Phantom Digital

31/12/2025

BREAKING: Meta Just Dropped Over $2 Billion On An AI Startup Nobody Saw Coming.

While everyone was watching OpenAI and Google fight it out, Mark Zuckerberg just made a move that caught the entire industry off guard.

Meta acquired Manus AI — a Singapore-based startup building autonomous AI agents. The price tag? Over $2 billion.

Let me break down why this matters way more than the headlines are telling you.

Meta bought Manus AI for over $2 billion — one of their largest AI acquisitions ever.

Who is Manus?

A company building AI agents that can actually do things autonomously. Not just chat. Not just answer questions.

We're talking:

→ Writing and debugging code
→ Conducting research
→ Analyzing data
→ Executing multi-step workflows
→ Operating with minimal human supervision

Think of it as the difference between asking ChatGPT a question versus having an AI employee that completes entire projects.

WHY THIS IS DIFFERENT FROM TYPICAL ACQUISITIONS

Most tech acquisitions are talent grabs or patent plays.

This one? Meta bought a functioning business with real revenue and millions of users.

Manus was already generating over $100 million in annual recurring revenue. That's not a research lab. That's a profitable company Meta just absorbed.

Translation: Meta didn't just buy technology. They bought an existing customer base and proven monetization.

THE THREE REASONS THIS CHANGES THE AI GAME

REASON 1: Meta Was Falling Behind

Let's be real. Meta has been playing catch-up in AI.
OpenAI has ChatGPT. Google has Gemini. Anthropic has Claude.

Meta had... Llama models that people respect but don't really use.
Manus gives them something none of their competitors have yet: autonomous AI agents that people are already paying for.

REASON 2: From Chatbots To Workers

Every major AI company is racing toward the same goal: AI that can complete tasks, not just have conversations.

OpenAI has been working on agents. Google is building them. But they're still mostly experimental.

Manus already has a working product with millions of users.

Meta just skipped years of development by buying the company that's already there.

REASON 3: The Geopolitical Play

Here's the part most people are missing:

Manus was founded in China but relocated to Singapore amid regulatory tensions.

As part of this deal, Meta is severing all Chinese ownership ties and shutting down Manus's China operations.

This isn't just a tech acquisition. It's a geopolitical positioning move.

Meta is saying: "We're building AI infrastructure outside of Chinese influence." That matters to regulators. That matters to enterprise customers. That matters strategically.

WHAT META PLANS TO DO WITH THIS

Manus won't just disappear into Meta's labs.

Here's the playbook:

→ Integration across platforms: Manus agent technology will power Meta AI, WhatsApp, Messenger, and Facebook

→ Continued standalone operation: Manus will keep selling its subscription service independently

→ Talent acquisition: Manus's entire team, including the CEO, is joining Meta's AI division

This is a hybrid strategy: absorb the tech AND keep the business running.
This isn't happening in isolation.

Meta also:

→ Invested billions in Scale AI earlier this year
→ Hired Alexandr Wang (Scale AI's CEO) as Chief AI Officer
→ Launched Superintelligence Labs with massive budgets
→ Is spending over $60 billion on AI infrastructure in 2025

Zuckerberg isn't just competing in AI. He's buying his way to the front of the race.

If you're building AI products, this should terrify you.

Meta just showed they're willing to spend billions to acquire companies with working technology and real users.

That means:

→ AI startups with traction become instant acquisition targets
→ Smaller competitors get priced out
→ The AI race consolidates around companies with infinite capital

We're moving from "can you build great AI?" to "can you outspend everyone else?"

Everyone wanted AI to be this open, decentralized revolution.

Instead, we're watching it become a winner-take-all game played by companies with billion-dollar war chests.

Meta just proved that if you have enough money, you don't need to build the future.

You can just buy it.

20/12/2025

Japan has shattered internet speed records by reaching 1.02 petabits per second, a connection so fast it could download the entire Netflix library in a single second. This achievement sounds unreal, but it marks a real world leap in how quickly data can move across networks.

The breakthrough shows just how far global connectivity has progressed and sets a new benchmark for the future of the internet. Speeds like this could transform cloud computing, artificial intelligence, and real time communication on a massive scale.


Source

04/09/2025

As detailed in a yet-to-be-peer-reviewed study, coauthors Petter Törnberg, AI and social media assistant professor, and research assistant Maik Larooij simulated a social media platform that was populated entirely by AI chatbots, powered by OpenAI's GPT-4o large language model, to see if there was anything we could do to stop social media from turning into echo chambers.

Researchers simulated a social media platform that was populated entirely by AI to see if we can stop them from turning into echo chambers.

02/09/2025

Takealot has relaunched its business-to-business (B2B) product, Takealot for Business, a procurement offering with 30-day interest-free terms, up to 5% rebates, and unlimited free delivery via TakealotMORE.

Takealot will showcase its offerings, including Takealot for Business, at the NSBC’s The Business Show Africa 2025 in September.

06/08/2025

China is building underwater data centers that use seawater for cooling and 97% wind energy for power to meet rising AI demand.

A $223M facility six miles off Shanghai will hold up to 792 AI-capable servers and launch in September.

It’s expected to train models like GPT-3.5 in a day while using 30% less electricity than land-based centers. Microsoft tested a similar idea with Project Natick in 2018 but shelved it.

Meanwhile, China is scaling fast. Environmental concerns remain, but South Korea and Japan are already exploring similar offshore data solutions.

02/08/2025

OpenAI just exposed thousands of private ChatGPT conversations to Google Search. And most users had no idea it was happening.

Your private chats were literally one toggle away from going public and showing up in Google results.

Here's what actually happened - and it's worse than you think.

OpenAI quietly introduced a "Make link discoverable" option within their new Shared Chats feature. Sounds innocent, right?

Wrong.

If you clicked that option, your chat became a public webpage that search engines could immediately index.

Google started crawling these conversations, and thousands of ChatGPT chats - many containing personal, professional, or sensitive content - began surfacing in Google Search results.

Why this was so dangerous:

- No clear warning: There was zero indication that enabling this option would make your content publicly searchable on Google.

- Basic web dev failure: OpenAI didn't add a "noindex" directive - web development 101 for keeping content off search engines.

- High-risk data exposed: The indexed chats included job applications, personal identifiers, business plans, technical code, and internal documents.

For a company pushing AI for enterprise use, this was a massive data governance failure.

Once this story started spreading, OpenAI pulled the plug within 24 hours.

They removed the "discoverable" option, started coordinating with Google to remove indexed links, and urged users to review their shared chats.

But the damage was already done. Cached versions of these conversations may still appear in search results.

What you need to do right now:

1. Go to ChatGPT > Settings > Shared Links
2. Review every conversation you've made public
3. Delete any shared links you don't want on the web
4. Wait for Google to remove the indexed versions

Sam Altman previously warned about this - "There's no legal privacy shield around your AI chats yet - it's screwed up, and it needs fixing."

This wasn't some sophisticated hack. This was a feature working exactly as designed - with very poor ex*****on lol.

Thousands of users accidentally made their private conversations public because of one unclear toggle.

And this won't be the last AI privacy incident. It's just the first one that got caught.

27/07/2025

In 2022, Google quietly wired Apple a mind-blowing $20 billion. There was no scandal. No lawsuit. No broken systems. Just one of the most strategic, low-key deals in the history of tech.

And yet, most people don’t even know it happened.

All Google wanted was one thing—to remain the default search engine on every iPhone, iPad, and MacBook around the world. That’s it. One slot. One silent placement that gave Google access to more than half the world’s mobile traffic.

That deal, that little setting most users never even change, fuels over a third of Google’s mobile ad revenue. Because when you open Safari and search something, it’s Google that serves the result. And Apple? Well, they collect a massive cut of that cash.

Back in 2014, this deal was just $1 billion. But as the stakes rose and iPhones spread across the globe, the price tag climbed with them. Fast forward a few years, and Google is paying Apple between $18 to $20 billion every single year. That’s not just pocket change—that’s more than the GDP of some countries.

Behind the scenes, this partnership is so valuable that Google considers losing it a “Code Red” event. They literally have internal crisis drills in case Apple ever walks away. On the flip side, Apple makes so much from this single partnership that some say it covers nearly 20% of its operating income.

What makes it even more fascinating is that Apple doesn’t even offer its own search engine. And despite being tech rivals in public, Apple continues to call Google the “best search engine” while raking in billions.

When Microsoft tried to get in on the deal, offering Apple 90% of Bing ad revenue, Apple didn’t even flinch. They stayed with Google. Not because they love them. But because the deal works—and it's invisible to most users.

This quiet alliance isn’t just lucrative—it’s also controversial. U.S. regulators have opened antitrust lawsuits over this deal, arguing that it locks out competition and gives Google unfair dominance. But while the courts argue, the checks keep clearing.

For entrepreneurs and founders, this story holds powerful lessons.

Google didn’t pay for ads. They paid for position.

And that position—being the default—unlocked one of the most valuable revenue streams in modern tech. Sometimes in business, the best seat at the table is not the loudest one. It’s the one people overlook. It’s the one baked into habit.

It also shows that traffic is the new gold. Not the app. Not the content. But the ability to direct and control where attention flows. Apple could have built their own search engine. But they didn’t need to. Letting someone else win—while getting paid billions—is a different kind of power.

We often think business success comes from being loud, innovative, or disruptive. But sometimes, the real power lies in quiet control. In unseen agreements. In being the one who decides what happens next—not necessarily the one who builds it.

And so, two of the biggest tech giants, who seem like enemies on the surface, are actually tied together in one of the smartest, most profitable deals ever made.

It’s a silent alliance. A strategic position. And a masterclass in how power really works in the modern world.

Let it sink in: Google pays Apple billions… just to stay in your search bar.

That’s the level of positioning every entrepreneur should aim for.

Not just to be seen—but to be default.

It does appear there is a pawn in this game. Can you guess who it is?

22/07/2025

Meta recently said it has removed about 10 million Facebook profiles for impersonating large content producers so far in 2025, in an effort to reduce spam. This mandate comes at a time when AI is making it easier to mass-produce content throughout social media platforms.

Meta on Monday said it has removed about 10 million Facebook profiles for impersonating large content producers so far in 2025, in an effort to reduce spam.

13/07/2025

Jack Dorsey has announced a new decentralized messaging app called Bitchat. It’s a peer-to-peer public domain messaging project that operates over Bluetooth Low Energy (BLE) mesh networks. Bitchat is said to offer a more secure method of messaging, protected by end-to-end encryption. The app doesn’t require internet, an account, or even a phone number to function.

25/06/2025

A massive data breach exposed 16 billion passwords from platforms like Apple, Google, and Facebook, the largest leak in history.

Reused or weak passwords heighten risks of identity theft and fraud, with datasets circulating among criminals.

Experts urge immediate password changes, using password managers, enabling 2FA, and adopting biometric passkeys.

24/06/2025

According to the company, ads won’t show up inside your chats. Instead, they will appear in the Updates tab, a feature launched in 2023 that lets users follow public Channels for news and announcements.

These channels, used by everyone from local governments to content creators, will now be able to pay for promoted placement, boosting visibility to WhatsApp’s 1.5 billion daily users.

Meta has announced that it is adding ads and subscriptions to its WhatsApp messaging app

24/05/2025

When asked at Anthropic's first developer conference, Code with Claude, when the first billion-dollar company with one human employee would happen, Amodei confidently responded, "2026."

Thinking of launching a startup? AI could be your staff.