AI Global News Recap.

Imagine waking up to a notification that Elon Musk just deposited a fresh chunk of cash into your bank account. If the tech mogul managed to grow his net worth to a staggering $2 trillion and decided to distribute it equally among everyone in the United States, you would be getting a very real, very surprising payday.

Here is exactly how much cash you would get, and what that kind of historic giveaway would look like.

Your Personal Share of the $2 Trillion Pie

If we divide $2 trillion by the current U.S. population of roughly 349 million people, every single American—from newborns to billionaires—would receive exactly $5,730.

For households, math gets even more exciting:

• Couples: A combined payout of $11,460 👥
• Families of Four: A massive $22,920 👨‍👩‍👧‍👦
• An Average Apartment of Three Roommates: A cool $17,190

While it is not exactly “retire early” money, it is more than enough to cover a few months of rent, buy a premium used car, or finally take that dream vacation.

What Could America Buy with Elon’s Money?

If all 349 million Americans pooled their $5,730 shares together, the collective spending power would completely reshape retail. With $2 trillion, the American public could theoretically buy:

• 36 million Brand New Tesla: Enough to hand the keys of a Model 3 to more than 10% of the population.
• 1.6 billion iPhone 17s: Enough to give nearly five brand-new iPhones to every single citizen.
• 400 billion Chipotle Burritos: Enough to feed every person in America three burritos a day for well over a year. As fun as it is to imagine the ultimate stimulus check, a few massive economic roadblocks stand in the way:
• It’s Not Cold, Hard Cash: Net worth is not a bank account balance. Musk’s wealth is tied up in stock and ownership of companies like Tesla, SpaceX, and xAI. To give it away, he would have to sell all his shares.
• Still, as a thought experiment, it is proof that a $2 trillion fortune is almost unfathomably massive!
• Would you use your $5,730 share to buy a Tesla, invest it, or finally pay off some bills?

Meta just rolled out an improved Meta Account-think of it as a universal remote control for all your Meta apps and devices. Instead of juggling separate logins for Facebook, Instagram, Threads, and Al glasses, you now have one centralized hub to manage everything. The best part? Your daily experience stays exactly the same. This is a behind- the-scenes upgrade that makes life easier without disrupting what you already do.

The rollout is happening gradually over the next year, so don’t expect everything to change overnight. When Accounts Center updates to Meta Account on your profile, you’ll get a notification. If you’ve linked WhatsApp before, it’ll carry over-but you can remove it anytime. If you haven’t added WhatsApp yet, it stays separate and encrypted as always. The transition is designed to be seamless and non-disruptive.

Login just got simpler and smarter. You can now set up a single password for all your Meta apps and devices, eliminating the password fatigue of remembering multiple logins. When you download a new Meta app or set up a fresh device, your Meta Account gets you started in seconds-no need to create yet another profile from scratch. It’s convenience meets efficiency.

Security is woven throughout Meta Account without requiring extra effort on your part. Passkeys let you log in using your fingerprint or face recognition instead of typing passwords, and this feature now works across Instagram, Facebook, and Messenger with more apps coming soon. Behind the scenes, Meta’s security systems run 24/7 to detect and block suspicious activity before it becomes a problem. You’ll also get personalized security recommendations tailored to your account.

Managing your settings just became less chaotic. Core settings like your password, two- factor authentication, and account email now live in one place instead of scattered across multiple apps. But here’s the smart part: app-specific settings stay where they belong. Your Facebook privacy preferences, Instagram tagging rules, and other personalized choices remain in each app, so you can customize your experience exactly how you want it. For parents, Family Center makes supervising teens’ accounts on Instagram, Facebook, and Messenger a breeze from one dashboard.

Whether you’re a casual user or someone juggling multiple Meta devices, you have flexibility. You can create different Meta Accounts for different purposes-say, one for Facebook and another for Instagram-or keep WhatsApp completely separate. There’s no one-size-fits-all approach; Meta Account works the way you want it to work.

This is just the beginning. Meta Account is designed to evolve, with new features rolling out over time to make your digital life even simpler and safer. If you already use Meta Quest or AI glasses, you’ve been using a meta account all along now it’s expanding to become the unified hub for your entire meta ecosystem.

Smart glasses are quickly becoming one of the most exciting frontiers in wearable technology, blending everyday eyewear with powerful AI features that make life easier, more connected, and more hands‑free. Among the rising options, the DPVR AI Smart Glasses stand out as an accessible, feature‑packed choice designed for travelers, creators, and anyone who wants to capture moments without pulling out a phone. With built‑in translation, an 8MP camera, open‑ear audio, and a sleek lightweight design, they offer a surprisingly advanced experience at a budget‑friendly price—making them an appealing alternative to premium models like the Meta Ray‑Ban Smart Glasses.

Most people want smart glasses but don’t want to spend $ 300 or more.

DPVR gives them:

• Hands‑free photos
• Hands‑free video
• Bluetooth audio
• Voice assistant
• AI features Meta doesn’t have
• All for under $100.

“Why spend $300+ when you can get the same core experience for under $100?”

2. “DPVR does things Meta can’t do.”

DPVR offers:

• 141‑language real‑time translation
• Object & landmark recognition
• Built‑in storage
• Photochromic lenses included

Meta Ray‑Ban doesn’t include any of these.

“DPVR isn’t just smart — it’s your personal translator, tour guide, and POV camera in one.”

3. “Perfect for travel, work, and everyday life.”

• Traveling abroad → instant translation
• Biking or hiking → hands‑free POV video
• Work or school → record notes or scenes
• Daily life → capture moments without pulling out your phone

“These glasses turn your day into a hands‑free highlight reel.”

4. “The best entry‑level smart glasses on Amazon.”

Based on the Amazon page, customers say they’re:

• Great quality for the price
• Easy to set up
• Sleek and lightweight
• Surprisingly capable

“If you want to try smart glasses without spending big, DPVR is the smartest first step.

“DPVR G6 gives you hands‑free photos, AI translation, open‑ear audio, and POV video — all for under $100. It’s like Meta Ray‑Ban, but smarter for your wallet.”

“Record your life from your eyes, not your hands. DPVR G6 captures smooth POV footage, stores everything internally, and even translates 141 languages on the fly.”

“Explore the world with glasses that translate languages, identify landmarks, and record every moment — without ever pulling out your phone.”

“Why pay $300+ for smart glasses? DPVR gives you the essentials — camera, AI, audio, translation — for under $100.”

OpenAI’s Codex Gets a Major Upgrade: A Glimpse Into the “Super App” Future

OpenAI is rolling out a significant update to its Codex desktop application today, introducing a suite of powerful new features that hint at the company’s larger vision: a unified “super app” combining ChatGPT, Codex, and its Atlas web browser.

Background Computer Use: AI That Works Alongside You

The headline feature is background computer use—Codex can now autonomously control your PC, seeing, clicking, and typing across all your applications without interfering with your own work. Multiple AI agents can operate in parallel on macOS, each handling separate tasks.

“With background computer use, Codex can now use all of the apps on your computer by seeing, clicking, and typing with its own cursor,” OpenAI explained. “Multiple agents can work on your Mac in parallel, without interfering with your own work in other apps.”

While competitors like Claude offer similar capabilities, OpenAI claims its “secret sauce” lies in execution—keeping the system responsive so your PC remains usable while Codex works in the background.

Task Scheduling and Proactive Assistance

Codex can now schedule work days, weeks, or even months in advance, automatically waking itself to execute tasks at the designated time. The app is also becoming more proactive, using memory features to recall previous tasks and suggest next steps—like reminding you to respond to a colleague’s Google Doc comment at the start of your day.

Built-In Browser with Collaborative Feedback

A new in-app web browser lets you monitor Codex’s progress in real-time and provide targeted feedback. Using an annotation system similar to design collaboration tools, you can comment directly on page elements with specific instructions. In one demo, a team member instructed Codex to adjust graph margins to prevent axis cutoff—a simple but practical example of human-AI collaboration.

Image Generation and Enhanced Developer Tools

Codex now integrates gpt-4o-mini-1.5 for image generation, enabling it to create mockups, UI designs, and game assets. Combined with screenshot verification, this allows Codex to self-correct and ensure outputs match your vision.

For developers specifically, the update adds:

• Multi-terminal support

---

• GitHub review comment integration
• 111 new plugins expanding workflows across development and general knowledge work

Building the “Super App” in the Open

The unifying theme behind these disparate features is OpenAI’s ambitious “super app” strategy. Last month, The Wall Street Journal reported that OpenAI is developing a unified desktop application merging ChatGPT, Codex, and Atlas. Today’s release isn’t that app—yet—but it’s a major steppingstone.

“We’re actually doing the sneaky thing where we’re building the super app out in the open and evolving it out of Codex,” said Thibault Sottiaux, head of Codex, during a media briefing. “This release is about developers. In the future, we will broaden it up to a wider audience.”

Availability and Rollout

The updated Codex is now rolling out to macOS users logged into their ChatGPT accounts. Computer use and memory features are initially available in the US, with EU and UK access coming soon.

---

Google has made its Personal Intelligence feature for Gemini available globally. This feature personalizes the AI assistant experience by using data from a Google account, including Gmail, Google Photos, and YouTube search history.

Availability

The feature is available in most major markets. It is not available in the United Kingdom, Nigeria, South Korea, Switzerland, and the European Economic Area. Google is rolling out the feature in phases. Subscribers to premium tiers—Plus, Pro, and Ultra—will get access first. Free users will get access in the coming weeks.

Setup

To enable Personal Intelligence, users can access the feature through Gemini’s web interface. An activation prompt will appear. Users can also go to the Connect apps section within Gemini’s settings. Then, they can turn on the Google services they want to integrate with the assistant.

The feature is off by default to protect user privacy. Google will not use personal data, such as Gmail and Google Photos, to train models.

User Control and Privacy

The platform lets users control the Personal Intelligence feature. Users can turn off the feature for individual conversations. To do this, they select the Tools icon in the chat interface and turn off Personal Intelligence. This setting only applies to the current session. Subsequent conversations will use linked Google services unless changed.

Note: The feature may not be immediately available to all users.

Google launched the Gemini app for Mac, which is the first desktop version of the AI assistant. It is free on macOS 15 Sequoia and higher. The application brings Google’s AI capabilities directly to Mac users without needing a web browser or switching between applications. The app has over 100 features and was built by a small team in under 100 days. The app competes with other AI leaders in the desktop AI space.

The Gemini Mac app offers accessibility and convenience through design features that integrate with macOS. Users can access the AI assistant using keyboard shortcuts—Option + Space for a quick mini chat or Option + Shift + Space for the full experience. Users can also click the Gemini icon in the menu bar or Dock. This allows users to ask questions, draft content, and brainstorm ideas without interrupting their workflow.

The app’s functionality caters to many user needs. Gemini delivers quick answers, help writing emails and documents, summarizing articles, debugging code, and analyzing images. The ability to share your screen provides context for accurate responses. Integration with Google Drive allows users to upload files and access previous conversations. The app supports content generation, including images, videos, and music.

Gemini’s role on Mac is the beginning of Google’s AI expansion on Apple devices. The company announced a partnership with Apple to power the next generation of Apple Intelligence and an upgraded Siri, starting with iOS 27 and macOS 27 in 2026. This collaboration positions Gemini as a cornerstone of Apple’s future intelligence features as Google continues to roll out new capabilities on the Mac app.

Consider the scenario of a new mother, concerned about her infant’s breathing, who opts for telehealth rather than visiting a crowded emergency room to ensure her child’s safety. This decision is widely regarded as prudent. Similarly, Woebot Health represents an important development in mental health care. Like telehealth, Woebot Health provides a safer and more accessible avenue for support, normalizing help-seeking behavior at any time of day. As this approach is still in its nascent stages, further development and refinement are anticipated.

Significance of this approach

The core advantage is that Woebot treats the therapeutic alliance not as an optional extra, but as the foundation of effective care. In traditional therapy, the bond between the patient and clinician is among the strongest predictors of positive outcomes. Woebot’s creators are essentially trying to replicate that dynamic in a digital format — not by pretending to be human, but by using AI to deliver the right supportive message at the right moment.

Woebot’s rules‑based system gives it a kind of precision that generative chatbots can’t guarantee. Every response is pre‑written by clinicians and writers, then selected by AI based on context. That means the bot can be warm, consistent, and clinically aligned without drifting into unsafe or improvisational territory. It’s empathy with guardrails.

Practical benefits

• It builds trust before intervention.
• It personalizes without improvising.
• It supports diverse populations.
• It integrates into care systems.
• It sustains a “working alliance.

additional details: Woebot Health

Artificial intelligence is quickly moving from the margins of mental-health care into the mainstream, marking a notable shift in how therapy is delivered and accessed. Yet despite the momentum, AI-driven psychotherapy remains largely untested territory. Its appeal is clear: with psychiatric disorders rising worldwide—especially in underserved regions—these systems promise scale, affordability, and constant availability. But adoption is outpacing understanding. Before these tools become embedded in care systems, patients and clinicians alike need a grounded view of how they function, where they add value, and where their limitations begin.

One of the most pressing concerns is algorithmic bias. AI systems learn from human data, and that data is rarely neutral. Many models are shaped by Reinforcement Learning from Human Feedback, meaning their outputs reflect the assumptions and demographics of those who trained them. Evidence suggests that some systems are disproportionately influenced by male-centered data, raising questions about how accurately they interpret the experiences of women and other underrepresented groups. The result is not just a technical flaw but a clinical risk—one that can affect the quality, fairness, and cultural sensitivity of care.

There is also a fundamental gap between what AI can simulate and what therapy requires. These systems are highly effective at recognizing patterns and generating language, but they do not possess judgment, lived experience, or emotional depth. Therapy, at its core, depends on nuance—on the ability to challenge, guide, and sometimes confront. AI, by contrast, tends to default to validation. While that may feel supportive in the moment, it can limit meaningful progress over time. The distinction between sounding empathetic and actually understanding emotion remains a critical fault line.

Another emerging risk is dependency. Unlike human therapists, AI chatbots are always available, offering immediate responses at any hour. For some users, particularly those prone to anxiety or depression, that constant access can reinforce reassurance-seeking behaviors rather than build resilience. Over time, reliance on instant validation may erode a person’s ability to self-regulate or tolerate uncertainty—skills that are central to long-term mental health.

Beyond individual use, the broader implications are harder to ignore. Questions around data privacy, accountability, and clinical oversight remain unresolved. The rapid growth of AI mental-health tools has also led to a crowded marketplace, where not all products meet the same standards of safety or evidence. In some cases, harm has already occurred when tools were deployed prematurely. Taken together, these concerns point to a larger challenge: ensuring that innovation does not outpace responsibility. AI may well become a valuable complement to therapy, but replacing human care is a far more complicated—and far riskier—proposition.

For more information.

Artificial intelligence (AI) in psychotherapy: A challenging frontier – PMC Artificial intelligence (AI) in psychotherapy: A challenging frontier – PMC

As ChatGPT becomes a tool people rely on for everything from daily tasks to deeply personal moments, its role in supporting users during emotional distress has grown more urgent. Recent cases involving individuals in crisis have underscored the responsibility AI systems carry when interacting with vulnerable people. The priority is clear: ensure the technology recognizes signs of distress, responds with empathy, and guides users toward real‑world help without ever worsening a difficult moment.

To meet that responsibility, ChatGPT is designed with layered safeguards. The system avoids providing self‑harm instructions, uses supportive language, and directs users to crisis resources such as 988 in the U.S., Samaritans in the U.K., and global helplines. A network of more than 90 physicians across 30 countries — including psychiatrists and pediatricians — helps shape these interventions. When users express intent to harm others, conversations are escalated to a specialized review team, with law‑enforcement referrals possible in cases of imminent danger. Self‑harm cases, however, are not referred to authorities to protect user privacy.

Young people require even stronger protections. New teen‑focused safeguards include tighter guardrails around sensitive content, parental controls for oversight, and the option for teens (with parental approval) to designate a trusted emergency contact. These measures reflect the reality that many teens turn to AI before turning to adults, and the system must be prepared to respond safely and responsibly.

A growing area of concern is the rise of dangerous viral challenges that spread rapidly across social platforms. ChatGPT is being trained to detect when a user is influenced by or asking about a harmful trend, provide clear warnings, encourage safer alternatives, and guide teens toward trusted adults. The goal is to prevent the model from unintentionally validating risky behavior — a critical step as children and teens face increasing exposure to online stunts, misinformation, and peer‑pressure‑driven challenges.

Artificial intelligence is rapidly reshaping Brazil’s landscape—from major innovation hubs to urban centers and even more remote regions. As one of the most dynamic economies in South America and the Caribbean, Brazil is positioning itself as a regional leader in AI development.

Significant investment from both the government and international partners has accelerated this momentum, with billions of dollars committed to strengthening the country’s role in the global AI ecosystem. A key initiative driving this transformation is the Brazilian Artificial Intelligence Plan (PBIA), a multi-year strategy running from 2024 to 2028.

The PBIA spans critical sectors, including healthcare, education, social development, agriculture, trade, environmental management, and public services. Its goal is not only to modernize infrastructure but also to ensure that AI benefits are distributed broadly across society.

At the same time, workforce readiness is expanding at scale. More than 5 million young people and adults in Brazil have developed foundational AI skills, supported in part by programs like Microsoft’s Connect AI. This growing talent pool is positioning the country for a surge in new job creation and economic opportunities.

As Brazil continues to invest in both technology and human capital, it is laying the groundwork for a future where AI enhances everyday life, drives innovation, and fuels long-term growth.