Welcome to Thursday’s Newsletter

In today’s scoop 🍨 

• 🔎 DeepSeek vs. OpenAI: AI Showdown or Data Theft?

• 🚖 Tesla’s Robotaxi Revolution (or Another Musk Mirage?)

• 💰 Big Tech's AI Spending Spree: No Brakes, No Budget Caps

• 🔧 3 Trending AI Tools

🔎 DeepSeek vs. OpenAI: AI Showdown or Data Theft?

OpenAI claims DeepSeek may have harvested its AI-generated data to train its own rival models—a move that could violate OpenAI’s terms of service. This technique, called distillation, involves training one AI model using outputs from another. Think of it as the academic equivalent of copying someone’s homework and then winning the science fair.

📈 The Evidence (So Far)

📃 Microsoft detected what looked like DeepSeek-linked groups siphoning massive amounts of data via OpenAI’s API. They flagged it, and OpenAI is now "reviewing indications" that its tech was "inappropriately distilled."

🌐 DeepSeek’s new AI model, R1, is reportedly performing at OpenAI and Google Gemini levels—without needing the same billion-dollar compute resources. That’s like building a Ferrari with spare parts from a used Toyota.

🕵️ The U.S. government is watching. AI czar David Sacks told Fox News that there’s “substantial evidence” DeepSeek piggybacked on OpenAI’s work.

🎮 The Bigger AI Power Game

If DeepSeek did train its models using OpenAI’s data, this raises some spicy questions:

• 🤨 Does OpenAI have a legal case? The AI industry has been a bit…gray on data usage rules. OpenAI itself is facing multiple lawsuits for allegedly scraping copyrighted content.

• 💪 Can Silicon Valley keep its edge? DeepSeek’s ability to compete with far less compute power could shake up assumptions about AI’s cost barriers.

• 🚫 Will API access get stricter? If OpenAI and Microsoft fear more “AI cloning,” we might see tighter restrictions on who can use AI APIs and how much data they can pull.

🚀 What’s Next?

Expect more tech world drama as OpenAI digs deeper into what happened. If DeepSeek did cut corners, it could face backlash from the industry and regulators. But if it found a way to build competitive AI without crossing ethical or legal lines? Then OpenAI—and the rest of Silicon Valley—might need to rethink the way they play the game.

One thing’s for sure: AI isn’t just a battle of compute power anymore. It’s a battle of strategy, speed, and, apparently, who can out-maneuver whose terms of service.

🚖 Tesla’s Robotaxi Revolution (or Another Musk Mirage?)

Elon Musk is at it again—promising the future, one ambitious deadline at a time. This time, he’s rolling out self-driving robotaxis in Austin, Texas, this June. No drivers. No steering wheels. No pedals. Just pure, unfiltered AI taking the wheel. What could possibly go wrong?

🚗 The Details (or Lack Thereof)

Musk announced that Tesla will finally launch a paid ride-hailing service, using its own fleet of vehicles equipped with a yet-to-be-released “unsupervised” Full Self-Driving (FSD) software. Key points:

• 🚖 No human drivers – These cars will be fully autonomous, at least according to Musk.

• 📍 Austin gets the first taste – The trial starts in Texas, with plans to expand later.

• ⏳ Customers' cars stay parked – Unlike past promises, Tesla owners won’t be able to list their personal vehicles in the fleet until a later date yet to be announced.

• 🛶 “Toe in the water” approach – Musk claims Tesla is starting slow to ensure safety, but he’s also hyping 2025 as “the most important year in Tesla’s history.”

👏 A Bold Play, But Can Tesla Deliver?

Tesla has been teasing self-driving capabilities for years, with Musk promising autonomy was just one or two years away… repeatedly. Now, the timeline has drastically shortened—Tesla’s claiming this isn't a “far-off, mythical situation” but a real-world launch in just a few months.

But there’s skepticism:

• 🚩 Regulators are watching – The National Highway Traffic Safety Administration (NHTSA) is already investigating Tesla’s FSD tech after multiple collisions.

• 🤖 No lidar? No problem? – Unlike Waymo and other AV companies, Tesla relies only on cameras, avoiding lidar sensors entirely.

• ⚠️ Safety questions linger – Tesla says FSD (Supervised) has logged over 3 billion miles, but critics argue its safety metrics lack transparency.

🎭 Hype or History in the Making?

If Tesla actually pulls this off, it could redefine urban transportation. If not? Well, it wouldn't be the first time Musk set a deadline he couldn’t keep. Either way, Austin is about to become the ultimate self-driving experiment.

💰 Big Tech's AI Spending Spree: No Brakes, No Budget Caps

In the AI arms race, Big Tech is throwing down cash like a billionaire at a Vegas casino—except the stakes are even higher, and the competition just got a whole lot cheaper.

China’s DeepSeek turned heads when it claimed to have developed an AI model that rivals Western tech giants, all while spending only $5.6 million (peanuts compared to the billions burned by Microsoft, Meta, and friends). But if you thought this would make Big Tech slam the brakes on spending, think again.

Instead, Mark Zuckerberg and Satya Nadella are doubling down, defending their massive AI budgets as the key to long-term dominance.

📊Meta: No Budget? No Problem.

• Zuck isn't fazed. Meta is set to spend a staggering $60-65 billion this year on AI infrastructure, an eye-watering 70% higher than what analysts expected.

• The company sees AI personalization as its secret weapon, betting that more computing power means better, smarter AI services.

• As for DeepSeek? Zuck acknowledged its innovations but insisted that heavy investment in AI hardware and infrastructure will be a long-term advantage.

• Oh, and Meta’s next big AI model, Llama 4, is aiming to be the world’s best, taking on even closed models like OpenAI’s ChatGPT.

💰 Microsoft: Spending to Win

• Microsoft is keeping pace, planning to pour $80 billion into AI this fiscal year, with spending continuing to grow in 2026.

• Satya Nadella believes that as AI gets cheaper, demand will skyrocket—so why not build the infrastructure to cash in?

• Microsoft is also making its AI infrastructure more flexible, ensuring its "fungible fleet" of data centers can pivot between training and deploying AI models worldwide.

• Meanwhile, its cozy relationship with OpenAI is raising eyebrows, with rumors swirling that DeepSeek may have trained its model by scraping OpenAI’s work.

🔑 Takeaway

Big Tech is not hitting the brakes—if anything, they’re flooring it. Meta and Microsoft are betting that AI’s future belongs to those who can scale, personalize, and dominate the infrastructure game. But with investors getting nervous about sky-high AI spending, one big question looms: Will these billion-dollar bets pay off, or will DeepSeek prove that innovation doesn’t have to cost a fortune?

Either way, the AI race isn’t slowing down—it’s just getting more expensive. 

🔧3 Trending AI Tools

• 🎬 Pika 2.1 - AI-powered video creator that transforms your wildest ideas into stunning 1080p visuals, making the impossible possible—whether it's landing on the moon or remixing reality.

• 📅 EpicTopia AI - AI-powered personal pursuit manager that streamlines goal-setting, planning, and journaling with multi-goal timelines and growth insights to help you stay proactive and fulfilled.

• 🗣️ PitchAvatar - AI-powered chat avatars that adapt to roles, speak and listen in multiple languages, and present dynamic content like slides, images, and videos for engaging, goal-oriented interactions.

Thank you for reading!

Got thoughts on today’s edition? Something you loved, learned, or laughed at? Hit reply and let us know—your feedback fuels the fun!

And if you’ve got a friend who’s into AI or loves staying sharp, pass this along. Let’s spread the AI insights and grow this community!

Until tomorrow—stay curious! 👋