect on what effective software engineers actually do. When you watch someone who knows what they are doing, you'll see them looping over the following steps: Build a mental model of the requirements Write code that (hopefully?!) does that Build a mental model of what the code actually does Identify t

Philips Hue’s new bridge could turn your lights into motion sensors

is a senior reviewer focused on smart home and connected tech, with over twenty years of experience. She has written previously for Wirecutter, Wired, Dwell, BBC, and US News. Posts from this author will be added to your daily email digest and your homepage feed. Smart lighting company Philips Hue appears to be launching a new version of its bridge, the first in a decade. According to a product page briefly published on Hue’s website and since removed, the Hue Bridge Pro will be faster, have m

Canada’s House of Commons investigating data breach after cyberattack

The House of Commons of Canada is currently investigating a data breach after a threat actor reportedly stole employee information in a cyberattack on Friday. While the lower house of the Parliament of Canada has yet to issue a public statement regarding this incident, CBC News reports that House of Commons staff were notified of a breach on Monday via email. The alert states that the attacker exploited a recent Microsoft vulnerability to gain access to a database containing sensitive informat

What's the strongest AI model you can train on a laptop in five minutes?

What’s the strongest model I can train on my MacBook Pro in five minutes? I’ll give the answer upfront: the best 5-minute model I could train was a ~1.8M-param GPT-style transformer trained on ~20M TinyStories tokens, reaching ~9.6 perplexity on a held-out split. Here’s an example of the output, with the prompt bolded: Once upon a time, there was a little boy named Tim. Tim had a small box that he liked to play with. He would push the box to open. One day, he found a big red ball in his yard.

New protein therapy shows promise as antidote for carbon monoxide poisoning

University of Maryland School of Medicine (UMSOM) researchers, along with their colleagues, engineered a new molecule that appears promising as an effective antidote for carbon monoxide poisoning with fewer side effects than other molecules currently being tested, according to a new study published in the journal PNAS. Carbon monoxide poisoning accounts for 50,000 emergency room visits in the U.S. each year and causes about 1,500 deaths. These deaths may occur when carbon monoxide released from