As the Integrity spacecraft, carrying four astronauts on board, splashed down in the Pacific Ocean off the coast of San Diego, California, it marked a momentous occasion in the history of space exploration. The successful landing of the Orion craft, after an 10-day journey around the Moon, marked a major milestone in NASA’s Artemis II mission, which aimed to test the capabilities of the Space Launch System (SLS) and Orion spacecraft.

The development of solid-state batteries has long been touted as the “Holy Grail” of battery technology, offering a solution to the limitations of traditional lithium-ion batteries. With their promise of high-density, durable, and fast-charging capabilities, solid-state batteries have captured the imagination of experts and enthusiasts alike. Recently, Finnish startup Donut Lab claimed to have achieved a breakthrough in this field, sparking both excitement and skepticism.

The debate surrounding Donut Lab’s achievement is not new; indeed, many experts have been skeptical about the potential for commercialization of solid-state batteries. The lack of transparency and verifiable research from the startup has only fueled these doubts. However, Donut Lab’s decision to launch a website dedicated to independent testing results from VTT Technical Research Centre of Finland has provided some level of assurance that their claims are more than just hype.

As the technology landscape continues to evolve, Google’s AI Mode has emerged as a game-changer in the world of artificial intelligence. The latest update brings not only a redesigned interface but also expanded capabilities that cater to users worldwide. In this article, we’ll delve into the details of this significant development and explore its implications for the future of AI-driven applications.

The revamped prompt box, reminiscent of Google’s Gemini app, offers a sleeker and more user-friendly experience. Gone is the traditional pop-up menu on mobile devices, replaced by a bottom sheet that provides easy access to key features. The “Tools” section has been streamlined, with “Create images” being the only option available on mobile devices at present. The “Gemini 3 models” switcher remains intact, allowing users to toggle between Auto and Pro modes.

The latest update to Google Messages for Android has brought a much-needed feature to the table - a Trash folder, designed to serve as a safety net against accidental deletions. This change comes on the heels of last month’s announcement, and it’s now widely available with stable version 20260327_00_RC00.

One of the most significant benefits of this new feature is its “grace period” - conversations will be deleted permanently only after 30 days (or 7 days on Android Go devices to conserve storage space). This means that users have a month’s time to recover or delete individual threads, giving them greater control over their digital communication history. The Trash folder can be accessed from the account menu by tapping the profile image in the top-right corner, where it appears alongside the existing “Archived” category.

In the latest episode of Pixelated, a podcast by 9to5Google, hosts Damien, Abner, and Will delved deep into the world of Gemini, dissecting Google’s frequent user interface (UI) changes and exploring how the service can evolve to better meet user needs. The discussion begins with an examination of Gemini’s UI redesigns, which have been a staple of Google’s product updates in recent times.

The trio dives into their individual experiences using Gemini, highlighting both the benefits and drawbacks of this dynamic service. Damien shares his own experience using Gemini with Android Auto, revealing that it’s a mixed bag - while some features work seamlessly, others leave room for improvement. This nuanced perspective underscores the complexity of user adoption and highlights the need for ongoing refinement.

The age-old adage “nothing is certain except death and taxes” has been replaced by a new truism in the digital era: “nothing is private except for what you don’t know about.” The latest salvo in the ongoing battle between privacy and progress comes in the form of a class-action lawsuit filed in San Francisco’s federal court. The suit, brought forth by several Californians, alleges that Sutter Health and MemorialCare have been using an AI transcription tool called Abridge to record doctor visits without patients’ consent, raising serious concerns about privacy violations.

After a thrilling journey around the Moon, NASA’s Orion spacecraft successfully splashed down in the Pacific Ocean on Friday, marking a major milestone in human spaceflight history. The spacecraft, named Integrity, returned with its four astronauts, Commander Reid Wiseman, Pilot Victor Glover, and Mission Specialists Christina Koch and Jeremy Hansen, safely ending their voyage to the Moon.

The journey home was nothing short of remarkable. As Orion re-entered Earth’s atmosphere, temperatures outside the capsule reached a scorching 5,000 degrees Fahrenheit, causing a plasma sheath to form around the spacecraft and temporarily disrupt radio communications with Mission Control in Houston. The spacecraft then steered towards its splashdown zone southwest of San Diego, where a US Navy recovery ship, USS John P. Murtha, awaited the crew’s homecoming.

The recent banishment of Peter Steinberger, creator of OpenClaw, from accessing Claude by Anthropic has sent shockwaves through the AI model community. The temporary suspension was met with a mixture of outrage and speculation, with many questioning the true motivations behind Anthropic’s decision. At the heart of the controversy lies the evolving landscape of AI models and their relationships with developers.

The ban

Steinberger’s frustration with Anthropic is nothing new. He has long been vocal about the company’s closed nature and perceived attempts to stifle innovation. The recent events may be seen as part of a larger narrative, where Steinberger’s decision to join OpenAI, an Anthropic rival, has led to increased scrutiny and tension between the two companies.

The Artemis program, NASA’s ambitious plan to return humans to the moon by 2024, hangs precariously in the balance as it faces its most daunting challenge yet: reentry. The final stage of a lunar mission, which lasts mere minutes, has the potential to make or break the entire program.

For astronauts, engineers, and mission managers at NASA, reentry is not just a minor hurdle but rather the most critical moment of a trip to the moon. The Orion capsule’s return journey from the moon will reach speeds of 11 kilometers per second, roughly twice that of a traditional reentry from the International Space Station. This incredible velocity generates intense friction with the atmosphere, creating a plasma cocoon where temperatures soar to an astonishing 2,700 degrees Celsius (4,900 Fahrenheit).

The recent announcement by Anthropic of its Mythos Preview model has sent shockwaves throughout the cybersecurity community, sparking a heated debate about whether this represents a turning point in the evolution of software defense strategies or simply another iteration of AI hype. As we delve into the implications of this technology, it’s essential to separate fact from fiction and explore the potential consequences for organizations and individuals alike.

At its core, Mythos Preview is a generative AI model capable of discovering vulnerabilities in virtually any operating system, browser, or software product and autonomously developing working exploits for hacking. This level of capability has significant implications for existing cybersecurity strategies, which rely heavily on manual detection and patching processes. The fact that Anthropic is only releasing this model to a select few organizations as part of the Project Glasswing consortium underscores the gravity of the situation.