verge-rss

Illustration by Kristen Radtke / The Verge; Getty Images

Elon Musk’s AI company, xAI, is making progress on adding multimodal inputs to its Grok chatbot, according to public developer documents. What this means is that, soon, users may be able to upload photos to Grok and receive text-based answers.
This was first teased in a blog post last month from xAI which said Grok-1.5V will offer “multimodal models in a number of domains.” The latest update to the developer documents appear to show progress on shipping a new model.
In the developer documents, a sample Python script demonstrates how developers can use the xAI software development kit library to generate a response based on both text and images. This script reads an image file, sets up a text prompt, and uses the xAI SDK to generate a response.

Image: xAI

This is a big update for Grok, which xAI first released in November 2023 and is available to users who pay for the X Premium Plus subscription. The last update was Grok 1.5 in March, which came with improved reasoning capabilities.
The model is trained “on a variety of text data from publicly available sources from the Internet up to Q3 2023 and data sets reviewed and curated by … human reviewers,” according to a blog post from X. Grok-1 was not trained on X data (including public X posts), the blog added. However, Grok does have “real-time knowledge of the world,” including posts on X.
xAI, founded by Elon Musk in March 2023, is relatively new in the AI field and trails behind competitors such as OpenAI’s ChatGPT. However, according to a blog post from xAI, their Grok 1.5 model is closing the gap with GPT-4 on various benchmarks that span a wide range of grade school to high school competition problems. It’s important to note that benchmarks for large language models are often criticized because the models can perform well on benchmarks if those benchmarks are included in their training data. It’s sort of like memorizing test answers, rather than actually learning the material.
Multimodal conversational chatbots seem to be the next frontier for AI, with multiple advancements announced at Google I/O and OpenAI releasing GPT-4o, so Grok lacking multimodal capabilities has put it behind the curve — until now.

Illustration by Kristen Radtke / The Verge; Getty Images

Elon Musk’s AI company, xAI, is making progress on adding multimodal inputs to its Grok chatbot, according to public developer documents. What this means is that, soon, users may be able to upload photos to Grok and receive text-based answers.

This was first teased in a blog post last month from xAI which said Grok-1.5V will offer “multimodal models in a number of domains.” The latest update to the developer documents appear to show progress on shipping a new model.

In the developer documents, a sample Python script demonstrates how developers can use the xAI software development kit library to generate a response based on both text and images. This script reads an image file, sets up a text prompt, and uses the xAI SDK to generate a response.

Image: xAI

This is a big update for Grok, which xAI first released in November 2023 and is available to users who pay for the X Premium Plus subscription. The last update was Grok 1.5 in March, which came with improved reasoning capabilities.

The model is trained “on a variety of text data from publicly available sources from the Internet up to Q3 2023 and data sets reviewed and curated by … human reviewers,” according to a blog post from X. Grok-1 was not trained on X data (including public X posts), the blog added. However, Grok does have “real-time knowledge of the world,” including posts on X.

xAI, founded by Elon Musk in March 2023, is relatively new in the AI field and trails behind competitors such as OpenAI’s ChatGPT. However, according to a blog post from xAI, their Grok 1.5 model is closing the gap with GPT-4 on various benchmarks that span a wide range of grade school to high school competition problems. It’s important to note that benchmarks for large language models are often criticized because the models can perform well on benchmarks if those benchmarks are included in their training data. It’s sort of like memorizing test answers, rather than actually learning the material.

Multimodal conversational chatbots seem to be the next frontier for AI, with multiple advancements announced at Google I/O and OpenAI releasing GPT-4o, so Grok lacking multimodal capabilities has put it behind the curve — until now.

Read More 

Photo: Allison Johnson / The Verge

Microsoft launched a range of Copilot Plus PCs yesterday that includes new AI features built directly into Windows 11. Behind the scenes, the company now has more than 40 AI models running on Windows 11 thanks to a new Windows Copilot Runtime that will also allow developers to use these models for their apps.
At Microsoft Build today, the company is providing a lot more details about exactly how this Windows Copilot Runtime works. The runtime includes a library of APIs that developers can tap into for their own apps, with AI frameworks and toolchains that are designed for developers to ship their own on-device models on Windows.
“Windows Copilot Library consists of ready-to-use AI APIs like Studio Effects, Live Captions Translations, OCR, Recall with User Activity, and Phi Silica, which will be available to developers in June,” explains Windows and Surface chief Pavan Davuluri.

Image: Microsoft
The new Windows Copilot Runtime.

Developers will be able to use the Windows Copilot Library to integrate things like Studio Effects, filters, portrait blur, and other features into their apps. Meta is adding the Windows Studio Effects into WhatsApp, so you’ll get features like background blur and eye contact during video calls. Even Live Captions and the new AI-powered translation feature can be used by developers with little to no code.
Microsoft demonstrated its Recall AI feature yesterday, allowing Copilot Plus PCs to document and store everything that you do on your PC so you can recall memories and search through a timeline. This is all powered by a new Windows Semantic Index that stores this data locally, and Microsoft plans to allow developers to build something similar.
“We will make this capability available for developers with Vector Embeddings API to build their own vector store and RAG within their applications and with their app data,” says Davuluri.

Photo: Allison Johnson / The Verge

Developers will also be able to improve Windows’ new Recall feature by adding contextual information to their apps that feeds into the database powering this feature. “This integration helps users pick up where they left off in your app, improving app engagement and users’ seamless flow between Windows and your app,” says Davuluri.
All of these improvements inside Windows for developers are the very early building blocks for more AI-powered apps on top of its new Arm-powered systems and the NPUs coming from AMD and Intel soon. While Microsoft is building the platform for developers to create AI apps for Windows, it’s now banking on this being an important part of the next decade of Windows development. Onstage at Build today, Davuluri stood in front of a slide that read “Windows is the most open platform for AI,” signaling just how important this moment is for Microsoft.

Photo: Allison Johnson / The Verge

Microsoft launched a range of Copilot Plus PCs yesterday that includes new AI features built directly into Windows 11. Behind the scenes, the company now has more than 40 AI models running on Windows 11 thanks to a new Windows Copilot Runtime that will also allow developers to use these models for their apps.

At Microsoft Build today, the company is providing a lot more details about exactly how this Windows Copilot Runtime works. The runtime includes a library of APIs that developers can tap into for their own apps, with AI frameworks and toolchains that are designed for developers to ship their own on-device models on Windows.

“Windows Copilot Library consists of ready-to-use AI APIs like Studio Effects, Live Captions Translations, OCR, Recall with User Activity, and Phi Silica, which will be available to developers in June,” explains Windows and Surface chief Pavan Davuluri.

Image: Microsoft
The new Windows Copilot Runtime.

Developers will be able to use the Windows Copilot Library to integrate things like Studio Effects, filters, portrait blur, and other features into their apps. Meta is adding the Windows Studio Effects into WhatsApp, so you’ll get features like background blur and eye contact during video calls. Even Live Captions and the new AI-powered translation feature can be used by developers with little to no code.

Microsoft demonstrated its Recall AI feature yesterday, allowing Copilot Plus PCs to document and store everything that you do on your PC so you can recall memories and search through a timeline. This is all powered by a new Windows Semantic Index that stores this data locally, and Microsoft plans to allow developers to build something similar.

“We will make this capability available for developers with Vector Embeddings API to build their own vector store and RAG within their applications and with their app data,” says Davuluri.

Photo: Allison Johnson / The Verge

Developers will also be able to improve Windows’ new Recall feature by adding contextual information to their apps that feeds into the database powering this feature. “This integration helps users pick up where they left off in your app, improving app engagement and users’ seamless flow between Windows and your app,” says Davuluri.

All of these improvements inside Windows for developers are the very early building blocks for more AI-powered apps on top of its new Arm-powered systems and the NPUs coming from AMD and Intel soon. While Microsoft is building the platform for developers to create AI apps for Windows, it’s now banking on this being an important part of the next decade of Windows development. Onstage at Build today, Davuluri stood in front of a slide that read “Windows is the most open platform for AI,” signaling just how important this moment is for Microsoft.

Read More 

The stylish Sonos Ace come in black and white. | Photo by Chris Welch / The Verge

At long last, the eagerly anticipated Sonos Ace have arrived. On Tuesday, the company announced that its first pair of wireless headphones will be available on June 5th. Even better, you can already preorder them ahead of launch for $449.

We’re still testing the Sonos Ace, but in our limited time playing around with them during a recent demo session, we came away impressed. The plush noise-canceling headphones are exceptionally comfortable to wear, with magnetic ear pads made of vegan leather and a memory foam headband. Plus, if you own a Sonos soundbar, you can quickly transfer TV audio from the soundbar to the headphones with a push of a button, allowing for a more immersive audio experience nobody else can hear.
Of course, we have yet to see how their overall sound, ANC, and transparency modes stack up against rivaling headphones from Bose, Sony, and Apple over extended listening sessions. We’ll be publishing our full review in the coming days, but in the meantime, here’s how to ensure you’re one of the first to get your hands on Sonos’ forthcoming headphones.
Where to preorder the Sonos Ace
The Sonos Ace are available to preorder from Sonos and Best Buy in black or white for $449 ahead of their June 5th release date. You can also preorder them at Amazon and B&H Photo starting on May 28th or sign up to receive release updates from either retailer now.

As previously said, we need more time to form a solid opinion about the Sonos Ace, but so far, we’re impressed by their capabilities and design. With just a click of a button, you can sync them with the Sonos Arc and — soon — Sonos’ other soundbars, including the Sonos Ray and the second-gen Sonos Beam. Combine that with their support for head-tracking spatial audio, and the headphones essentially let you create a private cinematic listening experience.

Granted, it’s a shame they can’t play music over Wi-Fi and you can’t group the Ace with Sonos speakers, but at least the headphones have other things going for them. They’re stylish and exceptionally comfortable to wear, for example, and we loved how much attention Sonos paid to detail. Sonos even includes a USB-C and headphone cable pouch inside the carrying case, which attaches magnetically. The company also added a fingerprint-resistant coating to reduce smudges, along with physical buttons for an intuitive experience, making for what seems so far like a well-rounded pair of headphones.

The stylish Sonos Ace come in black and white. | Photo by Chris Welch / The Verge

At long last, the eagerly anticipated Sonos Ace have arrived. On Tuesday, the company announced that its first pair of wireless headphones will be available on June 5th. Even better, you can already preorder them ahead of launch for $449.

We’re still testing the Sonos Ace, but in our limited time playing around with them during a recent demo session, we came away impressed. The plush noise-canceling headphones are exceptionally comfortable to wear, with magnetic ear pads made of vegan leather and a memory foam headband. Plus, if you own a Sonos soundbar, you can quickly transfer TV audio from the soundbar to the headphones with a push of a button, allowing for a more immersive audio experience nobody else can hear.

Of course, we have yet to see how their overall sound, ANC, and transparency modes stack up against rivaling headphones from Bose, Sony, and Apple over extended listening sessions. We’ll be publishing our full review in the coming days, but in the meantime, here’s how to ensure you’re one of the first to get your hands on Sonos’ forthcoming headphones.

Where to preorder the Sonos Ace

The Sonos Ace are available to preorder from Sonos and Best Buy in black or white for $449 ahead of their June 5th release date. You can also preorder them at Amazon and B&H Photo starting on May 28th or sign up to receive release updates from either retailer now.

As previously said, we need more time to form a solid opinion about the Sonos Ace, but so far, we’re impressed by their capabilities and design. With just a click of a button, you can sync them with the Sonos Arc and — soon — Sonos’ other soundbars, including the Sonos Ray and the second-gen Sonos Beam. Combine that with their support for head-tracking spatial audio, and the headphones essentially let you create a private cinematic listening experience.

Granted, it’s a shame they can’t play music over Wi-Fi and you can’t group the Ace with Sonos speakers, but at least the headphones have other things going for them. They’re stylish and exceptionally comfortable to wear, for example, and we loved how much attention Sonos paid to detail. Sonos even includes a USB-C and headphone cable pouch inside the carrying case, which attaches magnetically. The company also added a fingerprint-resistant coating to reduce smudges, along with physical buttons for an intuitive experience, making for what seems so far like a well-rounded pair of headphones.

Read More 

Photo by Justin Sullivan / Getty Images

Like parent company Disney, Pixar is now laying off employees as part of Bob Iger’s big plan to get the company’s spending under control.
According to Variety, Deadline, and The Hollywood Reporter, Pixar is laying off 175 staffers — 14 percent of its workforce — as it shifts its focus back to projects meant for theatrical rather than streaming releases. These layoffs come months after Disney began dismissing employees by the thousands in an effort to reduce spending by around $5.5 billion. Though Pixar was never guaranteed to be shielded from the layoffs, it was able to avoid them last year due to production schedules that were already in place at the time.
In a memo sent to staff on Tuesday, Pixar president Jim Morris stressed that the layoffs — which many employees knew were coming — are part of the studio’s efforts to “return to our focus on feature films” and “one of the hardest changes” executive leadership ultimately decided to make.
“Despite the challenges in our industry over the past few years, you have all consistently shown up to contribute, collaborate, innovate, lead, and do great work at this studio,” Morris said. “I give you my deepest thanks, and for those who will be leaving us, I am hopeful that our paths will cross again, both professionally and personally.”
Inside Out 2, Pixar’s latest film, is scheduled to debut on June 14th.

Photo by Justin Sullivan / Getty Images

Like parent company Disney, Pixar is now laying off employees as part of Bob Iger’s big plan to get the company’s spending under control.

According to Variety, Deadline, and The Hollywood Reporter, Pixar is laying off 175 staffers — 14 percent of its workforce — as it shifts its focus back to projects meant for theatrical rather than streaming releases. These layoffs come months after Disney began dismissing employees by the thousands in an effort to reduce spending by around $5.5 billion. Though Pixar was never guaranteed to be shielded from the layoffs, it was able to avoid them last year due to production schedules that were already in place at the time.

In a memo sent to staff on Tuesday, Pixar president Jim Morris stressed that the layoffs — which many employees knew were coming — are part of the studio’s efforts to “return to our focus on feature films” and “one of the hardest changes” executive leadership ultimately decided to make.

“Despite the challenges in our industry over the past few years, you have all consistently shown up to contribute, collaborate, innovate, lead, and do great work at this studio,” Morris said. “I give you my deepest thanks, and for those who will be leaving us, I am hopeful that our paths will cross again, both professionally and personally.”

Inside Out 2, Pixar’s latest film, is scheduled to debut on June 14th.

Read More 

Illustration: The Verge

Google only just rolled out AI summaries in search results — and now they’re getting ads. In an update on Tuesday, Google says it will soon start testing search and shopping ads within AI Overviews for users in the US.
In the example shared by Google, the search engine’s AI overview lists a response to the question: “how do I get wrinkles out of clothes?” Beneath the AI-generated suggestions, there’s a new “Sponsored” section with a carousel showing wrinkle spray you can buy from places like Walmart and Instacart.

GIF: Google

Google says it will display ads in AI Overviews when “they’re relevant to both the query and the information in the AI Overview.” Advertisers that already run certain campaigns through Google will automatically become eligible to appear in AI Overviews. “As we move forward, we’ll continue to test and learn new formats, getting feedback from advertisers and the industry,” Google writes.
There aren’t any details on when exactly Google will start testing ads in AI Overviews, but it kind of defeats the purpose of using AI to get a quick summary of what you’re looking for. If anything, it looks like it will just clutter that experience even more.

Illustration: The Verge

Google only just rolled out AI summaries in search results — and now they’re getting ads. In an update on Tuesday, Google says it will soon start testing search and shopping ads within AI Overviews for users in the US.

In the example shared by Google, the search engine’s AI overview lists a response to the question: “how do I get wrinkles out of clothes?” Beneath the AI-generated suggestions, there’s a new “Sponsored” section with a carousel showing wrinkle spray you can buy from places like Walmart and Instacart.

GIF: Google

Google says it will display ads in AI Overviews when “they’re relevant to both the query and the information in the AI Overview.” Advertisers that already run certain campaigns through Google will automatically become eligible to appear in AI Overviews. “As we move forward, we’ll continue to test and learn new formats, getting feedback from advertisers and the industry,” Google writes.

There aren’t any details on when exactly Google will start testing ads in AI Overviews, but it kind of defeats the purpose of using AI to get a quick summary of what you’re looking for. If anything, it looks like it will just clutter that experience even more.

Read More 

Cath Virginia / The Verge

It seems no industry is safe from consolidation, and the latest target is the gaming media. IGN Entertainment has acquired the website portfolio of UK publisher Gamer Network, which operates a number of beloved games-focused publications. That list includes Gamesindustry.biz, Eurogamer, Rock Paper Shotgun, VG247, and the tabletop site Dicebreaker. The network also holds shares in sites like Nintendo Life and Digital Foundry.
Terms of the deal were not disclosed. GI.biz reports that “some redundancies” have been made across the sites, though it’s not clear how many workers have been impacted. According to several posts on X, editors at both RPS and GI.biz have been laid off. IGN is owned by Ziff Davis which, in addition to IGN’s site, also operates other subsidiaries like the Humble Bundle.
The news comes as the games media, like the rest of the industry, has been thrust into a period of dramatic upheaval, brought about in part by the sudden rise of AI, changes to Google, and shifting editorial strategies. In the wake of this, several independent sites — like Aftermath, formed by former Kotaku staff — have appeared.

Cath Virginia / The Verge

It seems no industry is safe from consolidation, and the latest target is the gaming media. IGN Entertainment has acquired the website portfolio of UK publisher Gamer Network, which operates a number of beloved games-focused publications. That list includes Gamesindustry.biz, Eurogamer, Rock Paper Shotgun, VG247, and the tabletop site Dicebreaker. The network also holds shares in sites like Nintendo Life and Digital Foundry.

Terms of the deal were not disclosed. GI.biz reports that “some redundancies” have been made across the sites, though it’s not clear how many workers have been impacted. According to several posts on X, editors at both RPS and GI.biz have been laid off. IGN is owned by Ziff Davis which, in addition to IGN’s site, also operates other subsidiaries like the Humble Bundle.

The news comes as the games media, like the rest of the industry, has been thrust into a period of dramatic upheaval, brought about in part by the sudden rise of AI, changes to Google, and shifting editorial strategies. In the wake of this, several independent sites — like Aftermath, formed by former Kotaku staff — have appeared.

Read More 

The Snapdragon Dev Kit for Windows in a transparent shell; most if not all units will be black. | Image: Qualcomm

Excited for an era of long-lasting powerful Windows laptops with Arm chips and 45 TOPS for AI but would rather pay less and plug into the wall? Qualcomm has just revealed a Mac Mini-esque box that’s ostensibly just for developers.
The new $899.99 “Snapdragon Dev Kit for Windows” houses the most powerful of Qualcomm’s lineup of Snapdragon X Elite chips — the one with the 4.6 TFLOP GPU — as well as 32GB of LPDDR5x RAM, 512GB of NVMe storage, and lots of ports, all within roughly the same volume as Apple’s mini desktop.

Image: Qualcomm

Image: Qualcomm

While the spec sheet unfortunately suggests the two-pound PC will only ship in black — not the transparent model at the top of this post — the eight- by seven- by 1.3-inch PC does have Wi-Fi 7, Bluetooth 5.4, and enough oomph to run three 4K monitors at once.
Ports include three USB4 ports with Type-C connectors, a pair of USB 3.2 ports with Type-A, ethernet, 3.5mm audio, and HDMI. So if you want to go the multi-monitor route, expect to plug in at least one USB-C cable or adapter.

Image: Qualcomm

Speaking of adapters, what you won’t find in Qualcomm’s images here is the power brick: it’s got an external 180W AC adapter.

While the Dev Kit might be more cost-effective than the new wave of Microsoft Copilot Plus PC laptops that include similar Qualcomm chips, it’s not by much: the Surface Laptop 6 starts at $999.99 and obviously comes with its own screen. And it’s not clear just anyone will be able to buy the Dev Kit: Qualcomm’s landing page currently points to a contact form that you may need to fill out to get one.

The Snapdragon Dev Kit for Windows in a transparent shell; most if not all units will be black. | Image: Qualcomm

Excited for an era of long-lasting powerful Windows laptops with Arm chips and 45 TOPS for AI but would rather pay less and plug into the wall? Qualcomm has just revealed a Mac Mini-esque box that’s ostensibly just for developers.

The new $899.99 “Snapdragon Dev Kit for Windows” houses the most powerful of Qualcomm’s lineup of Snapdragon X Elite chips — the one with the 4.6 TFLOP GPU — as well as 32GB of LPDDR5x RAM, 512GB of NVMe storage, and lots of ports, all within roughly the same volume as Apple’s mini desktop.

Image: Qualcomm

Image: Qualcomm

While the spec sheet unfortunately suggests the two-pound PC will only ship in black — not the transparent model at the top of this post — the eight- by seven- by 1.3-inch PC does have Wi-Fi 7, Bluetooth 5.4, and enough oomph to run three 4K monitors at once.

Ports include three USB4 ports with Type-C connectors, a pair of USB 3.2 ports with Type-A, ethernet, 3.5mm audio, and HDMI. So if you want to go the multi-monitor route, expect to plug in at least one USB-C cable or adapter.

Image: Qualcomm

Speaking of adapters, what you won’t find in Qualcomm’s images here is the power brick: it’s got an external 180W AC adapter.

While the Dev Kit might be more cost-effective than the new wave of Microsoft Copilot Plus PC laptops that include similar Qualcomm chips, it’s not by much: the Surface Laptop 6 starts at $999.99 and obviously comes with its own screen. And it’s not clear just anyone will be able to buy the Dev Kit: Qualcomm’s landing page currently points to a contact form that you may need to fill out to get one.

Read More 

Image: Microsoft

Microsoft’s new Surface Laptop could be the most impressive laptop the company has ever made. It has a modern chassis with thin display bezels and a new Arm-based CPU that might just give Apple laptops a run for their money when it comes to performance. That’s no mean feat, considering Apple’s striking lead since 2020, but we won’t know how it compares until we get a Surface Laptop in hand. That didn’t stop me from diving deep on their specs to see how they compare on paper.
Before we get to the real nitty-gritty, let’s take a look at some of the most obvious differences you’ll see between the Surface Laptop and the MacBook Air.
What’s it gonna cost you?
The M3 MacBook Air starts at $1,099 for an 8GB model, which already puts it at a disadvantage since it’s $100 more than the $999.99 Surface Laptop, which also gets 16GB of RAM out of the gate (more on that later).
Both can get much spendier, though. Spec up the Surface, and you can spend $2,499.99 for the preorder-exclusive 64GB RAM 15-inch model with 1TB SSD and a Snapdragon X Elite. For the same price, you can get a 15-inch MacBook Air with 24GB of RAM and 2TB SSD.
Let’s get physical

Image: Microsoft / Apple
The Surface Laptop (left) and MacBook Air (right): laptops on a path of convergent evolution.

At first glance and from certain angles, the Surface Laptop and MacBook Air look very similar. Both laptops are encased in metal and feature thin display bezels, chicklet-style keys, and big haptic trackpads.
The Surface Laptop uses Windows Hello face authentication, rather than a fingerprint sensor, for logging in. The MacBook Air, despite having a notch drooping down into the screen since 2022, has no Face ID and uses Touch ID fingerprint scanning instead, which works well, but still.
The 13-inch MacBook Air has a higher native resolution (2560 x 1664) compared to the 13.8-inch Surface Laptop (2304 x 1536). Likewise, the 15-inch Surface Laptop has a 2496 x 1664 resolution to the 15-inch Air’s 2880 x 1864 resolution. That means the MacBook Air can show a little more information on its 16:10 ratio screen than the 3:2 ratio Surface. But Microsoft’s laptop has a touchscreen, which no Macs have, and a 120Hz variable refresh rate, which only MacBook Pros get.
The Surface has more ports than the Air, too. Both companies’ laptops have 3.5mm headphone jacks and dual USB 4 Type-C ports, but the Surface Laptop also has a USB-A port and the 15-inch model gets a microSDXC card reader. MagSafe helps the Air a bit by freeing up a USB-C port when charging.
Microsoft says the Surface is its thinnest, lightest laptop ever, but it’s still a bit chonky. It weighs 2.96 pounds or 3.67 pounds and is 0.69 inches or 0.72 inches thick, depending on whether you get the small or large one. The MacBook Air, on the other hand, sits at a waifish 0.44 inches thick (or 0.45 for the 15-inch) and weighs 2.7 pounds or 3.3 pounds. Small differences, yes, but they’re quite noticeable when you’re handling a laptop.
What’s inside counts

Image: Qualcomm / Apple
Which square is for you?

The crucial factor in Apple’s performance has been how much power it actually uses to get there. Plenty of Intel and AMD laptops have offered better raw performance, but nobody has really touched Apple when it comes to the energy cost, at least until now.
The Surface and Air have similar battery capacities. Microsoft rates the Surface Laptop at either a nominal 54Wh in the 13.8-inch or 66Wh for the 15-inch. Microsoft says that amounts to 20 or 22 hours of video and 13 or 15 hours of web browsing, depending on whether you’re using the smaller or bigger Surface. The MacBook Air is similar, with 52.6Wh for the 13-inch model or 66.5Wh for the 15-inch Air, good for 18 hours of video and 15 hours of web browsing, according to Apple.
But don’t count on those numbers. The two companies detail on their specs pages how they tested, but it’s not particularly helpful for comparing without real-world testing. For instance, Microsoft bases the 20-hour number on playing a 1080p video on a Surface with 256GB storage, 16GB RAM, and a Snapdragon X Plus, with brightness set to 150 nits. Apple, on the other hand, tested an M3 MacBook Air, also using 1080p video, with the brightness set “8 clicks from the bottom.” Depending on the panel, that can be similar to 150 nits, but there’s no guarantee.
Microsoft’s laptop also has 16GB out of the gate, while the MacBook Air has 8GB to start. You can argue about whether that’s enough for basic use and about the merits of Apple’s “unified memory” approach, but 16GB of RAM is still double the working memory, and the Surface machine can be configured up to 32GB (or 64GB in early preorders), whereas Apple’s machine tops out at 24GB. That means you can probably do more memory-intensive tasks on a Surface Laptop before it starts bogging down.
But as far as just pure performance, Microsoft hasn’t been specific about how the Surface Laptop performs with a 10-core Snapdragon X Plus or 12-core Elite chip. Qualcomm says those are all “performance” cores, rather than splitting the work between those and “efficiency” cores the way Apple and others do. That means the Snapdragon X chips don’t farm easier work to cores that aren’t as performative but that use less power. Still, the company also claims they’re more efficient than Apple’s chips. So we’ll have to wait and see how they compare in practice.
But Microsoft did claim at its May 20th Surface AI event that so-called “Copilot Plus PCs” will be “58 percent faster” than an M3 MacBook Air. The company wasn’t specific about whether this was Snapdragon X-equipped machines, but those are the first Copilot Plus PCs, so it’s possible.
Microsoft spent more time talking about the AI chops because it clearly wants these machines used for AI tasks. The company says the neural processing unit (NPU) in the Surface can carry out 45 trillion operations per second, or TOPS (that’s the number of mathematical equations it can solve per second). That’s more than double the 18 TOPS Apple claims for its Neural Engine chip in the M3. It’s even more than the M4 that’s in the new iPad Pro, which Apple says can pull off 38 TOPS. That’s impressive, though TOPS is a bit arbitrary since it doesn’t tell us much about workload being carried by the chip.
But all of this processor talk is just talk at the moment — comparing specs and reading tea leaves. We still need to see these devices in action, outside of demos, to get a real sense of their power. Thankfully, we don’t have to wonder for much longer how the Surface Laptop and all of the other Snapdragon X-equipped machines will do. The Surface is up for preorder and ships on June 18th, and others are coming from the likes of Dell, Asus, Lenovo, and Samsung.

Image: Microsoft

Microsoft’s new Surface Laptop could be the most impressive laptop the company has ever made. It has a modern chassis with thin display bezels and a new Arm-based CPU that might just give Apple laptops a run for their money when it comes to performance. That’s no mean feat, considering Apple’s striking lead since 2020, but we won’t know how it compares until we get a Surface Laptop in hand. That didn’t stop me from diving deep on their specs to see how they compare on paper.

Before we get to the real nitty-gritty, let’s take a look at some of the most obvious differences you’ll see between the Surface Laptop and the MacBook Air.

What’s it gonna cost you?

The M3 MacBook Air starts at $1,099 for an 8GB model, which already puts it at a disadvantage since it’s $100 more than the $999.99 Surface Laptop, which also gets 16GB of RAM out of the gate (more on that later).

Both can get much spendier, though. Spec up the Surface, and you can spend $2,499.99 for the preorder-exclusive 64GB RAM 15-inch model with 1TB SSD and a Snapdragon X Elite. For the same price, you can get a 15-inch MacBook Air with 24GB of RAM and 2TB SSD.

Let’s get physical

Image: Microsoft / Apple
The Surface Laptop (left) and MacBook Air (right): laptops on a path of convergent evolution.

At first glance and from certain angles, the Surface Laptop and MacBook Air look very similar. Both laptops are encased in metal and feature thin display bezels, chicklet-style keys, and big haptic trackpads.

The Surface Laptop uses Windows Hello face authentication, rather than a fingerprint sensor, for logging in. The MacBook Air, despite having a notch drooping down into the screen since 2022, has no Face ID and uses Touch ID fingerprint scanning instead, which works well, but still.

The 13-inch MacBook Air has a higher native resolution (2560 x 1664) compared to the 13.8-inch Surface Laptop (2304 x 1536). Likewise, the 15-inch Surface Laptop has a 2496 x 1664 resolution to the 15-inch Air’s 2880 x 1864 resolution. That means the MacBook Air can show a little more information on its 16:10 ratio screen than the 3:2 ratio Surface. But Microsoft’s laptop has a touchscreen, which no Macs have, and a 120Hz variable refresh rate, which only MacBook Pros get.

The Surface has more ports than the Air, too. Both companies’ laptops have 3.5mm headphone jacks and dual USB 4 Type-C ports, but the Surface Laptop also has a USB-A port and the 15-inch model gets a microSDXC card reader. MagSafe helps the Air a bit by freeing up a USB-C port when charging.

Microsoft says the Surface is its thinnest, lightest laptop ever, but it’s still a bit chonky. It weighs 2.96 pounds or 3.67 pounds and is 0.69 inches or 0.72 inches thick, depending on whether you get the small or large one. The MacBook Air, on the other hand, sits at a waifish 0.44 inches thick (or 0.45 for the 15-inch) and weighs 2.7 pounds or 3.3 pounds. Small differences, yes, but they’re quite noticeable when you’re handling a laptop.

What’s inside counts

Image: Qualcomm / Apple
Which square is for you?

The crucial factor in Apple’s performance has been how much power it actually uses to get there. Plenty of Intel and AMD laptops have offered better raw performance, but nobody has really touched Apple when it comes to the energy cost, at least until now.

The Surface and Air have similar battery capacities. Microsoft rates the Surface Laptop at either a nominal 54Wh in the 13.8-inch or 66Wh for the 15-inch. Microsoft says that amounts to 20 or 22 hours of video and 13 or 15 hours of web browsing, depending on whether you’re using the smaller or bigger Surface. The MacBook Air is similar, with 52.6Wh for the 13-inch model or 66.5Wh for the 15-inch Air, good for 18 hours of video and 15 hours of web browsing, according to Apple.

But don’t count on those numbers. The two companies detail on their specs pages how they tested, but it’s not particularly helpful for comparing without real-world testing. For instance, Microsoft bases the 20-hour number on playing a 1080p video on a Surface with 256GB storage, 16GB RAM, and a Snapdragon X Plus, with brightness set to 150 nits. Apple, on the other hand, tested an M3 MacBook Air, also using 1080p video, with the brightness set “8 clicks from the bottom.” Depending on the panel, that can be similar to 150 nits, but there’s no guarantee.

Microsoft’s laptop also has 16GB out of the gate, while the MacBook Air has 8GB to start. You can argue about whether that’s enough for basic use and about the merits of Apple’s “unified memory” approach, but 16GB of RAM is still double the working memory, and the Surface machine can be configured up to 32GB (or 64GB in early preorders), whereas Apple’s machine tops out at 24GB. That means you can probably do more memory-intensive tasks on a Surface Laptop before it starts bogging down.

But as far as just pure performance, Microsoft hasn’t been specific about how the Surface Laptop performs with a 10-core Snapdragon X Plus or 12-core Elite chip. Qualcomm says those are all “performance” cores, rather than splitting the work between those and “efficiency” cores the way Apple and others do. That means the Snapdragon X chips don’t farm easier work to cores that aren’t as performative but that use less power. Still, the company also claims they’re more efficient than Apple’s chips. So we’ll have to wait and see how they compare in practice.

But Microsoft did claim at its May 20th Surface AI event that so-called “Copilot Plus PCs” will be “58 percent faster” than an M3 MacBook Air. The company wasn’t specific about whether this was Snapdragon X-equipped machines, but those are the first Copilot Plus PCs, so it’s possible.

Microsoft spent more time talking about the AI chops because it clearly wants these machines used for AI tasks. The company says the neural processing unit (NPU) in the Surface can carry out 45 trillion operations per second, or TOPS (that’s the number of mathematical equations it can solve per second). That’s more than double the 18 TOPS Apple claims for its Neural Engine chip in the M3. It’s even more than the M4 that’s in the new iPad Pro, which Apple says can pull off 38 TOPS. That’s impressive, though TOPS is a bit arbitrary since it doesn’t tell us much about workload being carried by the chip.

But all of this processor talk is just talk at the moment — comparing specs and reading tea leaves. We still need to see these devices in action, outside of demos, to get a real sense of their power. Thankfully, we don’t have to wonder for much longer how the Surface Laptop and all of the other Snapdragon X-equipped machines will do. The Surface is up for preorder and ships on June 18th, and others are coming from the likes of Dell, Asus, Lenovo, and Samsung.

Read More 

Image: The Verge

Expect a bunch of new developer tools to go along with Microsoft’s fresh Copilot Plus PCs announcement. Microsoft is kicking off its three-day Build developer conference on Tuesday, May 21st, with a livestream starting at 11:30AM ET / 8:30AM PT. It’ll lead into an in-person keynote led by CEO Satya Nadella, which commences at 12PM ET / 9AM PT, followed by developer sessions that will come available to check out online.
Build is Microsoft’s developer conference where the company provides in-depth sessions for developers and professionals alike to get familiar with tools supporting new Windows 11 and Microsoft 365 features. This time, we’re expecting plenty of new AI announcements and sessions related to just-announced AI features like Recall.
Microsoft made a huge splash in the PC world on Monday announcing new Arm-powered “Copilot Plus PCs,” including a brand-new Surface Laptop and tablet. The new devices come with an emulation layer called Prism that promises seamless compatibility with x86 apps on Windows — taking a page out of Apple’s successful transition to its own M-series chips.
However, as Tom Warren wrote in his Notepad newsletter after the Surface event, “Apple’s success with the M1 was thanks to developers quickly porting apps to be fully native. Windows needs that same level of support from its developer community.” We’ll see this week if Microsoft has all the tools it needs to make that happen.
Read on for all the latest Build news.

Image: The Verge

Expect a bunch of new developer tools to go along with Microsoft’s fresh Copilot Plus PCs announcement.

Microsoft is kicking off its three-day Build developer conference on Tuesday, May 21st, with a livestream starting at 11:30AM ET / 8:30AM PT. It’ll lead into an in-person keynote led by CEO Satya Nadella, which commences at 12PM ET / 9AM PT, followed by developer sessions that will come available to check out online.

Build is Microsoft’s developer conference where the company provides in-depth sessions for developers and professionals alike to get familiar with tools supporting new Windows 11 and Microsoft 365 features. This time, we’re expecting plenty of new AI announcements and sessions related to just-announced AI features like Recall.

Microsoft made a huge splash in the PC world on Monday announcing new Arm-powered “Copilot Plus PCs,” including a brand-new Surface Laptop and tablet. The new devices come with an emulation layer called Prism that promises seamless compatibility with x86 apps on Windows — taking a page out of Apple’s successful transition to its own M-series chips.

However, as Tom Warren wrote in his Notepad newsletter after the Surface event, “Apple’s success with the M1 was thanks to developers quickly porting apps to be fully native. Windows needs that same level of support from its developer community.” We’ll see this week if Microsoft has all the tools it needs to make that happen.

Read on for all the latest Build news.

Read More 

Microsoft is adding custom emoji to Teams. | Image: The Verge

Microsoft is adding a new feature to its Teams communications platform that enables users to upload their own custom emoji to use in reactions and messages. Announced during its Build developer conference on Tuesday, Microsoft says the new custom emoji will be available to try next month via the Teams public preview, with the goal of helping Teams users collaborate and express themselves “more creatively and authentically.”
IT admins for businesses that use Teams will have the ability to limit which users can upload or delete custom emoji, or they can turn the feature off entirely. Once custom emoji are uploaded into Teams, they’ll only be visible within the same organization domain. Microsoft says that general availability for custom emoji is expected sometime this July.

Image: Microsoft
Here’s an example of how the custom emoji options will be presented to Teams users.

The announcement comes three months after Microsoft employees discovered an early version of the feature on an internal Teams build, finding an animated emoji of the widely memed Pepe the Frog character in some reactions and messages. Currently, Microsoft Teams only supports official Unicode emoji — the standardized array of public domain emoji supported by most smartphones and social media platforms.
In contrast to other communications platforms, Microsoft is very late to the custom emoji party. It’s been a prominent feature on Slack and Discord for years already, while Google Chat rolled out a similar feature for Workspace users back in 2022.

Microsoft is adding custom emoji to Teams. | Image: The Verge

Microsoft is adding a new feature to its Teams communications platform that enables users to upload their own custom emoji to use in reactions and messages. Announced during its Build developer conference on Tuesday, Microsoft says the new custom emoji will be available to try next month via the Teams public preview, with the goal of helping Teams users collaborate and express themselves “more creatively and authentically.”

IT admins for businesses that use Teams will have the ability to limit which users can upload or delete custom emoji, or they can turn the feature off entirely. Once custom emoji are uploaded into Teams, they’ll only be visible within the same organization domain. Microsoft says that general availability for custom emoji is expected sometime this July.

Image: Microsoft
Here’s an example of how the custom emoji options will be presented to Teams users.

The announcement comes three months after Microsoft employees discovered an early version of the feature on an internal Teams build, finding an animated emoji of the widely memed Pepe the Frog character in some reactions and messages. Currently, Microsoft Teams only supports official Unicode emoji — the standardized array of public domain emoji supported by most smartphones and social media platforms.

In contrast to other communications platforms, Microsoft is very late to the custom emoji party. It’s been a prominent feature on Slack and Discord for years already, while Google Chat rolled out a similar feature for Workspace users back in 2022.

Read More