Space Power, Physics Fails
· The Fluency Briefing
The Fluency Briefing
Your Guide to What's Happening in AI and Why It Matters to You
Monday, April 27, 2026
An 8-year-old is running a six-figure business with ChatGPT as his design team, Meta is trying to beam solar power from space to keep its data centers humming at night, and a new study just gave every AI model a flat zero on reproducing physics experiments. The thread connecting these stories?
AI keeps rewriting what's possible - and what's clearly not - in ways nobody predicted.
Today in AI:
- Two Kids, One AI, and a Stuffed Animal Empire - An 8-year-old and his 10-year-old brother built a $100K custom plush toy business using ChatGPT to turn hand-drawn sketches into manufacturer-ready designs. Their clients include Reddit and a marketing agency, proving that AI-native entrepreneurship starts younger than you think. Forbes
- Meta Wants to Beam Sunlight From Space (Yes, Really) - Meta signed a deal with startup Overview Energy to potentially deploy a thousand satellites that beam infrared light to solar farms at night, powering data centers around the clock. The company used over 18,000 gigawatt-hours of electricity in 2024 alone, so the motivation is clear. TechCrunch
- AI Flunks Physics Lab - A Peking University preprint found that every LLM tested, including OpenAI's Codex powered by GPT-5.3, scored a 0% success rate at reproducing numerical results from experimental physics papers. They understood the methodology fine - they just couldn't do the actual work. LessWrong
- Your Mall Is Getting an AI Makeover - A McKinsey and ICSC survey of 3,000+ US consumers found that AI is reshaping how people discover and buy products, meaning store visits will become less frequent but more intentional. Retailers have three to five years to figure out which physical locations still earn their keep. Onestepretail
- AI Layoffs Aren't Really About AI - Writer Daniel Miessler argues the real dynamic behind AI-driven layoffs is that companies no longer want to pay for average performers when top performers armed with AI can do 10x the work. The takeaway: become irreplaceable, or become independent. Writer
- Spotify Gets Sweaty With Peloton - Spotify launched a fitness hub Monday with over 1,400 Peloton-led workout classes baked into Premium subscriptions, aiming to become a daily wellness habit rather than just a music app. The move pits them directly against Apple Fitness+ and YouTube for your workout attention. Routenote
- Your AI System Is Quietly Lying to You - VentureBeat reports that the most dangerous enterprise AI failures produce no errors and trigger no alerts - they just give confidently wrong answers built on stale data and broken orchestration logic. Traditional monitoring tools weren't designed to catch this kind of drift. VentureBeat
- Your Brain on ChatGPT Needs a Firewall - A new practical guide based on 30+ studies outlines concrete steps to protect your cognitive skills while using AI chatbots, covering everything from critical thinking erosion to emotional dependency. Think of it as digital hygiene for the age of AI assistants. The Algorithmic Bridge
Today's Takeaway:
Here's the thing about AI in 2026: it's simultaneously enabling an 8-year-old to run a real business and flunking basic physics homework. That contradiction isn't a bug - it's the clearest picture we have of where this technology actually stands. The Fuller brothers' plush toy company, as reported by Forbes, works because turning a kid's sketch into a product rendering is exactly what generative AI excels at: pattern matching, visual transformation, creative interpolation. Meanwhile, the Peking University study covered by LessWrong shows that the moment you need genuine reasoning about physical systems - translating real-world conditions into precise numerical simulations - every model falls apart.
Now layer in the enterprise picture. VentureBeat describes AI systems that look perfectly healthy on every dashboard while confidently delivering wrong answers, and Daniel Miessler's analysis argues companies are betting that top performers plus AI can replace entire teams. Connect those dots and you get something uncomfortable: businesses are restructuring around a tool that's brilliant at some tasks and silently incompetent at others. The implication chain matters here. Companies fire average performers. Remaining staff rely more on AI. But if nobody's monitoring whether the AI is actually correct - and most monitoring tools can't tell - the whole strategy rests on a foundation nobody is stress-testing.
💡 Fluency Moment - Building your AI fluency, one term at a time.
"Zero-shot Learning"
In plain English: AI solves tasks it was never explicitly trained on, using general knowledge alone.
Think of it like: Like a student acing a test on a topic they never studied, just by thinking it through.
Why you'll hear about it: AI scored zero on physics labs despite never being taught to fail - limitations matter.
🧰 Your Toolkit
Try This Prompt: Exploring AI in Everyday Life
Explain how AI is changing the way people shop in stores, like I'm hearing about this for the first time. Give me 3 real examples I might notice next time I visit [YOUR FAVORITE STORE, e.g., a grocery store or clothing shop]. I want to understand if AI can think or feel emotions. Explain it simply, like you're talking to someone with no tech background, and use a comparison to something from everyday life like [A PET, A CALCULATOR, OR A TOASTER]. Summarize the biggest tech news from this week in plain English, focusing on stories about [AI, COMPUTER CHIPS, OR GAMING], and explain why each story matters to someone like me who just wants to stay informed. I'm thinking about building or buying a new computer for [GAMING, WORKING FROM HOME, OR SCHOOL]. In simple terms, what should I look for, and what questions should I ask before spending my money? Without using technical words, explain what computer chips have to do with AI getting smarter, and why companies like [APPLE, NVIDIA, OR TSMC] are talked about so much in the news lately.
For the best results, always tell the AI a little about yourself at the start - for example, 'I'm new to tech' - so it keeps the answer simple and relevant to you.
The Bottom Line
The Pattern: AI is brilliant at creative translation and terrible at precise reasoning, yet businesses and families alike are restructuring their lives around it as though it's uniformly capable. The gap between what AI can do impressively and what it quietly botches is the defining tension of this moment.
Why It Matters: If you're a business owner handing more responsibility to AI tools, or a professional competing against colleagues who are, understanding that gap isn't academic - it's survival. The companies that thrive won't be the ones that adopt AI fastest. They'll be the ones that know exactly where AI's confidence outstrips its competence.
Your Move: This week, pick one AI-assisted workflow you rely on and audit the output like you would a new hire's first assignment. Trust, but verify - before the silent failures become expensive ones.
What We're Working On
✨ Founding Cohort Special - 60% Off! - Use code MAF20 to join for just $20/month (regularly $50). Get weekly group sessions & workshops, self-paced courses for all levels, access to tools & templates, challenges with peer feedback, and 24/7 support community. → Join Now
✨ Free 30-Minute AI Consultation - Discover how My AI Fluency can help your business unlock the potential of AI. We'll discuss your goals, explore practical AI opportunities for your industry, and outline clear next steps. → Schedule Free Call
✨ How AI-Fluent Are You? - Test your AI fluency with our interactive quiz. See how you stack up and discover what to learn next. → Take the Quiz
💬 Community | 📞 Book a Consultation | 🌐 Website
Fluently yours, The My AI Fluency Team