AI in a Nutshell - Week 37 - AI Becomes Minister, Claude Muscles Up, and Job Losses Increase
Albania hires an AI minister, OpenAI teases a LinkedIn rival, Anthropic beefs up Claude, and Amazon rolls robotaxis down the Vegas Strip.
Fellow human, it’s week 37. AI hasn’t taken over the world yet. And here’s your lazy-man version of what went down in the AI world in the past week.
Speed-read: Albania hires an AI minister, OpenAI teases a LinkedIn rival, Anthropic beefs up Claude, and Amazon rolls robotaxis down the Vegas Strip.
What You Must Know
Albania Appoints an AI Government Minister → Albania made history by swearing in Diella, an AI-powered “minister” handling procurement by voice command. On one hand, it could fight corruption and cut bureaucracy. On the other hand, a bot now sits in government. What could go wrong? :)
AI Job Decline Study: 13% Fewer Junior Roles → Stanford found entry-level roles are vanishing, 13% fewer junior positions since AI tools gained serious adoption.
Mind-Reading Wearable Emerges → A new AI-powered headset reads your brain signals to control devices with your thoughts. Accessibility win, privacy nightmare. Just wondering what’ll happen to intrusive thoughts.
Claude Adds Team Memory Features → Anthropic’s Claude now remembers chat context across tasks for Team and Enterprise users.
Claude Can Now Create and Edit Files → Anthropic didn’t stop at memory. Claude can now create and edit spreadsheets, PDFs, and more, right inside chats.
OpenAI Jobs Platform to Rival LinkedIn → OpenAI announced a 2026 launch for an AI-powered jobs platform. Word on the street is that it’s like LinkedIn with AI-driven certifications and smarter matching.
OpenAI to Build Its Own AI Chips → OpenAI is teaming up with Broadcom to build custom chips, cutting dependence on Nvidia. If it works, it will lead to cheaper computing and faster products. Maybe subscriptions will even be cheaper, or maybe not :)
What’s Good to Know
Amazon Zoox Launches Robotaxis in Las Vegas → Zoox finally hit the Vegas Strip. The cars are built from scratch, passengers sit facing each other (awkward eye contact included), and rides are free, for now.
Apple Updates Siri With World Knowledge Answers → Apple finally decided Siri should know stuff. World-knowledge answers are rolling out, meaning it might stop being the dumbest assistant in the room.
FTC Launches Inquiry into AI Chatbots → The FTC is probing Google, Meta, OpenAI, and friends about how their bots handle user input, generate responses, and maybe cause harm. Basically: “hey, we saw your ToS, but do you really do that?”
Sam Altman Warns: Internet is Full of “AI Slop” → Altman says bots are drowning social media in low-quality AI content, which he calls “AI slop”.
U.S. Hosts White House Tech Summit for AI Policy → Trump gathered tech CEOs to talk about AI dominance, global policy, and spending. The U.S. wants to run the table on AI geopolitics. Big stage, big promises.
Hugging Face Brings Open LLMs to VS Code Copilot Chat → VS Code users can now tap open-source LLMs directly inside Copilot Chat. More transparency, less lock-in.
AI Tools Worth Knowing
Fin.ai → AI chatbot for Customer Support.
Auth0 → Authentication built for the AI era.
Reddtrends → Find business opportunities and gaps from Reddit conversations.
LiteLLM → Gateway to 100+ LLMs in one unified format.
Doraverse → AI coworker for office work.
Kaizen Corner - Why does “AI” forget things mid-conversation?
Chatbots forget things mid-conversation because they have limited memory, like a small chalkboard.
As you keep talking, new info pushes out older info to make space. So if you told it something early on and then talked about 100 other things, that old info might get erased.
You would think that the AI is dumb, but it’s just that its “brain space” is small and gets full fast.
Extra note: It’s worse when using the free version of any LLM.
Meme of the Week
Just Curious…
Would you trust an AI government minister to run parts of your country?
A) Yes, humans already screw it up.
B) No, I’ve seen Black Mirror.
C) Maybe, depends on my mood.
That’s your week in AI.
If you learned something, tell a friend. And if you didn’t, blame yourself.
Until next Sunday,
Kay - your fellow human
P.S. If this email lands in spam, that’s just your inbox trying to stop you from staying plugged in. Fix it.