DeepSeek vs. The Zuck

Welcome back apprentices! 👋

Yes, yes, yes, we know — you’re probably so over DeepSeek news. It’s everywhere! 

But guess what? Millions of Americans clearly didn’t get the memo because they’ve downloaded it since January 15 like it’s the hottest new dating app.

Now, the fun part: NASA and the U.S. Navy have already hit “uninstall” faster than you close a spam email. Why? Because cybersecurity experts say DeepSeek’s privacy policy is about as solid as a wet napkin. If you love handing over your personal data like free candy, well... carry on.

Anyway, that’s (almost) all the DeepSeek drama for today — don’t worry, the AI chaos is just getting started!

In today's email

• AI for Hackers 

• AI’s Naughty List

• OpenAI Joins the U.S. National Labs

• Reid Hoffman’s New AI Startup

• Why Meta Isn’t Sweating DeepSeek

• More Tools & Updates

_{Read Time: 4 minutes}

Quick News 

🕵️‍♂️ Cybercriminals just unlocked easy mode with GhostGPT, an AI chatbot that writes malware, crafts phishing emails, and automates scams — all for a small fee on Telegram. Unlike regular AI that shuts down shady requests, this one leans in and takes notes. Even wannabe hackers with zero tech skills can launch cyberattacks in seconds.

🚨 The European Union has officially put AI systems with “unacceptable risk” in the penalty box — goodbye, creepy social scoring and Minority Report-style predictive policing!  This new ban, effective February 2, 2025, is all about keeping AI ethical and ensuring it doesn’t go full sci-fi dystopia on us.

🤖 OpenAI just teamed up with U.S. National Laboratories, giving 15,000 government scientists access to its top AI models for causal research — like nuclear security, cybersecurity, and disease prevention. Oh, and they’re also deploying AI on a supercomputer at Los Alamos (yep, that Los Alamos).

💊 LinkedIn co-founder Reid Hoffman just swiped right on drug discovery with his new AI-powered startup, Manas AI. Partnering with Pulitzer Prize-winning cancer researcher Siddhartha Mukherjee, the company just secured $24.6 million to supercharge AI-driven medical breakthroughs — because, apparently, networking and curing diseases are on his résumé. 

Together with The Rundown

Start learning AI in 2025

Everyone talks about AI, but no one has the time to learn it. So, we found the easiest way to learn AI in as little time as possible: The Rundown AI.

It's a free AI newsletter that keeps you up-to-date on the latest AI news, and teaches you how to apply it in just 5 minutes a day.

Plus, complete the quiz after signing up and they’ll recommend the best AI tools, guides, and courses – tailored to your needs.

Sign up to start learning.

Meta 
Why Meta Isn’t Sweating DeepSeek 

DeepSeek’s AI is making waves, racking up downloads, and getting government bans faster than a crypto Ponzi scheme — but Mark Zuckerberg? Not impressed. 

On Meta’s recent Q4 earnings call, he basically patted DeepSeek on the head, called it “cute,” and went back to counting billions while working on Llama 4, Meta’s next-gen AI.

The message? DeepSeek is the new kid on the AI playground, but Meta owns the entire school.

Meta’s Financial Flex: “We’re Rich, Thanks for Asking”

While AI startups are still burning investor money like it’s a campfire, Meta’s Q4 2024 earnings look like this:

• Revenue: $48.39 billion (+21% YoY) – That’s nearly $550,000 per minute.

• Net Profit: $20.84 billion (+50%) – Or, $228 million per day.

• Ad Revenue: $46.78 billion – Because Instagram is free, but somehow still makes more than entire nations.

Moral of the story: AI may be the future, but Meta’s bank account is already there.

Llama 4: The Bigger, Smarter, Hungrier AI Model

Zuckerberg confirmed that Llama 4 is coming, and instead of just pumping it full of more GPUs, Meta is making it smarter.

• Better Reasoning: AI that doesn’t just predict words, but actually thinks (like a college student after coffee).

• Mass Adoption: Goal? 1 billion users using Meta’s AI assistant by the end of 2025.

• AI Assistant Everywhere: Think AI in your Instagram DMs, WhatsApp chats, and Messenger conversations — predicting your responses before you even type. Creepy? Maybe. Useful? Absolutely. 

Basically, Meta wants its AI to be your personal assistant, best friend, and nosy neighbor — all at once.

Why Meta Isn’t Worried About DeepSeek (Or Anyone Else, Really)

Zuck laid out several reasons why Meta isn’t even blinking at DeepSeek:

1. Meta Has More Users Than DeepSeek Will Ever Dream Of

   3.35 billion people use a Meta app daily — which means Meta has more real-world AI training data than DeepSeek, OpenAI, and Anthropic combined. 

2. Meta’s Infrastructure Budget Is Bigger Than Some Countries’ GDP

   💸 $60-65 billion investment in AI infrastructure for 2025 alone. Meta is building new AI supercomputers like it’s a Call of Duty DLC drop. 

3. DeepSeek Has No Business Model, Meta Has a Printing Press

   📈 AI startups are still figuring out how to make money. Meanwhile, Meta’s $47 billion ad business is funding its AI dreams on autopilot.

In short: Meta doesn’t need to “win” the AI race — it already owns the stadium, the scoreboard, and the referees.

What This Means for You

Meta isn’t just playing the AI game — it’s rigging it in its favor. Here’s what anyone should learn from Zuck’s latest flex:

🔹 AI isn’t just about raw power — it’s about distribution. Meta’s massive user base gives it a monopoly on AI reach that few can match.
🔹 Smarter AI > Bigger AI. The shift toward reasoning-focused AI means businesses should think about how AI can enhance decisions, not just automate tasks.
🔹 AI-driven products need deep pockets. If you’re an AI startup, find a solid revenue model (or a very patient investor) before taking on giants like Meta.

Final Takeaway

DeepSeek’s AI is good — but Zuckerberg’s response was basically:

"That’s cute. Now watch us build something even bigger."

Meta isn’t flinching. If anything, it’s doubling down, spending more, and setting the AI agenda for the next decade.

DeepSeek? Might be the underdog story we all love.

Meta? Still the one writing the rules.

Tools & Updates 

📞 Hate making phone calls? Google gets it. Their new AI-powered features — "Ask for Me" and "Talk to a Live Representative" — now handle awkward business inquiries and soul-crushing customer service hold times for you. One books your nail appointment (without the small talk), while the other waits on hold (so you don’t have to listen to that awful elevator music).

🎵 Ever wanted to drop a hit single but lack talent, instruments, or rhythm? No problem — Fuzz is here. Riffusion’s new free AI music platform lets you create full-length tracks from text prompts, audio snippets, or even images (because apparently, your cat photos can inspire bangers now). Plus, it learns your musical taste over time, meaning your AI DJ will know you better than your Spotify Wrapped. Oh, and The Chainsmokers are on board as advisors, because why not?

📊 Hate wrestling with spreadsheets? Google heard you again. Gemini AI just landed in Google Sheets, letting users analyze trends and create charts — without touching a single formula. Just ask, and Gemini will generate insights and static visualizations (but don’t expect them to update automatically — this AI still makes you work a little). Rolling out now for Workspace and Google One AI Premium users ($19.99/month), with full access by February 20th.

Let's Chat About AI

This is all for today!

Replay to this email with your thoughts.

AI is more than just a buzzword. It’s a shift in how we live and work. And understanding it a bit better means you can make smarter choices about the tech you use every day.

_FEEDBACK