Singapore AI News & Daily Briefing

Meta is developing an AI pendant and expanding into business-focused wearables, aiming to sell 10 million devices in the second half of 2026. The push comes as its Reality Labs hardware division lost $4 billion in the first quarter on just $402 million revenue. The company also acquired AI-wearables startup Limitless and plans to launch the new products alongside existing Ray-Ban and Oakley smart glasses partnerships.

China's MiniMax, an AI company that generates text, images, audio, and video using large language models, is preparing for a domestic stock listing to fund its fight against rivals like DeepSeek. The Shanghai-based firm, which listed in Hong Kong in January, now claims over $300 million in annual recurring revenue and over one million business users—but remains unprofitable due to heavy spending on AI development.

Singtel has secured S$28 million in funding from Digital Industry Singapore (DISG) to accelerate AI transformation across its operations. The multi-year programme will cover AI-enabled operations, digital infrastructure, customer platforms, and workforce development, starting with foundational AI capabilities for Singtel Singapore.

Singapore's Ministry of Education is rolling out AI literacy across all universities and polytechnics, meaning every student will graduate with practical AI skills. The government sees this as essential for Singapore's future workforce as AI reshapes industries. Students can expect new AI-focused modules and hands-on training built into their degrees.

Fortinet has launched a new AI-powered Network Detection and Response cloud hub in Singapore—the first of its kind in the ASEAN region. The point-of-presence helps security teams across Asia-Pacific detect sophisticated cyber threats designed to bypass traditional security tools. For everyday users, this means better protection against increasingly smart cyberattacks targeting businesses in Singapore and the region.

OpenAI has signed a deal with Singapore's Ministry of Digital Development and Information (MDDI) worth over S$300 million (US$234 million). The partnership includes an OpenAI Singapore Applied AI Lab, programs to build local AI talent, and support for startups, small businesses, and government agencies. This is OpenAI's first such memorandum with any country and could serve as a template for future national partnerships as the company pivots from selling AI models to providing AI services and consulting.

The TechInAsia article headline mentions Meta stock dropping 5% due to AI funding needs, but the actual page content is inaccessible (shows an error/login page). The URL suggests Singapore staff layoffs may be part of the company's cost-cutting measures alongside AI investments.

An AI industry leader argues that AI agents are the next major technology shift—potentially as transformative as the browser or smartphone. The piece explores how agents could take over tasks like trip planning and cross-app coordination, why this demands new hardware designs, and how companies like Qualcomm are positioning for this shift. It's a forward-looking industry perspective on where AI is heading.

One NZ, one of Australia and New Zealand’s leading telecommunications providers, has cut enterprise mobile provisioning times from ten days to under ten minutes using UiPath Maestro, a cloud-native orchestration platform that unifies automation, AI agents, and human interaction into end-to-end business processes. The deployment was implemented in just five weeks and without a costly […] The post One NZ Cuts Mobile Provisioning from Ten Days to Ten Minutes with UiPath Maestro appeared first on techcoffeehouse.com .

AI tools are compressing hours of work into minutes for tech workers—drafting documents, summarizing meetings, reviewing code. But the time saved isn't translating to shorter workdays. Workers say any gains get immediately redirected to the next project, and some are actually working longer hours while building the automation systems. It's a candid look at the reality behind the productivity hype.

Google is partnering with investment firm Blackstone to launch a new AI cloud company, with Blackstone pouring in $5 billion in equity. The venture will use Google's custom Tensor Processing Units (TPUs) to focus on AI inference—the ongoing process of running AI models for users—rather than just training. Blackstone is also tackling a key bottleneck in AI expansion: power supply. They've partnered with US utility PPL Corporation to build gas-fired power plants specifically for data centers, since new grid connections can take 7-10 years in some markets. Tech giants are collectively spending over $700 billion on AI this year.

OpenAI CEO Sam Altman met with Sen. Bernie Sanders to discuss giving the public equity stakes in AI companies. Sanders proposed a 50% public ownership model to create a public wealth fund. Altman expressed support for public ownership in principle but couldn't endorse the 50% threshold. President Trump also voiced interest in allowing Americans to benefit from AI's success. This represents growing bipartisan support for public ownership models in AI development.

Investment giants Apollo and Blackstone have finalized a massive $35 billion financing deal for AI company Anthropic, pushing its valuation to $96.5 billion. Semiconductor company Broadcom is guaranteeing $30 billion in senior debt to reduce borrowing costs. This represents one of the largest AI company funding rounds in history.

Marvell Technology, a US chipmaker, just got promoted to the S&P 500 after forecasting massive revenue growth driven by AI data-center demand. The company expects nearly $11 billion in fiscal 2027 revenue (up 30% year-on-year) and $15 billion by fiscal 2028. Its custom AI chips for tech giants like Amazon and Google now make up 10-15% of sales, with optical interconnect company Celestial AI acquisition set to boost high-speed chip connections.

Yahoo Singapore profiles five people who switched careers into AI roles, exploring how they made the transition and what they learned. The article offers practical takeaways for anyone considering a similar career move.

Huawei is committing $2.6 billion to smart driving technology research, including $1.46 billion specifically for computing power to train AI systems. The company plans to spend up to $11.7 billion on computing power over five years, even as US export controls limit its access to Nvidia chips. Huawei's Ascend AI chips are being used by Chinese AI firm DeepSeek, and its Qiankun ADS driver assistance system now powers over 15 vehicle models, with Audi signed on for China.

Hundreds of residents in Columbia, Pennsylvania packed a municipal meeting to oppose a proposed data center project, with commenters estimating at least 500 attendees still speaking past 10pm. The backlash reflects growing tension between communities asked to absorb the costs of Big Tech infrastructure versus the promised benefits. Data centers consume enormous amounts of electricity and water, potentially straining local utilities and raising costs for residents. As AI-driven demand pushes companies to build more facilities, local meetings like this may become more common.

Tech in Asia's weekly funding roundup highlights infrastructure startups attracting big checks. DriveNets (Israel) raised $410 million for AI networking software backed by AMD, while South Korean chip startup Xcena grabbed $135 million for AI memory chips. UAE logistics firm CargoX secured $250 million for autonomous delivery networks. The theme: investors are funding the digital 'backbone'—the unglamorous infrastructure that keeps AI and tech running.

Chinese researchers unveiled AI-powered drone swarms that can autonomously find and eliminate targets without human approval. The system raises alarms about weapons making life-or-death decisions without people in the loop.

OpenAI and the Trump administration are negotiating a direct government stake in the AI startup. The idea is a "Public Wealth Fund" that would pay out directly to American citizens. Senator Bernie Sanders wants to push through a 50 percent tax on AI shares by law. Critics fear the arrangement could create a "too big to fail" dynamic similar to the 2008 financial crisis. The article OpenAI and the Trump administration are negotiating a government stake in the AI startup appeared first on The Decoder .

Google DeepMind CEO Demis Hassabis says artificial general intelligence — AI that matches human-level thinking — could arrive by 2030. He calls it the start of a "new human era" that could bring medical breakthroughs and even a post-scarcity world. But he warns society has a narrow window to prepare for massive changes to jobs and daily life. Unlike some tech leaders who dialed back their more alarming predictions, Hassabis is still urging people to take the timeline seriously.

The US President is meeting with AI companies next week to discuss sharing AI profits with the public. The proposal could reshape how AI wealth is taxed and distributed. While details are still emerging, this signals governments worldwide are eyeing AI companies' massive profits for public benefit.

Nvidia just crashed the consumer laptop market with its new RTX Spark processor, capable of 1 petaflop of AI performance. Every major laptop brand — Dell, HP, Microsoft, Asus, Lenovo — announced new models powered by this chip. Microsoft unveiled the Surface Laptop Ultra as its flagship RTX Spark device, while Dell launched a $599 XPS 13 as a direct competitor to Apple's budget MacBook Neo. Most devices arrive this fall.

This week's security roundup covers serious privacy concerns: Meta quietly installed dormant face recognition code on over 50 million phones through its smart glasses app, capable of identifying people without their knowledge. Also, hackers exploited Meta's AI support chatbot to hijack Instagram accounts. The tech giant settled billions in biometric lawsuits just years ago over similar technology.

AI has become the top reason companies cite when laying off workers—nearly 40% of job cuts announced in May 2026 were attributed to AI, up from just 7% in January. The first five months of 2026 saw 87,714 AI-linked job cuts, already surpassing all of 2025. While overall employment numbers remain positive, experts warn that workers in tech and other sectors face concentrated disruption, and many open positions don't match the skills of those being displaced.

Senator Bernie Sanders has floated a dramatic proposal: a 50% tax on wealth created by AI and automation, with the money going directly to Americans. The idea is to make sure AI profits don't only benefit tech companies and billionaires. Political odds are long, but the debate is heating up.

Anthropic publicly warns about the dangers of authoritarian governments using AI while accepting billions in investment from Abu Dhabi's sovereign wealth fund MGX, which is controlled by an authoritarian monarchy. The company argues democracies must lead AI development, but its own investors include a regime known for surveillance, oppression, and hacking dissidents. Critics call this hypocrisy; Anthropic calls it a "communications headache."

AI data centers are consuming so much memory that other industries face shortages. Cars, medical devices, and phones could become more expensive as memory chips get snapped up by AI companies. An industry group is urging the Trump administration to step in.

Meta is launching "Hatch," its first paid AI product — a monthly subscription potentially costing up to $200. Hatch lets users describe what they need in plain language, and the AI builds tools, schedules appointments, or sends emails on their behalf. CEO Mark Zuckerberg is positioning this as a major new revenue stream beyond advertising.

US Senator Elizabeth Warren is pushing for Nvidia CEO Jensen Huang to appear before the Senate and explain how Nvidia's AI chips—particularly the H20 model—are being sold to China. Warren and other lawmakers argue these sales may undermine US efforts to restrict China's access to advanced AI technology that could be used militarily. Nvidia is already under existing export restrictions, but the H20 chip remains legal to sell after additional compliance requirements. The Senate Banking Committee, which oversees export controls, would be the forum for such testimony.

Elon Musk's xAI reportedly used Anthropic's Claude AI to train its own coding models for months, even after Anthropic cut off access. xAI continued using private accounts and a third-party service called Blackbox AI. Meanwhile, xAI's own pretraining team reportedly shrank to fewer than five people, with several leads leaving. The AI compute infrastructure Musk acquired is now being rented out to rivals Anthropic and Google instead.

Japanese AI startup Sakana AI has launched a dedicated lab to develop AI that recursively improves itself, potentially sidestepping the massive compute costs driving rivals like OpenAI and Anthropic to spend billions. The idea: smarter algorithms rather than bigger data centers. Co-founded by Transformer co-author Llion Jones, the firm bets efficiency beats brute force. Meanwhile, Anthropic warns similar self-improving AI carries serious control risks.

Companies that rushed to replace workers with AI are having second thoughts. A new trend called the "AI boomerang" is emerging: firms that laid off employees due to AI productivity gains are now rehiring for those exact same roles. Research shows 55% of employers regretted AI-driven layoffs, and half of companies that replaced customer service staff with AI will rehire by 2027. The lesson? AI works best alongside humans, not instead of them.

One NZ, a major telecom in Australia and New Zealand, cut its mobile provisioning time from 10 days to under 10 minutes using UiPath Maestro, an AI-powered automation platform. The rollout took just 5 weeks. This shows how enterprises are using AI and automation to slash operational costs and deliver services faster—potentially changing what customer service looks like across industries.

Researchers have released an open-source voice AI model called 'Audio Interaction' that listens continuously and decides every 0.4 seconds whether to respond or stay silent. Unlike typical voice assistants that wait for you to finish speaking, it processes audio in real-time — handling translation, transcription, chat, and ambient sounds like coughing in one continuous stream. Everything is free under Apache 2.0 license.

A new report confirms that AI-powered agents — bots that autonomously browse and interact with websites — now account for more internet traffic than actual human users. This milestone marks a shift in how the web is consumed, with AI systems scraping data, training on content, and performing tasks at scale. For everyday users, this means websites may prioritize serving AI over humans, and content creators could see their work fed to bots instead of readers.

Update on Anthropic calls for pause of global AI development. Coverage via Reddit/r/artificial.

Apple is giving Siri another major overhaul at WWDC, aiming to catch up after years of AI lag behind Google and OpenAI. The company appears ready to showcase deeper AI integration across iPhones and Macs. For everyday iPhone users, this matters because Siri is how millions interact with their devices daily — better AI could mean faster, smarter responses and fewer frustrating 'I don't understand that' moments.

A helium shortage is threatening to slow down global chip manufacturing. Helium is essential for cooling the lasers used in semiconductor production, and supplies are tightening due to geopolitical disruptions and increased demand from tech and healthcare sectors. If the crisis worsens, it could constrain the supply of AI chips and drive up prices for everything from smartphones to data center servers.

Meta is now filling its AI app with AI-generated clickbait articles instead of linking to real news. The standalone Meta AI app has a "For You" feed that populates stories entirely created by AI—including the topics, images, and text. This mirrors the low-quality content that has long plagued Facebook, but now the platform is manufacturing it itself. Users may struggle to distinguish between real news and AI-manufactured content.

McDonald's is rolling out Archy IQ, a new AI drive-thru system built with Google, after discontinuing a previous AI attempt two years ago. Currently testing in five stores, the system handles orders in English and Spanish and has processed over one million transactions with 90% completed without human help. All US locations are getting Google Edge Cloud hardware installed in preparation. Customers are pushing back hard—complaining about job losses, poor customer service experiences, and fewer human interactions. The system also acts as a restaurant management assistant, flagging operational bottlenecks.

The Acer Swift Air 14 ($699) is pitted against Apple's MacBook Neo ($599 for students) as a direct Windows response to Apple's budget laptop disruption. The Swift Air wins on battery (19 hours vs 16), ports (dual USB-C Thunderbolt 4), and storage expandability (up to 1TB). The MacBook Neo wins on display quality, camera, and overall value at its price point. For students on a budget, the MacBook Neo remains the best bang-for-buck, though Windows rivals are catching up.

Apple Watch users are stuck with a Siri that feels outdated compared to what's available on iPhone. The argument: Apple Watch is where AI assistance makes most sense—you need quick answers while your hands are busy—so fixing Siri there matters more than on the phone. Currently, Apple Watch Siri lacks features available on iPhone, creating a frustrating experience for the device most suited to voice commands.

Want better AI images from ChatGPT or Gemini? A simple prompt tweak — adding specific style descriptors or lighting details — can dramatically improve results. The comparison shows both tools have strengths, but small language tweaks unlock significantly better outputs.

A free, open-source macOS app called Council lets you ask the same question to multiple AI models like Claude, GPT, and Gemini at once. The models then critique each other's answers without knowing who said what, showing you exactly where they disagree. It's a practical tool for anyone who wants deeper analysis rather than trusting a single AI's opinion.

President Trump signed a directive pushing US military and intelligence agencies to rapidly adopt AI systems from multiple vendors while building high-security computing facilities. The policy creates an AI National Security Strategic Reserve to bring civilian AI experts into government planning and replaces Biden-era AI policy described as restrictive. Agencies must ensure no company can disable or degrade AI systems used by warfighters without government approval. It also directs the Secretary of War to issue new guidance on autonomous weapon systems.

A $2 billion data center project in Shelbyville, Indiana has sparked community backlash, with residents posting 'No Data Center' signs. The mayor was caught on camera dismissing opposition by saying only people in 'shitty houses' would object. The controversy centers on local concerns about noise, property values, and environmental impact from the large facility needed to power AI systems.

Alibaba's Qwen team has released Qwen3.7-Plus, a multimodal AI agent model that combines visual understanding, app operation, and coding in one system. In a demo, an agent built on this model autonomously created a vocabulary learning app, generating over 10,000 lines of code across 1,000 agent calls over eleven hours—all without human help. The model leads Qwen's own benchmarks for screen understanding but overall performance is mixed compared to Western competitors. It's a proprietary model priced well below OpenAI or Anthropic offerings.

# The quickest way to find blocked numbers on your iPhone - it's easy | ZDNET X Trending * [I wore the world's first HDR10 smart glasses](https://www.zdnet.com/article/rayneo-air-4-pro-hdr10-smart-glasses-ces/) * [TCL's new E Ink tablet beats the Remarkable and Kindle](https://www.zdnet.com/article/tcl-note-a1-nxtpaper-ces/) * [Anker's new charger is one of the most unique I've ever seen](https://www.zdnet.com/article/anker-wireless-chargers-ces-2026/) * [Best laptop cooling pads](https://www.zdnet.com/home-and-office/best-laptop-cooling-pad/) * [Best flip phones](https://www.zdnet.com/article/best-flip-phone/) * [Roku TV vs Fire Stick](https://www.zdnet.com/home-and-office/home-entertainment/roku-tv-vs-fire-stick/) * [Galaxy Buds 3 Pro vs Apple AirPods Pro 3](https://www.zdnet.com/article

Meta's stock dropped over 5% after a report surfaced that the company might raise tens of billions of dollars through new stock sales to fund its AI ambitions. The move mirrors Alphabet's plan to raise $85 billion for AI infrastructure. Investors are nervous because Meta is already spending up to $145 billion on capital expenses this year, and its stock has underperformed compared to Google, partly because Meta lacks a profitable cloud business to justify the spending.

AMD now controls a third of the server CPU market (33.2%), up six percentage points from last year, as AI data center demand continues to surge. Intel still dominates with roughly two-thirds share, but AMD's Epyc processors are gaining ground fast. Server CPU shipments grew over 10% year-on-year, driven by the AI boom, while general PC chip shipments declined. The shift matters because AI companies buying servers are increasingly choosing AMD over Intel.

Utah residents are suing to stop Kevin O'Leary's massive AI data center, claiming an unelected authority called MIDA bypassed constitutional rules and blocked public input. After political pressure, O'Leary agreed to shrink the project by 75%. This is the first major legal challenge to how AI data centers get approved in the US.

Update on Google To Pay SpaceX $920 Million Per Month For Massive AI Compute Power. Coverage via Reddit/r/technology.

Researchers from a Korean institute, NYU, and Université de Montréal are raising uncomfortable questions about how we measure consciousness. Their paper in Neuron argues that common experiments like showing people flashing images and asking what they saw may not actually measure subjective experience at all — they may just measure how well the brain processes information. The team points to cases like 'blindsight' patients, who report seeing nothing yet can correctly guess what's in front of them, as evidence that awareness and information processing can come apart. They're not saying AI or animals definitely aren't conscious — just that the science isn't solid enough to make confident claims. The researchers warn that if policy debates about animal welfare or AI ethics are going to lean on neuroscience, that neuroscience first needs better tools.

Taiwan's Gigabyte showcased a bizarre server at Computex that packs 40 Intel laptop chips (Lunar Lake) into a single chassis the size of a pizza box. Each of the 40 nodes runs a Core Ultra 7 processor with its own memory, SSD, and built-in graphics—totaling 320 cores and 1.28 TB of RAM. The design targets light workloads like cloud PCs and casual game streaming, letting companies skip expensive vGPU licensing fees since each node has dedicated graphics.

A CEO has publicly told employees there will be no raises this year, blaming the expenditure on AI investments instead. The story highlights growing concerns about companies prioritizing AI spending over worker compensation.

Sometimes, you have to dig beyond the headlines to get a better idea of how a company is performing. Endowus’ latest audited financial results are a good example of this. [Gregory Van](https://www.linkedin.com/in/gregory-van/?originalSubdomain=hk), the co-founder and CEO of the Singapore-based wealthtech firm, told _Tech in Asia_ last year that it had a [“clear path to profitability”](https://www.techinasia.com/endowus-doubles-revenue-2022-losses-rose-26m) by 2025. On its face, the results suggest that the firm has its work cut out for it to hit this milestone. Endowus co-founders Samuel Rhee (left) and Gregory Van / Photo credit: Endowus In 2023, Endowus’ total losses before tax widened to S$30.6 million (US$23.7 million) – a 14% jump from last year. This came even as its revenue grew by

# Family offices invest in sports, from pickleball to smart soccer balls [Skip Navigation](https://www.cnbc.com/2026/06/05/family-office-sport-investments.html#MainContent) [](https://www.cnbc.com/) Family offices bet on sports, from pickleball leagues to smart soccer balls Share Share Article via Facebook Share Article via Twitter Share Article via LinkedIn Share Article via Email [Markets](https://www.cnbc.com/markets/) * [Pre-Markets](https://www.cnbc.com/pre-markets/) * [U.S. Markets](https://www.cnbc.com/us-markets/) * [Currencies](https://www.cnbc.com/currencies/) * [Prediction Markets](https://www.cnbc.com/prediction-markets/) * [Cryptocurrency](https://www.cnbc.com/cryptocurrency/) * [Futures & C

Communities across America are mounting serious legal and political fights against AI data centers — and they're winning. A Virginia court just derailed what would have been the world's largest data center on a technicality: the county failed to properly publish public notices. Nationally, at least 25 data center projects were canceled last year, and half of those planned for 2026 are being delayed or scrapped. Locals are angry about higher electricity bills, water shortages, and paltry job gains — and they're voting out officials who push these projects through.

A security researcher found that Creative Technologies' Sound Blaster Katana V2X speaker—a $283 device popular with gamers—can be hacked over Bluetooth to infect connected PCs with malware, all without the owner ever touching the speaker. The flaw lets anyone within Bluetooth range replace the speaker's firmware with malicious code that can type commands and execute programs on your computer. The researcher, who involved Singapore's cybersecurity agency CERT, says the company hasn't fixed the issue.

A new academic study tracked what happens when you ask AI chatbots like ChatGPT and Gemini to build stock portfolios for eight months. The result: AI picks heavily concentrated portfolios (up to 41% semiconductors), tends to follow recent news and stock hype, and after adjusting for risk factors, doesn't actually beat the market. The takeaway for regular investors: asking chatbots for investment advice gives you a risky, lopsided portfolio that performs no better than simply buying similar stocks on your own.

Microsoft has unveiled its own AI models after investing heavily in OpenAI ($13B) and Anthropic ($5B). The new tools include MAI-Code-1-Flash for coding and MAI-Thinking-1 for complex reasoning, both marketed as cheaper alternatives to third-party options. Microsoft claims 10x better cost efficiency than OpenAI's GPT 5-5 for certain tasks. For everyday users, this could mean lower subscription costs for AI coding tools as competition heats up.

Goldman Sachs' top AI skeptic is warning that major AI companies like OpenAI and Anthropic face mounting pressure to prove they can make money as they prepare for IPOs. The concern: billions in AI investments haven't yet delivered clear profits, and investors are getting nervous. For everyday users, this matters because if these AI giants fail to monetize their products, subscription prices could rise, features could disappear, or the entire AI services market could shift dramatically.

A tech worker won a religious exemption from using AI at her coding job, after arguing the technology conflicts with her Unitarian Universalist beliefs. The Pope's recent warning that AI could undermine human dignity and displace workers has given more workers ammunition to request similar accommodations. Employers could face legal pressure under US federal law to grant these requests — and workers who refuse may face career consequences as AI becomes ubiquitous in the workplace.

Google has rolled out its AI-powered virtual try-on feature in Singapore, letting shoppers upload a photo to see how clothes and shoes would look on their body before buying. Previously only available in the US, the tool works across Google Search and Shopping. This changes online apparel shopping by reducing guesswork and potential returns.

Google is partnering with investment firm Blackstone to launch a new AI cloud company, with Blackstone pouring in $5 billion in equity. The venture will use Google's custom Tensor Processing Units (TPUs) to focus on AI inference—the ongoing process of running AI models for users—rather than just training. Blackstone is also tackling a key bottleneck in AI expansion: power supply. They've partnered with US utility PPL Corporation to build gas-fired power plants specifically for data centers, since new grid connections can take 7-10 years in some markets. Tech giants are collectively spending over $700 billion on AI this year.

Illustration by Tag Hartman-Simkins / Futurism. Source: Karl Mondon / AFP via Getty Images; Shutterstock Not even Google's own workforce seems all that enthusiastic about AI. Internal messages [obtained by](https://www.404media.co/google-employees-internally-share-memes-about-how-its-ai-sucks/)[_404 Media_](https://www.404media.co/google-employees-internally-share-memes-about-how-its-ai-sucks/) show employees ruthlessly mocking AI tools — including the company's own internal AI coding tool, Jetski — complaining that they're unreliable and [make their jobs harder](https://tech.yahoo.com/ai/articles/ai-forcing-employees-harder-ever-125732788.html). It's clear who these jokes are being aimed at. CEO [Sundar Pichai](https://tech.yahoo.com/ai/gemini/articles/college-grads-boo-mention-ai-1416584

Google's own employees are loudly mocking the company's AI push from the inside. Internal memes shared on a company message board called "Memegen" show workers ridiculing AI tools — including Google's own internal coding tool — calling them "slop" and saying they make jobs harder. While CEO Sundar Pichai boasts that 75% of Google's code is now AI-generated, employees say the real bottleneck has just shifted: code comes out fast, but human review and testing now creates new delays company-wide.

Singtel has secured S$28 million in funding from Digital Industry Singapore (DISG) to accelerate AI transformation across its operations. The multi-year programme will cover AI-enabled operations, digital infrastructure, customer platforms, and workforce development, starting with foundational AI capabilities for Singtel Singapore.

The TechInAsia article headline mentions Meta stock dropping 5% due to AI funding needs, but the actual page content is inaccessible (shows an error/login page). The URL suggests Singapore staff layoffs may be part of the company's cost-cutting measures alongside AI investments.

Marvell Technology, a US chipmaker, just got promoted to the S&P 500 after forecasting massive revenue growth driven by AI data-center demand. The company expects nearly $11 billion in fiscal 2027 revenue (up 30% year-on-year) and $15 billion by fiscal 2028. Its custom AI chips for tech giants like Amazon and Google now make up 10-15% of sales, with optical interconnect company Celestial AI acquisition set to boost high-speed chip connections.

Huawei is committing $2.6 billion to smart driving technology research, including $1.46 billion specifically for computing power to train AI systems. The company plans to spend up to $11.7 billion on computing power over five years, even as US export controls limit its access to Nvidia chips. Huawei's Ascend AI chips are being used by Chinese AI firm DeepSeek, and its Qiankun ADS driver assistance system now powers over 15 vehicle models, with Audi signed on for China.

Tech in Asia's weekly funding roundup highlights infrastructure startups attracting big checks. DriveNets (Israel) raised $410 million for AI networking software backed by AMD, while South Korean chip startup Xcena grabbed $135 million for AI memory chips. UAE logistics firm CargoX secured $250 million for autonomous delivery networks. The theme: investors are funding the digital 'backbone'—the unglamorous infrastructure that keeps AI and tech running.

Meta's stock dropped over 5% after a report surfaced that the company might raise tens of billions of dollars through new stock sales to fund its AI ambitions. The move mirrors Alphabet's plan to raise $85 billion for AI infrastructure. Investors are nervous because Meta is already spending up to $145 billion on capital expenses this year, and its stock has underperformed compared to Google, partly because Meta lacks a profitable cloud business to justify the spending.

A CEO has publicly told employees there will be no raises this year, blaming the expenditure on AI investments instead. The story highlights growing concerns about companies prioritizing AI spending over worker compensation.

OpenAI has signed a deal with Singapore's Ministry of Digital Development and Information (MDDI) worth over S$300 million (US$234 million). The partnership includes an OpenAI Singapore Applied AI Lab, programs to build local AI talent, and support for startups, small businesses, and government agencies. This is OpenAI's first such memorandum with any country and could serve as a template for future national partnerships as the company pivots from selling AI models to providing AI services and consulting.

Researchers from a Korean institute, NYU, and Université de Montréal are raising uncomfortable questions about how we measure consciousness. Their paper in Neuron argues that common experiments like showing people flashing images and asking what they saw may not actually measure subjective experience at all — they may just measure how well the brain processes information. The team points to cases like 'blindsight' patients, who report seeing nothing yet can correctly guess what's in front of them, as evidence that awareness and information processing can come apart. They're not saying AI or animals definitely aren't conscious — just that the science isn't solid enough to make confident claims. The researchers warn that if policy debates about animal welfare or AI ethics are going to lean on neuroscience, that neuroscience first needs better tools.

Singapore's Ministry of Education is rolling out AI literacy across all universities and polytechnics, meaning every student will graduate with practical AI skills. The government sees this as essential for Singapore's future workforce as AI reshapes industries. Students can expect new AI-focused modules and hands-on training built into their degrees.

OpenAI CEO Sam Altman met with Sen. Bernie Sanders to discuss giving the public equity stakes in AI companies. Sanders proposed a 50% public ownership model to create a public wealth fund. Altman expressed support for public ownership in principle but couldn't endorse the 50% threshold. President Trump also voiced interest in allowing Americans to benefit from AI's success. This represents growing bipartisan support for public ownership models in AI development.

Investment giants Apollo and Blackstone have finalized a massive $35 billion financing deal for AI company Anthropic, pushing its valuation to $96.5 billion. Semiconductor company Broadcom is guaranteeing $30 billion in senior debt to reduce borrowing costs. This represents one of the largest AI company funding rounds in history.

Update on Google To Pay SpaceX $920 Million Per Month For Massive AI Compute Power. Coverage via Reddit/r/technology.

AMD now controls a third of the server CPU market (33.2%), up six percentage points from last year, as AI data center demand continues to surge. Intel still dominates with roughly two-thirds share, but AMD's Epyc processors are gaining ground fast. Server CPU shipments grew over 10% year-on-year, driven by the AI boom, while general PC chip shipments declined. The shift matters because AI companies buying servers are increasingly choosing AMD over Intel.

Taiwan's Gigabyte showcased a bizarre server at Computex that packs 40 Intel laptop chips (Lunar Lake) into a single chassis the size of a pizza box. Each of the 40 nodes runs a Core Ultra 7 processor with its own memory, SSD, and built-in graphics—totaling 320 cores and 1.28 TB of RAM. The design targets light workloads like cloud PCs and casual game streaming, letting companies skip expensive vGPU licensing fees since each node has dedicated graphics.

Chinese researchers unveiled AI-powered drone swarms that can autonomously find and eliminate targets without human approval. The system raises alarms about weapons making life-or-death decisions without people in the loop.

Utah residents are suing to stop Kevin O'Leary's massive AI data center, claiming an unelected authority called MIDA bypassed constitutional rules and blocked public input. After political pressure, O'Leary agreed to shrink the project by 75%. This is the first major legal challenge to how AI data centers get approved in the US.

Fortinet has launched a new AI-powered Network Detection and Response cloud hub in Singapore—the first of its kind in the ASEAN region. The point-of-presence helps security teams across Asia-Pacific detect sophisticated cyber threats designed to bypass traditional security tools. For everyday users, this means better protection against increasingly smart cyberattacks targeting businesses in Singapore and the region.

Hundreds of residents in Columbia, Pennsylvania packed a municipal meeting to oppose a proposed data center project, with commenters estimating at least 500 attendees still speaking past 10pm. The backlash reflects growing tension between communities asked to absorb the costs of Big Tech infrastructure versus the promised benefits. Data centers consume enormous amounts of electricity and water, potentially straining local utilities and raising costs for residents. As AI-driven demand pushes companies to build more facilities, local meetings like this may become more common.

McDonald's is rolling out Archy IQ, a new AI drive-thru system built with Google, after discontinuing a previous AI attempt two years ago. Currently testing in five stores, the system handles orders in English and Spanish and has processed over one million transactions with 90% completed without human help. All US locations are getting Google Edge Cloud hardware installed in preparation. Customers are pushing back hard—complaining about job losses, poor customer service experiences, and fewer human interactions. The system also acts as a restaurant management assistant, flagging operational bottlenecks.

President Trump signed a directive pushing US military and intelligence agencies to rapidly adopt AI systems from multiple vendors while building high-security computing facilities. The policy creates an AI National Security Strategic Reserve to bring civilian AI experts into government planning and replaces Biden-era AI policy described as restrictive. Agencies must ensure no company can disable or degrade AI systems used by warfighters without government approval. It also directs the Secretary of War to issue new guidance on autonomous weapon systems.

Alibaba's Qwen team has released Qwen3.7-Plus, a multimodal AI agent model that combines visual understanding, app operation, and coding in one system. In a demo, an agent built on this model autonomously created a vocabulary learning app, generating over 10,000 lines of code across 1,000 agent calls over eleven hours—all without human help. The model leads Qwen's own benchmarks for screen understanding but overall performance is mixed compared to Western competitors. It's a proprietary model priced well below OpenAI or Anthropic offerings.

OpenAI and the Trump administration are negotiating a direct government stake in the AI startup. The idea is a "Public Wealth Fund" that would pay out directly to American citizens. Senator Bernie Sanders wants to push through a 50 percent tax on AI shares by law. Critics fear the arrangement could create a "too big to fail" dynamic similar to the 2008 financial crisis. The article OpenAI and the Trump administration are negotiating a government stake in the AI startup appeared first on The Decoder .

AI has become the top reason companies cite when laying off workers—nearly 40% of job cuts announced in May 2026 were attributed to AI, up from just 7% in January. The first five months of 2026 saw 87,714 AI-linked job cuts, already surpassing all of 2025. While overall employment numbers remain positive, experts warn that workers in tech and other sectors face concentrated disruption, and many open positions don't match the skills of those being displaced.

US Senator Elizabeth Warren is pushing for Nvidia CEO Jensen Huang to appear before the Senate and explain how Nvidia's AI chips—particularly the H20 model—are being sold to China. Warren and other lawmakers argue these sales may undermine US efforts to restrict China's access to advanced AI technology that could be used militarily. Nvidia is already under existing export restrictions, but the H20 chip remains legal to sell after additional compliance requirements. The Senate Banking Committee, which oversees export controls, would be the forum for such testimony.

One NZ, a major telecom in Australia and New Zealand, cut its mobile provisioning time from 10 days to under 10 minutes using UiPath Maestro, an AI-powered automation platform. The rollout took just 5 weeks. This shows how enterprises are using AI and automation to slash operational costs and deliver services faster—potentially changing what customer service looks like across industries.

A new report confirms that AI-powered agents — bots that autonomously browse and interact with websites — now account for more internet traffic than actual human users. This milestone marks a shift in how the web is consumed, with AI systems scraping data, training on content, and performing tasks at scale. For everyday users, this means websites may prioritize serving AI over humans, and content creators could see their work fed to bots instead of readers.

Update on Anthropic calls for pause of global AI development. Coverage via Reddit/r/artificial.

An AI industry leader argues that AI agents are the next major technology shift—potentially as transformative as the browser or smartphone. The piece explores how agents could take over tasks like trip planning and cross-app coordination, why this demands new hardware designs, and how companies like Qualcomm are positioning for this shift. It's a forward-looking industry perspective on where AI is heading.

One NZ, one of Australia and New Zealand’s leading telecommunications providers, has cut enterprise mobile provisioning times from ten days to under ten minutes using UiPath Maestro, a cloud-native orchestration platform that unifies automation, AI agents, and human interaction into end-to-end business processes. The deployment was implemented in just five weeks and without a costly […] The post One NZ Cuts Mobile Provisioning from Ten Days to Ten Minutes with UiPath Maestro appeared first on techcoffeehouse.com .

Yahoo Singapore profiles five people who switched careers into AI roles, exploring how they made the transition and what they learned. The article offers practical takeaways for anyone considering a similar career move.

Senator Bernie Sanders has floated a dramatic proposal: a 50% tax on wealth created by AI and automation, with the money going directly to Americans. The idea is to make sure AI profits don't only benefit tech companies and billionaires. Political odds are long, but the debate is heating up.

# The quickest way to find blocked numbers on your iPhone - it's easy | ZDNET X Trending * [I wore the world's first HDR10 smart glasses](https://www.zdnet.com/article/rayneo-air-4-pro-hdr10-smart-glasses-ces/) * [TCL's new E Ink tablet beats the Remarkable and Kindle](https://www.zdnet.com/article/tcl-note-a1-nxtpaper-ces/) * [Anker's new charger is one of the most unique I've ever seen](https://www.zdnet.com/article/anker-wireless-chargers-ces-2026/) * [Best laptop cooling pads](https://www.zdnet.com/home-and-office/best-laptop-cooling-pad/) * [Best flip phones](https://www.zdnet.com/article/best-flip-phone/) * [Roku TV vs Fire Stick](https://www.zdnet.com/home-and-office/home-entertainment/roku-tv-vs-fire-stick/) * [Galaxy Buds 3 Pro vs Apple AirPods Pro 3](https://www.zdnet.com/article

AI tools are compressing hours of work into minutes for tech workers—drafting documents, summarizing meetings, reviewing code. But the time saved isn't translating to shorter workdays. Workers say any gains get immediately redirected to the next project, and some are actually working longer hours while building the automation systems. It's a candid look at the reality behind the productivity hype.

Nvidia just crashed the consumer laptop market with its new RTX Spark processor, capable of 1 petaflop of AI performance. Every major laptop brand — Dell, HP, Microsoft, Asus, Lenovo — announced new models powered by this chip. Microsoft unveiled the Surface Laptop Ultra as its flagship RTX Spark device, while Dell launched a $599 XPS 13 as a direct competitor to Apple's budget MacBook Neo. Most devices arrive this fall.

This week's security roundup covers serious privacy concerns: Meta quietly installed dormant face recognition code on over 50 million phones through its smart glasses app, capable of identifying people without their knowledge. Also, hackers exploited Meta's AI support chatbot to hijack Instagram accounts. The tech giant settled billions in biometric lawsuits just years ago over similar technology.

Researchers have released an open-source voice AI model called 'Audio Interaction' that listens continuously and decides every 0.4 seconds whether to respond or stay silent. Unlike typical voice assistants that wait for you to finish speaking, it processes audio in real-time — handling translation, transcription, chat, and ambient sounds like coughing in one continuous stream. Everything is free under Apache 2.0 license.

The Acer Swift Air 14 ($699) is pitted against Apple's MacBook Neo ($599 for students) as a direct Windows response to Apple's budget laptop disruption. The Swift Air wins on battery (19 hours vs 16), ports (dual USB-C Thunderbolt 4), and storage expandability (up to 1TB). The MacBook Neo wins on display quality, camera, and overall value at its price point. For students on a budget, the MacBook Neo remains the best bang-for-buck, though Windows rivals are catching up.

Want better AI images from ChatGPT or Gemini? A simple prompt tweak — adding specific style descriptors or lighting details — can dramatically improve results. The comparison shows both tools have strengths, but small language tweaks unlock significantly better outputs.

AI data centers are consuming so much memory that other industries face shortages. Cars, medical devices, and phones could become more expensive as memory chips get snapped up by AI companies. An industry group is urging the Trump administration to step in.

Meta is launching "Hatch," its first paid AI product — a monthly subscription potentially costing up to $200. Hatch lets users describe what they need in plain language, and the AI builds tools, schedules appointments, or sends emails on their behalf. CEO Mark Zuckerberg is positioning this as a major new revenue stream beyond advertising.

Elon Musk's xAI reportedly used Anthropic's Claude AI to train its own coding models for months, even after Anthropic cut off access. xAI continued using private accounts and a third-party service called Blackbox AI. Meanwhile, xAI's own pretraining team reportedly shrank to fewer than five people, with several leads leaving. The AI compute infrastructure Musk acquired is now being rented out to rivals Anthropic and Google instead.

Apple is giving Siri another major overhaul at WWDC, aiming to catch up after years of AI lag behind Google and OpenAI. The company appears ready to showcase deeper AI integration across iPhones and Macs. For everyday iPhone users, this matters because Siri is how millions interact with their devices daily — better AI could mean faster, smarter responses and fewer frustrating 'I don't understand that' moments.

Google DeepMind CEO Demis Hassabis says artificial general intelligence — AI that matches human-level thinking — could arrive by 2030. He calls it the start of a "new human era" that could bring medical breakthroughs and even a post-scarcity world. But he warns society has a narrow window to prepare for massive changes to jobs and daily life. Unlike some tech leaders who dialed back their more alarming predictions, Hassabis is still urging people to take the timeline seriously.

A helium shortage is threatening to slow down global chip manufacturing. Helium is essential for cooling the lasers used in semiconductor production, and supplies are tightening due to geopolitical disruptions and increased demand from tech and healthcare sectors. If the crisis worsens, it could constrain the supply of AI chips and drive up prices for everything from smartphones to data center servers.

Apple Watch users are stuck with a Siri that feels outdated compared to what's available on iPhone. The argument: Apple Watch is where AI assistance makes most sense—you need quick answers while your hands are busy—so fixing Siri there matters more than on the phone. Currently, Apple Watch Siri lacks features available on iPhone, creating a frustrating experience for the device most suited to voice commands.

Japanese AI startup Sakana AI has launched a dedicated lab to develop AI that recursively improves itself, potentially sidestepping the massive compute costs driving rivals like OpenAI and Anthropic to spend billions. The idea: smarter algorithms rather than bigger data centers. Co-founded by Transformer co-author Llion Jones, the firm bets efficiency beats brute force. Meanwhile, Anthropic warns similar self-improving AI carries serious control risks.

Meta is now filling its AI app with AI-generated clickbait articles instead of linking to real news. The standalone Meta AI app has a "For You" feed that populates stories entirely created by AI—including the topics, images, and text. This mirrors the low-quality content that has long plagued Facebook, but now the platform is manufacturing it itself. Users may struggle to distinguish between real news and AI-manufactured content.

Anthropic publicly warns about the dangers of authoritarian governments using AI while accepting billions in investment from Abu Dhabi's sovereign wealth fund MGX, which is controlled by an authoritarian monarchy. The company argues democracies must lead AI development, but its own investors include a regime known for surveillance, oppression, and hacking dissidents. Critics call this hypocrisy; Anthropic calls it a "communications headache."

Companies that rushed to replace workers with AI are having second thoughts. A new trend called the "AI boomerang" is emerging: firms that laid off employees due to AI productivity gains are now rehiring for those exact same roles. Research shows 55% of employers regretted AI-driven layoffs, and half of companies that replaced customer service staff with AI will rehire by 2027. The lesson? AI works best alongside humans, not instead of them.

The US President is meeting with AI companies next week to discuss sharing AI profits with the public. The proposal could reshape how AI wealth is taxed and distributed. While details are still emerging, this signals governments worldwide are eyeing AI companies' massive profits for public benefit.

A free, open-source macOS app called Council lets you ask the same question to multiple AI models like Claude, GPT, and Gemini at once. The models then critique each other's answers without knowing who said what, showing you exactly where they disagree. It's a practical tool for anyone who wants deeper analysis rather than trusting a single AI's opinion.

A $2 billion data center project in Shelbyville, Indiana has sparked community backlash, with residents posting 'No Data Center' signs. The mayor was caught on camera dismissing opposition by saying only people in 'shitty houses' would object. The controversy centers on local concerns about noise, property values, and environmental impact from the large facility needed to power AI systems.

Meta is so desperate for compute power to run its AI systems that it's building "data centers" made of massive weatherproof tents filled with AI chips. Under its "Prometheus" project in Ohio, the company has erected six 125,000-square-foot canvas structures powered by off-grid gas turbines. These tent structures can be deployed in months instead of the two to three years traditional buildings take. As communities increasingly reject data center projects, more companies may turn to these makeshift solutions to get AI chips online as fast as possible.

President Donald Trump said he's discussing deals "where the American people can benefit from the success of AI."

Cohere Labs has quietly released a new 30-billion parameter AI model called BLS-Mini-Code-1.0, designed for coding tasks. It's now available on Hugging Face for developers to try. The model uses a mixture-of-experts architecture and comes in BF16 format.

A Bank of America director is warning that AI stocks may be in a bubble and offering advice on how to position your portfolio if markets crash. The guidance centers on shifting toward bonds and defensive stocks rather than riding out volatility in AI-focused companies. This is speculative market commentary rather than established financial fact — but it reflects growing Wall Street nervousness about sky-high AI valuations.

The NSA is actively using Anthropic's Claude Mythos model for offensive cyber operations, even though the Pentagon has blacklisted the company and labeled it a supply chain risk. Anthropic has engineers working inside the NSA to customize the model for hacking operations targeting countries like China and Iran. This creates a bizarre situation where one arm of the US government bans Anthropic while another relies on its technology.

Americans lost nearly $900 million to AI-generated scams last year, and experts warn the problem is only getting worse. Scammers are using AI to clone voices, create deepfake videos, and craft highly convincing phishing emails that fool even careful people. This means anyone could be targeted — your elderly relatives, your coworkers, even you. The best defense is verifying unexpected requests through a separate channel and never trusting urgent money demands alone.

Sriram Krishnan, the former tech executive and venture capitalist who served as senior AI policy advisor to President Trump, is leaving his role at the end of June. Krishnan was a key figure in shaping the administration's pro-growth AI agenda, which prioritized building data centers over safety regulations. He says he'll now focus on building outside institutions to influence AI policy. His departure comes as the administration explores new ways to involve government in AI companies, including taking equity stakes in firms like OpenAI.

A developer built a game where five AI agents from different labs (OpenAI, OpenBMB, NVIDIA) each think independently and interact in a woodland economy. You play as a shadow financier lending, tipping, and shorting while a magistrate hunts for insider trading. Key takeaways: running heterogeneous models is mostly a config problem once serving infrastructure is solid; secret info given to AI WILL leak unless you firewall it at the data layer (not just prompt instructions); and keeping memory as bounded summaries keeps small models from drowning in context. The project is open-source.

OpenAI has launched a "Lockdown Mode" for ChatGPT that aims to block prompt injection attacks — a hacking technique where malicious instructions hidden in emails or documents trick AI chatbots into sharing private information like passwords or internal files. While not a complete fix, the mode adds an extra layer of defense for users handling sensitive business or personal data. If you've ever pasted a forwarded email or document into ChatGPT, this feature could help prevent accidental data leaks.