Bugs, Backlash, and Debt

2026-05-23 · The Fluency Briefing

The Fluency Briefing

Your Guide to What's Happening in AI and Why It Matters to You

Saturday, May 23, 2026

Here's something most people won't piece together from this Saturday's headlines: the same week AI found over 10,000 software vulnerabilities that humans missed, people started asking plastic surgeons to make their faces look like AI-generated selfies. Meanwhile, Oracle's massive AI bet is quietly teetering toward junk-bond territory. The gap between what AI can actually do and what we think it's doing has never been wider - and that gap is where the real story lives.

Today in AI:

• Your Face, But Make It Algorithmic - Plastic surgeons report a surge in patients arriving with AI-beautified selfies, asking for procedures to match impossible results. Doctors are calling it "AI face" - a new body-image crisis driven by chatbot recommendations. The Guardian
• Anthropic's Bug-Hunting Beast Hits 10,000 Vulnerabilities - Claude Mythos Preview, the model Anthropic considers too dangerous to release publicly, helped partners find over 10,000 software bugs in its first month. Cloudflare alone found 2,000, with 400 rated high or critical severity. Engadget
• Oracle's AI Debt Pile Is Starting to Smell - Oracle now carries over $160 billion in liabilities as it builds out massive AI data centers, including the Stargate campus in Texas. Its credit default swap market is surging, and its stock has fallen more than 43% since September's peak. Fast Company
• AI Gets Booed at Graduation - AI's public image took a beating this week: boos at commencement speeches, a book dragged for AI-fabricated quotes, and a Stanford poll showing less than half of Americans support charging ahead with AI innovation. Fast Company
• Huawei Engineers Around US Sanctions With a 122TB SSD - Cut off from advanced 3D NAND chips, Huawei built a workaround using Die-on-Board packaging that mounts more NAND dies directly on the PCB. The result: a 122TB SSD for AI data centers, with a 245TB version coming. Tom's Hardware
• Qualcomm's Stock Rockets 75% in a Month - Investors are betting big on Qualcomm's role in edge AI - the chips inside your glasses, car, and phone. A reported partnership with OpenAI to build an AI-agent-powered device is adding fuel. CNBC
• Google's Anything-to-Anything AI Model Gets Hands-On - The Verge tested Gemini's latest multimodal capabilities and found that making realistic AI-generated video is now shockingly easy and requires almost no expertise. The deepfake-to-fun-project pipeline is getting very short. The Verge
• AI Drug Discovery Targets Brain Diseases - UK researchers are using AI to analyze voice recordings, eye scans, and lab-grown brain cells to identify existing drugs that could treat conditions like motor neurone disease. The goal: find treatments in years instead of decades. BBC

Today's Takeaway:

On the surface, three of this week's stories seem unrelated: Oracle's debt spiral, the AI backlash at graduation ceremonies, and Anthropic sitting on a model it won't release. But together, they reveal a single uncomfortable truth - the AI industry is outrunning its own ability to prove its value. Oracle borrowed over $160 billion to build infrastructure for AI workloads that haven't fully materialized yet, as Fast Company details. Meanwhile, public trust is eroding. A Stanford and UC Berkeley poll found that less than half of Americans think the country should push forward aggressively on AI, per Fast Company. And Anthropic has a model that found 10,000 real vulnerabilities but won't release it because, in their own words, no one has built safeguards strong enough to prevent misuse, according to Engadget.

Here's the pattern: massive investment is flowing in, public patience is flowing out, and the companies closest to the frontier are voluntarily pumping the brakes. That's a tension that can't hold forever. The brands and businesses that survive this phase won't be the ones that moved fastest - they'll be the ones that could clearly articulate why they're using AI and what value it actually delivers. The era of "we added AI" as a selling point is over. Now you need receipts.

🏆 5-Minute AI Challenge

Prove AI Can Have Craft

The challenge: Take a photo of a handmade object you own - a mug, a quilt, a piece of furniture - and use AI to generate a story celebrating the human craft behind it, then create a matching piece of AI art inspired by it.

Step by step:

1. Grab your phone and snap a clear photo of any handmade or handcrafted item in your home.
2. Upload the photo to your favorite AI tool and ask it to identify the likely craft techniques used and write a 3-sentence story honoring the human skill behind the object.
3. Copy one vivid phrase from that story and paste it as an image generation prompt in an AI image tool (like Adobe Firefly, DALL-E, or Canva AI).
4. Compare the AI-generated image to your original photo - notice what details, texture, and imperfection the AI captured or missed.
5. Share both images side by side with the caption: 'Craft is alive - human AND AI.'

Why this matters: As today's headlines debate whether AI is killing craft and authenticity, this challenge lets you experience firsthand both the limits and surprising capabilities of AI when faced with real human-made artistry.

The Bottom Line

The Pattern: AI is splitting into two lanes - one where it's delivering genuinely powerful results (finding thousands of security bugs, accelerating drug discovery), and another where it's generating backlash, debt, and unrealistic expectations about human faces. The technology is real, but the gap between capability and trust is widening fast.

Why It Matters: If you run a business, this split affects you directly. Slapping AI on your marketing without substance is now a brand risk, not a brand boost. But ignoring the tools that actually work - like AI-powered security scanning or drug repurposing - means falling behind competitors who are quietly getting real results.

Your Move: This weekend, pick one process in your work that's slow, repetitive, or error-prone. Research one AI tool that addresses it specifically. Don't chase hype - chase the boring, useful stuff. That's where the value actually lives right now.

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