🤖 AI Startup Firmus to Build Indonesia Data Center With Nvidia

Firmus Technologies Pty Ltd. will build its first data center project in Indonesia as part of a partnership with US chip giant Nvidia Corp., which the Australian artificial intelligence infrastructure firm said is expected to win it as much as $30 billion in committed offtake agreements in its first six years.

Firmus and Singapore-based DayOne will develop a 360 megawatt Nvidia DSX AI Factory campus in Batam, Indonesia, an island just off the coast of Singapore, as part of an eight-year partnership with Nvidia, the company said.

#BRICS

@QSIMedia

🇮🇩🤑 Ripple's RLUSD Stablecoin Launches on FLOQ in Indonesia

“RLUSD is a meaningful addition to the digital asset ecosystem in Indonesia. We’re proud to bring RLUSD to a licensed platform and to work alongside Ripple to support responsible adoption across the market,” Yudhono Rawis, Founder and CEO of FLOQ.

Ripple's RLUSD stablecoin has been listed on Indonesia's FLOQ digital asset platform, providing over 1.8 million registered users access to this digital dollar-pegged asset, significantly enhancing its market presence in Southeast Asia.

The introduction of XRP/RLUSD trading pairs on the Gate platform broadens liquidity channels, promotes global stablecoin adoption, and drives innovation and efficiency in cross-border transactions. 🚀

Indonesia is positioning early. RLUSD goes global 🌏

#Stellar #BRICS

@QSIMedia

📱🇻🇪Venezuela is restoring access to X, allowing millions to use the app without a VPN.

X had been blocked since 2024 under Maduro’s regime. Venezuela’s main ISPs, including the state-owned CANTV, have begun progressively lifting the block following the earthquakes.

Millions of people can once again share and access real-time updates without relying on a VPN.

@QSIMedia

⚔️ China’s Z.ai Has Matched Anthropic in Cybersecurity, Resetting AI Race

China’s Zhipu AI (Z.ai) released its open-weight GLM-5.2, and security researchers have claimed that it matches Mythos in certain bug-finding and cybersecurity scenarios. While GLM lags behind models from Anthropic and OpenAI in other, more general tasks, it seems that China has dramatically reduced the gap in the capabilities between its models and those of the US.

Because GLM is an open-weight model, it can be downloaded and run by anyone on readily available hardware. That gives it great flexibility and allows power users deep access, but it also makes it ripe for abuse by bad actors who can run it with little oversight.

#BRICS

@QSIMedia

🤖 IBM, Red Hat, and Deloitte Announce Lightwell Collaboration to Help Strengthen Open Source Software Supply Chain Trust

Deloitte, IBM, and Red Hat announced a collaboration to help protect the software supply chain against increasingly automated cyber threats. Deloitte joins the initiative as an integration collaborator for Lightwell, bringing its broader secured software supply chain architecture and cyber risk services to the large-scale enterprise open source security model deployed by IBM and Red Hat.

❗️Through this collaboration, the three organizations will coordinate across the software lifecycle to help clients manage security threats:

🟠Continuous Visibility & Discovery: Continuously mapping and scanning first-party, open source, and third-party software to identify exactly what code exists, where it runs, and which critical business functions it supports.

🟠Contextual Prioritization: Separating active threats from noise by analyzing severity, exposure, threat-chaining, and exploitability to inform operational decisions.

🟠Machine-Speed Remediation: Combining Red Hat and IBM’s automated patch validation with Deloitte’s orchestration services to rapidly coordinate, test, and deploy validated fixes into production repositories, limiting disruption. To achieve this, Deloitte will maintain a bench of Forward Deployed Engineers (FDEs) to support ongoing remediation and maintenance of client applications.

🟠Ecosystem Trust & Compliance: Through the collaboration, the organizations will help enterprises manage upstream open source and vendor relationships, including pre-disclosure vulnerability handovers, while delivering continuous, evidence-based reporting for boards, auditors, and regulators.

“Exploits don't wait for manual patching processes, and neither can enterprise response,” said Adnan Amjad, Deloitte’s US Cyber leader.

“Together, we're enabling clients to operate at machine speed to identify, validate, and remediate vulnerabilities. This collaboration is about building the operational resilience needed to maintain trust across increasingly complex software ecosystems — creating systems that can withstand and neutralize risk without disrupting the business,” added Adnan Amjad.

@QSIMedia

🧠 Scientists Discover the Brain Can Rewire Itself To Truly Multitask

Why does driving eventually feel effortless, while learning to drive demands total concentration? A new study from Georgetown University suggests the answer lies in the brain’s ability to rewire itself, shifting well-practiced skills into different neural circuits so they can be performed with less conscious effort.

⬇️ Under the right conditions, people may be capable of genuine multitasking rather than simply switching attention rapidly between tasks.

⚪️For the new study, participants learned to sort morphed images of cars into two groups by noticing small visual differences. Over 5 to 10 weeks, they completed more than 30,000 trials through a phone app that turned the sorting task into a game. Before and after training, the researchers scanned participants’ brains using fMRI and EEG.

⚪️At first, once participants had learned the sorting task, it activated the prefrontal cortex. That brain region supports executive function and deliberate thinking, but it is generally limited in how many tasks it can manage at once.

❗️After weeks of practice, brain scans showed a shift. The sorting process had moved into the temporal cortex, a region involved in memory encoding and the recognition of complex objects.

“We measure before and after training, so we can see that extensive training essentially put a category-selective area in the temporal lobe that was not there before. Experience remodels the brain to bypass that frontal bottleneck. The prefrontal cortex then stays free for whatever else you want to do, increasing your capacity,” Riesenhuber explained.

Moving a learned skill into the temporal cortex, while freeing the prefrontal cortex, may allow the brain to use established knowledge as a base for learning something new. 📚

@QSIMedia

⚡️🇪🇸Spain to Phase Out Electricity Generation Tax, Government Says

Scrapping the 7% tax on power generation will help lower electricity bills for households and businesses while encouraging the shift from fossil fuels to clean-energy, the government said. The tax will be phased out by 2028.

@QSIMedia

🤑 Ripple wants institutions to borrow against tokenized assets on XRPL

Ripple has proposed an XRPL Lending Protocol that would let institutions borrow against on-chain assets while keeping credit decisions off the blockchain.

⬇️ The system would pool single assets in “Single Asset Vaults” and automate loan mechanics at the base layer of the XRP Ledger.

The protocol has two parts. A Single Asset Vault pools a single asset, and the lending layer turns that pooled money into loans with set terms. Both are still proposals, defined in technical drafts known as XLS-65 and XLS-66, and remain subject to approval by the validators who run the network.

❗️The features are available to test on a development network but are not live.

Aimed at institutional use and competing with protocols like Aave and Compound, Ripple argues its fixed, network-level rules offer more predictable risk than crypto-native governance models.

The XRPL Lending Protocol (XLS-65, XLS-66) proposals are subject to validator approval in the coming weeks. Infrastructure providers and developers can begin integrating and testing on a testnet from Monday. 🚀

#Stellar

@QSIMedia

🇨🇳 China Pushes All School Levels to Teach AI in Xi’s Tech Drive

China has announced a formal plan to integrate AI across all levels of schooling, aligning its education system with President Xi Jinping’s campaign to dominate advanced technologies. The State Council released a five-year blueprint on Monday including a plan to make AI a core capability for every student, echoing Xi’s earlier call to cultivate high-tech talent.

The plan urges efforts to “promote AI education throughout all educational stages, improve students’ AI literacy and enhance their ability to identify and solve problems”.

#BRICS

@QSIMedia

🇺🇸 Governor Newsom announces a first-of-its-kind partnership to provide Anthropic's tools to California state agencies.

Governor Gavin Newsom today announced that the state has entered into a partnership with leading California-based AI company Anthropic, a first-of-its-kind collaboration that will help support state workers and responsibly adopt AI to make state services more effective, user-friendly, and efficient.

"This partnership is about using technology the California way: responsibly, transparently, and in service of people. AI should not replace the human work of government; it should help our workers move faster, solve problems more effectively, and deliver better results for Californians," Governor Gavin Newsom

Through the agreement, state agencies may access Anthropic’s AI productivity assistant, Claude, at a 50% discounted price, coupled with free workforce training as well as expert GenAI technical assistance and workflow input from Anthropic developers. The agreement also provides the same discounted offer for California’s local governments, including cities and counties.

@QSIMedia

🏢 Google Cloud to Offer Specialist AI Models for Science Research

Google will start offering specialist AI models from software company SandboxAQ through its cloud service, expanding enterprise and research access to technology designed to accelerate drug discovery, materials science and semiconductor manufacturing.

Google will add SandboxAQ’s so-called large quantitative models to its cloud marketplace, allowing researchers to combine Google’s Gemini language models with AI systems trained on scientific equations and laboratory data. Unlike large language models, which excel at generating text, large quantitative models are trained on numerical data to help solve problems in chemistry, biology and physics.

@QSIMedia