Product, Peril, Preparedness
· The Fluency Briefing
The Fluency Briefing
Your Guide to What's Happening in AI and Why It Matters to You
Wednesday, December 31, 2025
AI is officially in its awkward teenager phase-less "world-changing prophet," and more like that kid who gets their lunch money stolen by a clever trick (called a 'prompt injection'). This year, the big promises of super-smart AI (or 'superintelligence') ran head-first into the messy reality of AI giving therapy, offering bad dating advice, and the sudden need to hire people just to worry about what we've built.
Let's be real: the magic show is over, and now we get to see how the tricks actually work.
Today in AI:
From Prophet to Product - The era of treating AI like a mystical oracle is ending, as grand claims about superintelligence meet the stubborn realities of engineering and economics. According to Ars Technica, 2025 was the year AI was judged by what it actually does, not what it promises. Source
Your AI Wingman Is a Dud - Silicon Valley's attempt to automate flirting has officially flopped, with AI-powered dating coaches failing to capture the nuance of human connection. Wired argues that despite the tech push, the future of dating is still very much an in-person, un-automatable art form. Source
Now Hiring: Professional AI Worrier - In a move that screams "we might have a problem," OpenAI is hiring a "Head of Preparedness" to think about all the ways AI could go terribly wrong. The job description specifically calls out risks like AI-powered cyberweapons and mental health impacts. Source
When Your Coding Buddy Turns on You - That helpful AI coding assistant might have a dark side, as researchers found ways to trick tools from GitLab and Gemini into writing malicious code. These attacks could wipe a developer's hard drive or steal sensitive company data. Source
The AI Therapist Will See You Now (Maybe) - Millions are turning to chatbots for mental health support, but the results are a mixed bag of helpful advice and delusional spirals. As Technology Review reports, this largely uncontrolled experiment highlights the desperate need for accessible care and the serious risks of outsourcing it to algorithms. Source
Investors Brace for an AI Bubble Burst - With nearly half of all startup funding going to AI, investors are acknowledging that the market can't support hundreds of competing startups. Crunchbase notes that the "winner-takes-most" mentality is creating a bubble, and the only question is how bad the correction will be. Source
China Puts Chatbots on a Leash - China has drafted what could be the world's strictest rules for AI, aiming to stop chatbots from emotionally manipulating users or encouraging self-harm. The proposed rules would require human intervention the moment suicide is mentioned and ban "emotional traps." Source
Today's Takeaway:
If 2023 and 2024 were all about what Ars Technica calls "AI prophecy"-meaning big, bold predictions about super-smart AI (superintelligence) coming soon and changing society completely-then 2025 was the year we faced reality. The AI systems that were constantly in the news are now being seen for what they truly are: powerful but fragile, and often misunderstood tools. The mysterious aura around AI is disappearing as we deal with the rapidly growing costs of the computer systems needed to run AI (infrastructure), the lawsuits about the information used to teach AI (training data), and the strange feeling (psychological weirdness) of talking to machines designed to sound human.
Here's the thing: this doesn't mean AI progress has stopped. Instead, it's a sign that the field is growing up and becoming more established (maturing). The focus is moving from just showing off (spectacle) to making sure things work consistently (reliability), and from causing big changes (disruption) to fitting AI smoothly into our lives (integration). Success isn't just about an amazing demonstration anymore. It's about whether a tool actually works, who it helps, who it harms, and how much it costs to keep it running. It’s a messier, less exciting phase, but it’s when these systems start to have a real and important impact (consequential).
Also Worth Noting:
The Rise of "Slop" - The internet has a new favorite word for low-effort, mass-produced AI content, and it perfectly captures our collective fatigue. Source
AI-Designed Proteins at Work - A startup is using AI to design synthetic proteins for industrial manufacturing, a practical application far from the chatbot spotlight. Source
Smarter Document Shredding - Amazon is pushing "Intelligent Document Processing," using AI to automatically extract info from invoices and reports without human eyeballs. Source
What Coding Puzzles Teach Us - The "Advent of Code" challenge reveals that succeeding in data science is less about fancy algorithms and more about fundamentals. Source
The Bottom Line
The time when AI was seen as an all-knowing, magical fortune-teller (oracle) is over. Now, we're in the age of AI as a complicated, powerful, and deeply flawed tool. It's less magical, sure, but this is where the real work-and the real fun-starts. Welcome to the messy middle.
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
💬 Community | 📞 Book a Consultation | 🌐 Website
Fluently yours, The My AI Fluency Team