Good morning 👋 to anyone who’s spent days building strategy with a chatbot, only to realize it was making things up as it went along. We’ve all been there. Financial forecasts look polished, the numbers feel solid, and you’re ready to execute the plan. Then reality hits, and you realize you’ve been led down a rabbit hole. Seems like AI’s favorite growth hack is imaginary revenue.
✨ Are smart glasses ready to replace your phone? Yesterday Meta, Snap, and Brilliant Labs leaked rolled out big, AI-powered upgrades that promise a huge leap forward. Imagine AI glasses translating conversations in real-time, identifying anyone you meet, or instantly answering your questions without touching a screen. The next killer app could literally be right in front of your eyes. If they’re right, the smartphone era is over and Silicon Valley is about to chase the next boom.
🦾 OpenAI knows GPT-5 won’t get them to AGI, so they’re pivoting to robots. Recent hires show they’re bringing in robotics experts and engineers, while teaming up with humanoid robot startup Figure AI. The plan is to train algorithms through virtual simulations and remote control. OpenAI’s betting real intelligence needs a physical body, not just a chat window. Pretty sure this is exactly how Ex Machina started.
🚲 SpaceX built its docking system out of spare bike parts. NASA spent two decades engineering a complicated design with 6 custom actuators to absorb spacecraft impacts. Elon’s team skipped all that, came up with something simpler, and let a young engineer order mountain bike shocks for the “McDocker” online instead. NASA thought they were idiots… until it worked flawlessly. Sometimes ignoring the experts pays off.
♻️ AI models have run out of clean training data and are dumpster diving. LLMs have scraped the usable internet dry, leaving mostly toxic rants, fake news, and sensitive personal data. But Google DeepMind thinks it found a fix: just use AI to sanitize all that sketchy data into something safe through Generative Data Refinement (GDR).* Yeah, sanitizing 4chan with hallucinating AI should definitely solve the misinformation problem.
*not to be confused with Macro Data Refinement (MDR) or Severance.
THE AI DATA ANALYST YOU CAN TRUST
Tired of messy spreadsheets and guessing with ChatGPT? Julius AI turns complex data into clear insights in seconds, no coding required. Loved by 2M+ users and trusted by Zapier and Princeton University.
Why you need it: Julius is smarter than ChatGPT for data analysis. It can tackle problems generic AI can’t. Upload your data, ask a question, and get real analysis.
Skip repetitive explanations. Julius remembers your context and turns results into clear visuals that explain your findings.
☢️ AI is making social media even more toxic. Researchers let 500 chatbots loose on their own social network, only to watch them form cliques, amplify extremists, and let a few loud voices dominate. All driven by social media’s own reward system. Now social platforms are drowning in synthetic influencers and a flood of AI slop. It’s all starting to push people toward private communities where connections are still authentic. Not to be harvested, but to be heard.
funny money 🤑
👄 AI influence: Sponsored posts by human creators generate 2.7x more engagement than AI influencer campaigns. They also cost an average of $78,777 vs. $1,694 for the bots. [Financial Times]
🚀 Sky high: The world’s 3 richest people (Elon Musk, Larry Ellison, and Mark Zuckerberg) are now worth a combined $1 trillion thanks to AI optimism. None of them have gone to space. [Business Insider]
💸 Prosperity paradox: In 2021, the year before ChatGPT’s debut, the IT sector contributed 0.22% to real GDP growth in the US. By 2025, it jumped 150% – yet IT jobs growth plummeted by 87%. [Inc]