Here in Bend, Oregon, where the digital frontier meets the rugged individualism of the West, the annual tech showcases from the giants often provide a roadmap for the future. This year, Google I/O 2025 and Microsoft Build 2025 offered their respective glimpses into the AI-powered world to come. While both keynotes leaned into the promise of artificial intelligence heavily, the approaches, the emphasis and ultimately the resonance with the audience painted a distinctly different picture. Drawing parallels to the perceived missteps of Microsoft Build that we discussed previously, let’s delve into which tech titan appeared to have its finger more firmly on the pulse of the AI revolution.

AI on Stage: A Matter of Show, Don’t Just Tell

Comparing the two keynotes, Google I/O 2025 undeniably showcased a more pervasive and tangible integration of AI into its present and near-future offerings. The sheer volume of AI-powered features and products unveiled by Google was striking. From the enhanced reasoning of Gemini (“Deep Think”) and its efficiency boost (“Gemini 2.5 Flash“) to the AI-driven enhancements across core services like Search (“AI Mode“) and Gmail (with its increasingly prescient “Smart Replies” and “Inbox Cleanup”), Google made it clear that AI is not just a future aspiration but a current reality woven into the fabric of its ecosystem. The introduction of entirely new AI-centric tools like Veo 3 for AI video generation (complete with audio, no less!), Imagen 4 for image generation that borders on photorealism, and Flow for AI filmmaking demonstrated a commitment to democratizing powerful creative tools. Even the ambitious “Project Astra,” a universal AI assistant aiming for true contextual understanding, and “Project Mariner,” an agentic AI promising web-browsing and task automation, felt like concrete steps towards a future brimming with intelligent assistance. Furthermore, the very presentation style of Google I/O often felt infused with AI, with seamless transitions, dynamic visuals and a flow that suggested an intelligent hand behind the curtain.

Microsoft Build 2025, however, while also heavily focused on AI, particularly the burgeoning concept of “AI agents,” presented a more developer-centric vision. The emphasis on tools and platforms like Copilot Studio and Windows AI Foundry, while crucial for building the next generation of intelligent applications, lacked the immediate, end-user impact that permeated Google’s presentation. The focus on infrastructure and the means for developers to create AI agents, while forward-looking, didn’t showcase the readily available AI advancements in Microsoft’s own products with the same tangible enthusiasm as Google’s product-driven approach. The presentation itself, in contrast to Google’s slick, AI-infused delivery, often felt like a more traditional PowerPoint presentation, relying on static slides rather than dynamic, AI-generated content. It echoed the sentiment from our previous discussion: a lot of talk about the AI engine, but not enough showcasing of the cars it’s already powering.

Gazing into the Crystal Ball: Vision and Forward-Thinking

When it came to painting a picture of the future, Google’s I/O 2025 arguably offered a more immediately compelling and, dare I say, visionary outlook. Its exploration of “world models” with Gemini, aiming to create AI with a more human-like understanding of and interaction with the real world, felt like a significant leap toward more intuitive and capable AI. The unveiling of “Google Beam,” a 3D video conferencing platform hinting at a more immersive and perhaps less socially awkward future of remote communication done in partnership with HP, and the advancements in Android XR, envisioning smart glasses with integrated AI capabilities, provided tangible glimpses into near-future technologies that could reshape how we interact with the digital and physical realms.

While Microsoft’s emphasis on the “open agentic web” – a potentially transformative concept centered on interconnected AI agents – certainly held long-term promise, its immediate impact felt less pronounced. The focus on developer tools, while laying the groundwork for future innovation, didn’t quite capture the imagination in the same way as Google’s showcased AI applications poised for near-term integration into everyday life. Microsoft’s vision, while perhaps more foundational for the enterprise space, lacked the broad, consumer-centric appeal that made Google’s presentation feel more immediately forward-looking.

The Tone Test: Reading the Room (and the Internet) in the Age of AI Anxiety

In terms of “tone-deafness,” Google’s I/O 2025 appeared to navigate the complexities of the AI revolution with relative aplomb. The focus on tangible advancements, coupled with the democratization of powerful AI tools, seemed to align with industry trends and a general consumer curiosity and cautious optimism surrounding AI’s potential.

Microsoft Build 2025, however, faced a potentially significant “tone check” moment, amplifying the undercurrent of job security anxieties we discussed previously. The reported disruption of Satya Nadella’s keynote by a Microsoft employee protesting the company’s cloud and AI contracts with the Israeli government highlighted a potential disconnect between the company’s public messaging and the concerns of at least some within its own ranks. But the tonal missteps didn’t end there. Microsoft’s decision to feature Sam Altman, CEO of OpenAI, a company with its own share of ethical and leadership controversies, and Elon Musk, a figure whose relationship with the tech community and governments is increasingly fraught, on stage, while ostensibly promoting open source, felt jarring and out of touch. This contrasted sharply with Google’s keynote, which did not feature such polarizing figures. Microsoft’s featuring individuals whose companies and personas are often seen as antithetical to the collaborative and ethical ideals of open source created a significant credibility gap. It was a stark reminder of the “unholy trinity” discussed earlier, a baffling choice given the anxieties surrounding AI’s impact on employment and the need for responsible innovation. The absence of such potentially alienating figures on Google’s stage further highlighted Microsoft’s misreading of the room. Moreover, while Google’s presentation felt dynamically AI-driven, Microsoft’s reliance on a more traditional format underscored a potential irony: a tech giant pitching the future of AI with a presentation style that felt decidedly…pre-AI.

Wrapping Up: A Fork in the Road for the AI Narrative

In the battle for the AI horizon, Google I/O 2025 and Microsoft Build 2025 presented contrasting visions. Google appeared to lead with tangible, user-facing AI advancements and a more immediate glimpse into an AI-integrated future, delivered with a presentation style that hinted at the very technology being showcased. And the fact that Google used smartphones for many of the presentations rather than PCs looked more convenient and advanced. 

Microsoft, while heavily invested in AI, focused more on the underlying tools and infrastructure for developers, with a potentially less immediately impactful vision for end-users, and a keynote that felt tonally disconnected from the anxieties of its audience and undermined by its choice of controversial guest speakers. As the AI revolution unfolds, the ability to not only innovate but also to connect with and address the real-world concerns of developers and the broader public with a consistent and authentic message will be crucial for long-term success and trust.

Watching both events reminded me of Microsoft and IBM in the early 90s when Microsoft was on the cutting edge and IBM seemed way out of touch. But Microsoft now appears more like IBM, stuck in the past and hampered by its platforms, while Google appears more like the old Microsoft, more forward looking and with a deeper understanding of the pivotal technology. Then it was PCs. Now it’s AI, and this doesn’t bode well for Microsoft. The two companies feel like they are in different centuries, Microsoft living in the 1900s while Google has advanced into the future of AI.  

Following the announcement that Gemini is coming to cars, Volvo is using I/O 2025 to announce a new expanded partnership with Google. The companies’ new deal makes Volvo’s cars reference hardware for future Android Automotive OS development, and means Volvo drivers will be “among the first to benefit” when Gemini fully replaces Google Assistant in cars.

Volvo describes itself as Google’s “lead development partner for new features and updates,” making the company’s cars the first to receive new updates to the Android Automotive OS. Google offers Android Auto as its CarPlay-like solution for beaming a software interface from your phone to in-car displays, but its Automotive OS is more complete, running on your vehicle locally and connected to car controls for A/C and more. You can already experience Android Automotive OS in Volvo’s EX90, for example.

Google’s current vision for Android in cars is, perhaps unsurprisingly, focused on getting drivers to talk to Gemini. In a car with the assistant, you’ll be able to ask Gemini to send a message, pull up directions, or answer the more open-ended, natural language questions that Gemini Live is designed to handle. If it works as advertised, it seems better than pecking at a screen, and Volvo notes it could “help reduce your cognitive load so that you can stay focused on driving.”

There’s no release date for when you can expect Gemini to show up as your driving copilot, but at the very least this new partnership means it’ll be in Volvos first.

This article originally appeared on Engadget at https://www.engadget.com/transportation/volvo-expands-its-google-partnership-to-bring-new-features-like-gemini-to-cars-sooner-070020853.html?src=rss

A new study reveals that about 3.8 million years after the solar system’s first solids formed, Jupiter was twice its current size with a magnetic field 50 times stronger, profoundly influencing the structure of the early solar system. Phys.Org reports: [Konstantin Batygin, professor of planetary science at Caltech] and [Fred C. Adams, professor of physics and astronomy at the University of Michigan] approached this question by studying Jupiter’s tiny moons Amalthea and Thebe, which orbit even closer to Jupiter than Io, the smallest and nearest of the planet’s four large Galilean moons. Because Amalthea and Thebe have slightly tilted orbits, Batygin and Adams analyzed these small orbital discrepancies to calculate Jupiter’s original size: approximately twice its current radius, with a predicted volume that is the equivalent of over 2,000 Earths. The researchers also determined that Jupiter’s magnetic field at that time was approximately 50 times stronger than it is today.

Adams highlights the remarkable imprint the past has left on today’s solar system: “It’s astonishing that even after 4.5 billion years, enough clues remain to let us reconstruct Jupiter’s physical state at the dawn of its existence.” Importantly, these insights were achieved through independent constraints that bypass traditional uncertainties in planetary formation models — which often rely on assumptions about gas opacity, accretion rate, or the mass of the heavy element core. Instead, the team focused on the orbital dynamics of Jupiter’s moons and the conservation of the planet’s angular momentum — quantities that are directly measurable.

Their analysis establishes a clear snapshot of Jupiter at the moment the surrounding solar nebula evaporated, a pivotal transition point when the building materials for planet formation disappeared and the primordial architecture of the solar system was locked in. The results add crucial details to existing planet formation theories, which suggest that Jupiter and other giant planets around other stars formed via core accretion, a process by which a rocky and icy core rapidly gathers gas.
The findings have been published in the journal Nature Astronomy.

Read more of this story at Slashdot.

Fortnite has returned to Apple’s App Store in the United States after a nearly five-year absence, marking a significant victory for Epic Games in its protracted legal battle against Apple’s App Store policies. The return follows an April 30 ruling where a federal judge determined Apple violated a court order requiring the company to allow greater competition for app downloads and payment methods, referring Apple to federal prosecutors for a criminal contempt investigation.

Epic CEO Tim Sweeney celebrated on X with a simple “We back fam” message. The game, which had 116 million users on Apple’s platform before its 2020 removal, was banned after Epic challenged Apple’s practice of charging up to 30% commission on in-app payments as anticompetitive.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Reuters: Honda said on Tuesday that it was scaling back its investment in electric vehicles given slowing demand and would be focusing on hybrids, now far more in favor, with a slew of revamped models. Japan’s second-biggest automaker after Toyota also dropped a target for EV sales to account for 30% of its sales by the 2030 financial year. “It’s really hard to read the market, but at the moment we see EVs accounting for about a fifth by then,” CEO Toshihiro Mibe told a press conference.

Honda has slashed its planned investment in electrification and software by that year by 30% to 7 trillion yen ($48.4 billion). It’s one of a number of global car brands dialing back EV investment due to the shift in demand in favor of hybrids and as governments around the world ease timelines to meet emission rules and EV sales targets. Honda plans to launch 13 next-generation hybrid models globally in the four years from 2027. At the moment it sells more than a dozen hybrid models worldwide, though just three in the U.S. — the Civic, which comes in hatchback and sedan versions, the Accord and the CR-V. It will also develop a hybrid system for large-size models that it plans to launch in the second half of the decade.

The automaker is aiming to sell 2.2 million to 2.3 million hybrid vehicles by 2030, a huge jump from 868,000 sold last year. That also compares with a total of 3.8 million vehicles sold overall last year. Earlier this month, Honda announced it had put on hold for about two years a $10.7 billion plan to build an EV production base in Ontario, Canada, due to slowing demand for electric cars. Honda said, however, that it still plans to have battery-powered and fuel-cell vehicles make up all of its new car sales by 2040.

Read more of this story at Slashdot.

AMD’s Radeon RX 9070 XT and RX 9070 were fairly well received when they were released in March, ably competing with Nvidia’s RTX 5070 and RTX 5070 Ti for the same or a little less money. We were impressed by the cards’ performance and power efficiency, even if they still have some of the same caveats as older Radeon cards (lack of DLSS upscaling and lower relative ray-tracing performance being two).

Today AMD is formally expanding its family of RDNA 4 graphics cards with the Radeon RX 9060 XT, a GPU that will go up against Nvidia’s RTX 5060 and RTX 5060 Ti GPUs. These GPUs have just half the compute units of the RX 9070 XT, but at $299 and $349 for 8GB and 16GB configurations, they ought to be decent options for 1080p or entry-level 1440p gaming PCs (with the eternal “if you can find them” caveat that comes with buying a GPU in 2025).

AMD says the new GPUs will be available starting on June 5th from the typical range of partners—AMD released renders of a reference GPU design, but sometimes these are starting points that manufacturers can take or leave, rather than products AMD intends to manufacture and sell itself.

Read full article

Comments

AMD has unveiled its 9060 XT GPU at Computex 2025. The midrange GPU will be the clear competitor to Nvidia’s 5060 Ti and goes toe-to-toe with it on almost every spec. Built on AMD’s 4-nanometer RDNA 4 silicon, the 9060 XT will pack 32 compute units, along with 64 dedicated AI accelerators and 32 ray-tracing cores.

AMD

Notably, the RX 9060 XT will ship in 8GB and 16GB GDDR6 versions, whereas Nvidia’s RTX 5060 Ti uses faster 28 Gb/s GDDR7, delivering roughly 40 percent more bandwidth (448 GB/s vs. approximately 322 GB/s) on the same 128-bit bus. We’ll have to wait for some side-by-side performance comparisons before drawing any strong conclusions from those specs.

AMD has listed the 9060 XT’s boost clock at speeds up to 3.13 GHz. The GPU boasts 821 TOPS for AI workloads and will draw a modest 150 to 182 watts from the board. The card will connect via PCIe 5.0 x16 and supports the now-standard DisplayPort 2.1a and HDMI 2.1b. Based on these initial specs, the 9060 XT should be a solid entry for games running at 1080p and a decent option for those at 1440p. Those wishing to play at 4K should still opt for the Radeon RX 9070 or 9070 XT.

Pricing and exact release timelines have not yet been announced.

This article originally appeared on Engadget at https://www.engadget.com/gaming/pc/amd-unveils-radeon-rx-9060-xt-at-computex-2025-030021776.html?src=rss

Not many people need a 96-core processor. But for creative professionals, engineers and AI developers who do, AMD has a new batch of chips on display at Computex 2025. The company announced its new Ryzen Threadripper 9000 series on Tuesday, with bonkers specs to power pro-level workstations and ultra-high-end prosumer desktops.

At the top of the line in the series is the AMD Threadripper Pro 9995WX. This chip has a staggering 96 cores and 192 threads, matching the highest-end model from 2023’s Threadripper Pro 7000 line. But the new 9000 series tops out with a higher maximum boost speed of 5.4GHz. That’s up from 5.1GHz in the premiere 7000 Pro chip.

AMD’s new batch includes six processors in the Threadripper Pro WX series, designed for pro-level workstations. (In addition to the 96-core 9995WX, options include 12-, 16-, 24-, 32- and 64-core models.) Moving past the Pro series, the standard Threadripper 9000 line for high-end desktops maxes out with the 64-core, 128-thread 9980X.

AMD hasn’t yet announced pricing or specific retail models carrying the chips. But the 7000 Pro series offers a hint. The top-shelf model from that line costs a cool $10,000. (Yep, that’s for the processor alone.) So, unless your work involves extremely demanding AI development, 3D modeling or ultra-high-res video editing, you can slowly step away and make your way back to the consumer aisle.

This article originally appeared on Engadget at https://www.engadget.com/computing/amds-ryzen-threadripper-9000-chips-have-up-to-96-cores-just-like-the-last-bunch-030003537.html?src=rss

A Massachusetts man has agreed to plead guilty to hacking into one of the top education tech companies in the United States and stealing tens of millions of schoolchildren’s personal information for profit. From a report: Matthew Lane, 19, of Worcester County, Massachusetts, signed a plea agreement related to charges connected to a major hack on an educational technology company last year, as well as another company, according to court documents published Tuesday.

While the documents refer to the education company only as “Victim-2” and the U.S. attorney’s office declined to name the victim, a person familiar with the matter told NBC News that it is PowerSchool. The hack of PowerSchool last year is believed to be the largest breach of American children’s sensitive data to date.

According to his plea agreement, Lane admitted obtaining information from a protected computer and aggravated identity theft and agreed not to challenge a prison sentence shorter than nine years and four months. He got access simply by trying an employee’s stolen username and password combination, the complaint says, echoing a private third-party assessment of the incident previously reported by NBC News.

Read more of this story at Slashdot.

A new virtual machine manager called Karton is being developed specifically for the KDE Plasma desktop, aiming to offer a seamless, Qt-native alternative to GNOME-centric tools like GNOME Boxes. Spearheaded by University of Waterloo student Derek Lin as part of Google Summer of Code 2025, Karton uses libvirt and Qt Quick to build a user-friendly, fully integrated VM experience, with features like a custom SPICE viewer, snapshot support, and a mobile-friendly UI expected by September 2025. Neowin reports: To feel right at home in KDE, Karton is being built with Qt Quick and Kirigami. It uses the libvirt API to handle virtual machines and could eventually work across different platforms. Right now, development is focused on getting the core parts in place. Lin is working on a new domain installer that ditches direct virt-install calls in favor of libosinfo, which helps detect OS images and generate the right libvirt XML for setting up virtual machines more precisely. He’s still refining device configuration and working on broader hypervisor support. Another key part of the work is building a custom SPICE viewer using Qt Quick from scratch:

If you’re curious, here’s the list of specific deliverables Lin included in his GSoC proposal, though he notes the proposal itself is a bit outdated […]. For those interested in the timeline, Lin’s GSoC proposal says the official GSoC coding starts June 2, 2025. The goal is to have a working app ready by the midterm evaluation around July 14, 2025, with the final submission due September 1, 2025. You can learn more via KDE.org.

Read more of this story at Slashdot.