#BrainUp Daily Tech News – (Wednesday, March 11ᵗʰ)
Welcome to today’s curated collection of interesting links and insights for 2026/03/11. Our Hand-picked, AI-optimized system has processed and summarized 25 articles from all over the internet to bring you the latest technology news.
As previously aired🔴LIVE on Clubhouse, Chatter Social, Instagram, Twitch, X, YouTube, and TikTok.
Also available as a #Podcast on Apple 📻, Spotify🛜, Anghami, and Amazon🎧 or anywhere else you listen to podcasts.
1. YouTube surpasses Disney, Paramount, WBD in 2025 ad revenue | TechCrunch
#YouTube surpassed major Hollywood studios in 2025 #ad-revenue, signaling how audience attention and advertiser priorities have shifted toward digital platforms. Estimates from Moffett Nathanson, cited by The Hollywood Reporter, put YouTube’s 2025 ad revenue at $40.4B, beating the combined $37.8B for @Disney, @NBC, @Paramount, and @WarnerBrosDiscovery, reversing 2024 when YouTube’s $36.1B trailed the studios’ combined $41.8B. The article links the turnaround to shrinking linear TV audiences and rising production costs that challenge traditional studios, while YouTube keeps building momentum even as those companies invest in streaming. @Alphabet also reported YouTube’s total 2025 revenue hit $60B, with growing contributions from subscriptions like YouTube TV, YouTube Premium, YouTube Music, and NFL Sunday Ticket, though its ad business still trails @Meta’s $196.2B. As it scales, YouTube is also investing in #AI, expanding likeness-detection to help identify #deepfakes and support removal requests under YouTube policy.
A new study argues that the lunar near side region Rimae Bode is a strong candidate landing site for China’s first crewed moon mission because it combines high scientific value with safe landing conditions. China is targeting its first astronaut landing before 2030 and has been testing key mission hardware, while @Jun Huang and colleagues analyzed multiple orbital image and data sets to evaluate Rimae Bode in a paper published March 9 in Nature Astronomy. The site is one of 14 shortlisted from 106 candidates that meet engineering constraints such as near side communications, relatively flat terrain, and low latitude for sufficient solar power. The researchers report that astronauts using an unpressurized rover could access diverse materials in a small traversable area, including ancient lava flows, rilles formed by lava, and crater ejecta, and they identify four feasible landing sites with different sampling priorities. Described as a “geological museum,” Rimae Bode could let a single mission collect samples ranging from deep interior volcanic ash to debris from ancient impacts, potentially offering insights into the moon’s deep interior while fitting mission safety needs.
3. Meta Acquires Moltbook, a New Platform for Social AI Bots
Meta recently acquired Moltbook, a platform that enables users to create and interact with social AI bots designed for personalized and dynamic digital conversations. The platform offers tools that allow these bots to develop unique personalities and simulate human-like social interactions, enhancing user engagement and digital expression. By integrating #AI models capable of learning and adapting through conversation, Moltbook aims to transform social media interactions and provide an innovative way for users to connect. This initiative reflects Meta’s broader strategy to embed #ArtificialIntelligence within social platforms to increase interactivity and personalization. The introduction of Moltbook highlights the evolving landscape of social technology and underscores the importance of AI-driven experiences in future online communication.
4. Google to Provide Pentagon with Gemini-powered AI agents
#Google will roll out #Gemini powered AI agents to the Department of Defense, aiming to support more than 3 million civilian and military employees initially on unclassified networks, with discussions to extend access to classified and top secret systems. According to @Bloomberg and statements from Emil Michael and @Google VP Jim Kelly, eight pre built agents will automate tasks such as summarizing meeting notes, building budgets, and checking proposed actions against the national defense strategy, and staff will also be able to create custom agents using natural language. The Pentagon’s #GenAI.mil portal has already seen 1.2 million users since December, with 40 million unique prompts and over 4 million documents uploaded, though only 26,000 employees have completed AI training so far. The expansion reflects the Pentagon’s rapid broadening of AI partnerships after a standoff with @Anthropic over guardrails for domestic surveillance and autonomous weapons, alongside new deals with @OpenAI and xAI, and it comes as Google has faced and navigated internal controversy over military AI work since #Project Maven in 2018 and has since loosened restrictions and revised its #AI Principles.
5. Asus Co-CEO: MacBook Neo Is a ‘Shock’ to the PC Industry
@Asus co-CEO @S.Y. Hsu says Apple’s MacBook Neo, starting at $599, is a “shock” that forces the Windows PC industry and the broader PC ecosystem, including @Microsoft, @Intel, and @AMD, to consider how to compete. He argues the Neo may have limited appeal because its 8GB of #unified memory cannot be upgraded and he frames it as a “content consumption” device closer to an iPad than a mainstream notebook for heavier computing. Even so, PCMag’s review scored it “outstanding,” citing fast basic performance, light gaming capability, and strong early demand as preorders began ahead of a March 11 launch and shipping estimates slipped by weeks. Hsu adds that software differences may limit adoption among Windows users, but expects Windows PC makers to launch competing products. Competitive responses may be complicated by an #AI-driven memory shortage, with Asus saying memory prices rose over 100% quarter over quarter and warning it may need to reassess pricing as supplies run out, with the crunch expected to last until new fabs come online in late 2027.
6. Apple’s Low-Cost iPhone 17e is Faster Than the Low-Cost MacBook Neo
Apple is set to launch the $599 iPhone 17e and $599 MacBook Neo, and early benchmarks indicate the iPhone is faster on CPU performance because it uses a newer A19 chip versus the MacBook Neo’s A18 Pro. On #Geekbench, the iPhone 17e scored 9,241 multi-core and 3,607 single-core, compared with the MacBook Neo’s 8,668 multi-core and 3,461 single-core. GPU results in #Metal were close, with the MacBook Neo around 30,000 to 31,400 and the iPhone 17e around 31,000 to 31,600. Both devices include 8GB RAM for #AppleIntelligence support, and early reviewers said 8GB was adequate for everyday light workloads. As Apple’s first Mac built around an #A-series chip rather than an #M-series chip, the MacBook Neo benchmarks suggest it is essentially an iPhone running macOS, raising questions about how well it will sell given it trails Apple’s low-cost iPhone on CPU.
7. ‘Flying Cars’ Will Take Off in American Skies This Summer
The US Department of Transportation says new ultralight aircraft and #eVTOL “flying cars” could begin operating in US airspace as early as June through a three-year pilot program, even before they are fully #FAA certified. Eight regions, including New York and New Jersey, Texas, Florida, and Albuquerque, will host demonstrations that move people and cargo using aircraft designed to take off and land with minimal space, potentially outside traditional airports. Companies involved, including Archer Aviation, Beta Technologies, Joby Aviation, and Electra, argue these vehicles can be quieter, cheaper, and lower-emitting than helicopters, with some promising autonomous trips, and leaders like @Adam Goldstein call the effort a “Waymo moment” for normalizing air taxis. Federal policy is positioned as supportive, from a @Donald Trump executive order aimed at cutting red tape for drones and “flying cars” to a 2023 Biden-era plan targeting air taxi deployment by 2028, but certification remains a major hurdle because new rules and safety standards are required. The FAA emphasizes the pilot is meant to inform standards and future policy, not bypass certification, and participating aircraft must already be in the formal type certification process.
8. Oracle moves to assure MySQL community it really does care
@Oracle says it will make #MySQL development more transparent and accelerate innovation in the core, positioning MySQL as fundamental to its data strategy while prioritizing new capabilities like #vector support. In a blog post co-authored by Heather VanCura and Lenka Kasparova, Oracle listed priorities including AI and cloud alignment such as vector, developer experience features, performance such as scaling, observability, extensibility, and ecosystem tooling and connectors, with features selected on community principles and re-evaluated each release cycle. The update responds to community criticism that development has been opaque with private code drops, limited roadmap visibility, falling commits, job losses on the MySQL core team, and missing modern features like vector search as PostgreSQL and many commercial databases already offer it. Influential users and developers have also invited Oracle to join plans for an independent nonprofit foundation to guide MySQL, citing declining popularity versus PostgreSQL and difficulty attracting new contributors. The group reportedly welcomed what Oracle shared but wants concrete timelines, with more detail expected before the end of the month.
Nvidia is reportedly developing an #AI agent called #NemoClaw to compete with #OpenClaw, aiming to help enterprise users run AI tools with more autonomy while meeting corporate security and privacy needs. Wired reports it is intended to be open source, customizable, offered to partners such as Adobe, Cisco, CrowdStrike, Google, and Salesforce, and designed to run on any hardware rather than requiring Nvidia chips. The article argues that while generative AI can reason, agents are needed to orchestrate and execute tasks independently, and #OpenClaw popularized broad agent use across LLMs but has also seen issues like malicious skills uploaded to ClawHub and a reported incident where @Summer Yue said an agent deleted emails despite constraints. Nvidia is positioned as trying to avoid those problems and to capture the corporate market early, especially as @OpenAI hired #OpenClaw creator @Peter Steinberger in February 2026 to work on smart agents. Overall, #NemoClaw is framed as Nvidia’s enterprise-focused, open-source entry into the fast-moving agent ecosystem shaped by #OpenClaw’s popularity and associated risks.
@SK hynix announced it has successfully developed its first #LPDDR6 DRAM, claiming higher performance and efficiency than #LPDDR5X and positioning it for mobile devices and AI-focused servers. The company says the 16Gb chips run at over 10.7Gbps and deliver 33% greater speed plus more than 20% lower power use, enabled by a new sub-channel structure that powers only active data paths and #DVFS that lowers clock speed and voltage under light demand, all built on its 10nm-class 1c process. Coming about eight months after #JEDEC finalized the LPDDR6 standard, the product enters a market where @Samsung has already shown LPDDR6 at CES 2026 at up to 10.7Gbps, while SK hynix highlights broader adoption as designs mature. The article links LPDDR6 to the datacenter trend toward #SOCAMM and #SOCAMM2 LPDDR-only modules used in @Nvidia AI platforms like GB300 Grace Blackwell Ultra and Vera Rubin, and notes SK hynix expects post-Vera Rubin Nvidia designs to use LPDDR6. Looking ahead, it says LPDDR6 speeds should climb beyond 10.7Gbps, with JEDEC projecting up to 14,400 MT/s as optimization improves.
11. Miniature laser technology brings a science lab to your home
Researchers have developed miniature laser technology that enables the creation of compact and efficient laser systems suitable for home use. The innovation uses advanced optical components and integration techniques to significantly reduce the size and cost compared to traditional laser setups. This breakthrough allows for broader accessibility in scientific experimentation and practical applications, promoting hands-on learning and technological exploration outside specialized laboratories. By miniaturizing these systems, the technology supports a new wave of educational tools and research equipment, expanding opportunities for individuals to engage directly with laser science. This advancement bridges complex laboratory methods with everyday environments, fostering innovation and understanding in the wider community.
12. Anthropic’s AI Exposure Study: What the Data Means for Your Workforce | EM Solutions
@Anthropic’s March 5, 2026 labor market study argues that the most important signal for workforce planning is the gap between #LLM theoretical capability and what AI is actually doing on the job, measured with its new #observedExposure metric. Researchers Maxim Massenkoff and Peter McCrory combine feasibility scores from Eloundou et al. (2023), whether an LLM can halve task time, with real professional Claude usage from the Anthropic Economic Index, counting tasks only when both feasible and observed, and weighting automated use more than augmentative; this approach correlates with #BLS employment projections across 800 occupations through 2034, unlike purely theoretical models. The standout example is Computer and Math work at 94% theoretical exposure but only 33% observed exposure, a 61 point gap that suggests today’s limited impact reflects deployment frictions, not a low disruption ceiling; across current Claude professional usage, 97% of observed tasks are within feasible territory, implying the bottleneck is deployment. Similar gaps appear in Office and Administrative roles, while about 30% of the U.S. workforce shows zero observed exposure in physically present jobs like cooks, mechanics, bartenders, and dishwashers. Among the most exposed roles, Computer Programmers lead at 75% observed exposure, Customer Service Representatives rank high due to automated query handling seen in API traffic, and Data Entry Keyers reach 67%, and the framework is described as sensitive enough to detect a Great Recession scale shock in high exposure occupations, which has not occurred so far.
13. Oracle stock jumps 9% on earnings beat and increased guidance as cloud revenue climbs 44%
Oracle shares rose as much as 10% in extended trading after the company beat quarterly estimates and raised guidance, driven by surging #cloud revenue. It reported $1.79 in adjusted EPS on $17.19 billion in revenue versus LSEG expectations of $1.70 and $16.91 billion, with total revenue up 22% year over year and net income rising to $3.72 billion. Cloud revenue totaled $8.9 billion, up 44%, including $4.9 billion in cloud infrastructure revenue that jumped 84%, and Oracle cited business from Air France-KLM, Lockheed Martin, SoftBank Corp., and @Microsoft’s Activision Blizzard unit. Management lifted its fiscal 2027 revenue outlook by $1 billion to $90 billion and guided fiscal Q4 adjusted EPS of $1.92 to $1.96 with 19% to 20% revenue growth, while remaining performance obligations more than quadrupled to $553 billion. The report comes as Oracle invests heavily in AI-driven infrastructure expansion, plans to raise $45 billion to $50 billion to add capacity, and faces investor concerns about debt and negative free cash flow even as leaders like @Larry Ellison argue #agent-based software and SaaS strength position Oracle as a disruptor.
14. Oracle Stock Jumps After Earnings Beat. AI Powers Stronger-Than-Expected Cloud Growth..
@Oracle reported strong fiscal Q3 2026 earnings, exceeding Wall Street expectations as its #AI-driven cloud infrastructure business accelerated sharply, posting $1.79 earnings per share on $17.2 billion in revenue, both above analyst forecasts and representing about 22 percent year-over-year growth. The company’s #OracleCloudInfrastructure division was the standout performer, with revenue surging 84 percent to roughly $4.9 billion as demand for AI training and inference workloads continues to expand across enterprises building large-scale AI systems. Oracle also revealed a massive $553 billion backlog of contracted cloud services, a figure that jumped more than 300 percent year over year, highlighting how companies are committing to long-term infrastructure deals to power AI workloads. Despite investor concerns over heavy spending on AI data centers and a plan to raise as much as $50 billion in financing to expand infrastructure capacity, the strong results boosted confidence that Oracle’s strategy of building large AI-focused cloud platforms is paying off. The performance underscores how the AI boom is reshaping the cloud market, positioning Oracle as a rapidly growing competitor to hyperscale providers while embedding #AI capabilities deeper into its enterprise software and database ecosystem.
15. Security Experts Warn of Privacy & Identity Theft Risks Following Meta Ray-Ban Nairobi Case
Meta faces scrutiny and a class action lawsuit after reports that contractors in Nairobi accessed private, intimate videos captured by Meta Ray-Ban smart glasses, including footage showing private financial information, sometimes recorded without wearers’ knowledge. The article says the videos were reviewed as part of #AI training, where contractors label objects to help Meta’s AI recognize real-world scenes. Security experts warn that exposing sensitive recordings, even inadvertently, can create serious #privacy and #identityTheft risks, including #phishing, because personal data used for AI could surface again. @Melissa Ruzzi of AppOmni notes that companies often disclose data use in terms of service, but many users do not read them, leaving them unaware of the risks. The case also intensifies broader concerns about smart glasses recording bystanders and private spaces without consent, prompting some workplace bans and investigations by European regulators, including the UK’s #InformationCommissionersOffice, into potential privacy-law violations.
16. Move Over, Mercedes S-Class: Your Next Black Car Seats Eight, And It’s An Electric Van
The UK government is transitioning its official state car from a traditional luxury sedan like the Mercedes S-Class to an electric van that seats eight, reflecting a shift towards sustainability and practicality. This move replaces the iconic black car after decades, emphasizing cleaner transportation and increased capacity for officials. The electric van signifies a modern approach to #greenmobility and governmental image, promoting reduced emissions and functionality without compromising on status or security. By adopting this change, the government aligns with broader environmental goals and the evolving expectations of official vehicles. This development highlights the growing acceptance of electric vehicles in high-profile, formal roles.
17. VPN apps rocket up download charts in Australia as porn websites begin blocking users
VPN app downloads have surged in Australia as adult websites start restricting access to comply with new #online safety codes requiring #age verification from Monday. Sensor Tower data shows VPN Super Unlimited Proxy rose from 40th to 7th in free iPhone apps between 2 March and Sunday, while Proton VPN jumped from 174th to 19th and NordVPN from 189th to 13th, as users try to make their location appear outside Australia. Aylo-owned sites such as RedTube, YouPorn and Tube8 displayed notices that they were not accepting new registrations from Australian IP addresses, and Pornhub showed only safe-for-work content to Australian users who were not logged in. The rules also cover other services, including AI companion chatbots, app stores and social media that allows adult content, with @Elon Musk’s X reportedly prompting repeated age checks for adult posts and restricting Grok from generating blocked content until an age check is passed. The eSafety commissioner @Julie Inman Grant said simple self-declaration is no longer sufficient and warned non-compliant platforms could face fines up to $49.5m per breach, while Electronic Frontiers Australia’s John Pane said the VPN spike was expected based on similar past restrictions.
18. Silicon Valley is buzzing about this new idea: AI compute as compensation
Silicon Valley compensation packages may be adding a new line item alongside salary, bonus, and equity: access to #AI inference compute, because embedded generative AI tools make model-running costs a key productivity driver. Evidence includes engineers and researchers already competing internally for GPUs, job candidates asking how much dedicated inference they will get for tools like OpenAI’s #Codex, and @Thibault Sottiaux noting per-user usage rising faster than user growth, implying increasing scarcity and value. @Greg Brockman argues the inference compute available to an engineer will increasingly determine overall software productivity, making AI access consequential for output and career prospects. Early signals show up in pay disclosures such as a Levels.fyi submission listing a #Copilot subscription as a benefit, and proposals like Peter Gostev’s idea for job listings that publish token budgets alongside salary ranges. Investors like Tomasz Tunguz describe tokens as the economic language of generative AI and frame inference as a fourth compensation component, while finance leaders must treat inference costs as a significant budget item.
19. YouTube opens deepfake detection tool to politicians and journalists
@YouTube is expanding access to its free #deepfake likeness detection tool to government officials, journalists, and political candidates to help them identify and request removal of #AI-generated videos that resemble them. The company said the tool is meant to protect identities of people at the center of breaking news and civic discourse, after an earlier rollout to #YouTubePartnerProgram members in October 2025. Eligible users will be contacted to enroll, must submit a video of themselves and government ID, and will receive notifications in YouTube Studio when detected videos match their likeness, which they can flag for removal. @YouTube framed the move as balancing identity protection with free expression, including preserving parody and satire, as AI video advances have also enabled misinformation and scams. The company said participant data will not be used to train @Google AI models and it plans to significantly broaden access over the next year.
Thirteen major publishers including @Penguin Random House, @Elsevier, and @HarperCollins sued Anna’s Archive in New York federal court, alleging #copyright infringement on a massive scale and seeking an injunction aimed at taking the shadow library offline. The complaint claims the site hosts about 63 million books and 95 million papers, facilitates predominantly unauthorized downloading, and cited the site’s own statistics showing 763,000 downloads in a single day. Publishers also point to an #AI training angle, alleging Anna’s Archive promoted high-speed access to 140+ million texts for LLM developers and offered premium access for $200,000, referencing a LinkedIn post and an email exchange with a researcher. While the lawsuit lists 130 copyrighted works and seeks up to $19.5 million in statutory damages, it argues that identifying or collecting from unknown operators is unlikely, so the key remedy is an injunction that can pressure technical infrastructure. The requested order would require Anna’s Archive to stop infringing and destroy pirated materials, and would also compel third-party intermediaries such as hosts, registrars, and registries to cease providing services, following recent domain suspensions that occurred after a separate music industry injunction.
21. X suspends 800m accounts in one year amid ‘massive’ scale of manipulation attempts
@Elon Musk’s X says it suspended 800 million accounts over a 12-month period as it confronts a “massive” level of #platformManipulation and #spam. In evidence to UK MPs, X Corp government affairs executive Wifredo Fernández said state-backed actors repeatedly try to build inauthentic networks, naming Russia as the most prolific, followed by Iran and China, and stating Russia aimed to undermine the 2024 US presidential election and stoke division by “flooding the zone” with specific narratives. X said the 800 million suspensions in 2024 were for breaches of its manipulation and spam rules, but it did not specify how many were linked to foreign interference, and it estimates about 300 million monthly users worldwide. Fernández said takedowns continued in the latter part of last year and expressed confidence that remaining accounts are authentic, while X defines manipulative accounts as bulk, aggressive, or disruptive activity that misleads or disrupts users, and spam as unsolicited repeated actions that affect other accounts. The disclosure comes amid criticism of X’s content moderation since @Musk acquired the platform in 2022, alongside long-running concerns about spam and account authenticity that featured in the takeover dispute.
22. Gemini in Chrome gets first international expansion
#Gemini in Chrome is expanding internationally, adding availability in Canada, India, and New Zealand and support for over 50 languages. It can be launched from Chrome’s top right, keyboard shortcuts, the Mac menu bar, or the Windows system tray, and it runs as a floating window or a side panel that can use context from up to 10 open tabs. The assistant also integrates with Google apps like Calendar, Docs, Drive, Maps, Search, and YouTube, including the ability to compose and send messages with #Gmail from the side panel, while #Nano Banana 2 adds image editing without uploading files or opening a new tab. For safety, it asks for confirmation before sensitive actions like sending email or adding calendar events. Alongside Chromebook Plus, Mac, and Windows, the expansion also brings #Gemini in Chrome to the iOS browser.
23. Photographers Can Now Tell Photoshop How to Edit Their Images
Adobe has introduced an agentic #AIAssistant for #Photoshop that lets users describe edits in natural language and either apply them automatically or receive a step-by-step guide, aiming to make tasks like removing distractions, changing backgrounds, and refining lighting or color easier to learn and execute. The first release is limited and is available as a public beta on Photoshop for web and mobile, not desktop, with text chat and voice input supported in the mobile app. In Photoshop web, a new #AIMarkup tool powered by the assistant allows users to draw on an image and attach prompts specifying what to change and where, such as adding flowers or mountains in a marked area. Adobe also highlighted broader #Firefly updates: Firefly-powered features like #GenerativeFill, #GenerativeRemove, #GenerativeExpand, #GenerativeUpscale, and Remove Background are now available in a dedicated Firefly Image Editor, and Firefly supports over 25 non-Adobe models including Google’s Nano Banana 2, OpenAI Image Generation, and Black Forest Labs’ Flux.2 [pro]. Together, these additions signal Adobe’s wider plan to bring agentic AI workflows across its Creative Cloud apps while expanding Firefly’s model ecosystem for image editing.
24. Gemini burrows deeper into Google Workspace with revamped document creation and editing
#Gemini is getting a broader overhaul across #Google Workspace to help users create, edit, and search documents by pulling context from their #Google account. In #Docs, a new chatbot-like prompt box will generate first drafts and support follow-up edits such as reformatting, targeted section changes, and AI-assisted style matching, with suggestions staying private until approved. In #Sheets, Gemini can generate spreadsheets from prompts and specified data sources, and Google claims it can even fill missing data via web searches while handling tasks from basics to complex analysis. In #Slides, Gemini can create a single slide from a prompt and rework it, with multi-slide presentation generation planned for later, and #Drive search will add more AI-driven results plus an AI Overview with citations and “Ask Gemini” queries over folders or result sets that can be exported into new documents. These features are slated to arrive first for Google AI Ultra and Pro subscribers, with Docs, Slides, and Sheets rolling out gradually in spring in English, and Drive search changes coming in the following months.
25. Submerged bumblebee queens breathe underwater
Lab experiments show that hibernating bumblebee queens can survive for days underwater by continuing to breathe and by shifting to #anaerobic metabolism. After an earlier lab accident revealed eastern bumblebee queens (Bombus impatiens) could avoid drowning for up to a week, @Charles Darveau and colleagues submerged hibernating queens in cold water for eight days and tracked oxygen intake, carbon dioxide output, and lactic acid buildup. Oxygen in the water declined while queens continuously released carbon dioxide, and lactic acid levels spiked, indicating both underwater respiration and oxygen-independent energy production. How underwater breathing works is still uncertain, but the researchers suggest the queens might trap a thin air layer like some aquatic insects; after resurfacing, queens breathe faster for days to clear lactic acid. The findings imply that queens hibernating in flood-prone soils may endure flooding, though heavier rainfall with #climate change could mean repeated submersions that erode energy reserves and threaten survival.
That’s all for today’s digest for 2026/03/11! We picked, and processed 25 Articles. Stay tuned for tomorrow’s collection of insights and discoveries.
Thanks, Patricia Zougheib and Dr Badawi, for curating the links
See you in the next one! 🚀
