- AI Fire
- Posts
- 😬 OAI’s $5,000-Night Trip Turns Ugly
Sponsored by
Plus: 4 Personal AI Agents that Actually Save You Time at Work (Automate 99% of Your Life). You'll see how simple it is to create those systems.
OpenAI’s first luxury influencer retreat sparked a HUGE backlash. Creators were accused of selling out over AI’s environmental impact, even one removed her video.
What's on FIRE 🔥
IN PARTNERSHIP WITH CATCH AGENT
The AI Agent You Can Trust
The best assistants don't multitask their attention across a hundred tools. Neither does Catch. It's an AI agent that focuses on one thing — the admin work you'd rather not touch — and does it exceptionally well.
Scheduling, flights, restaurants, follow-ups, vendors, clients. You hand it over; Catch handles the back-and-forth and comes back with it done.
No context-switching. No dropped balls. Just your admin, quietly cleared — so your focus stays on the work only you can do.
Meet the agent built for admin, and it'll be ready to work before your next meeting.
Get started at catchagent.ai — and give your attention back to what matters.
AI INSIGHTS
Mexico’s top university, UNAM, is facing a national admissions scandal after suspiciously high scores appeared in its first fully online entrance exam, involving AI and leaked questions. It disqualified more than 3,000 online entrance exams.
More than 158,000 applicants took the online exam.
The share scoring above 100 correct answers jumped from roughly 3.5% historically to 16.3%.
UNAM ordered affected applicants to take another supervised exam.
The testing platform used browser restrictions, webcams, microphones, AI monitoring, and human proctors.
Students questioned whether the system could reliably distinguish cheating from legitimate high performance.
UNAM is now requiring a control exam for tens of thousands of applicants whose scores fell within the affected range. President Claudia Sheinbaum also called for the case to be investigated.
Key takeaway: Universities are using AI both to enable and detect cheating, but automated monitoring still cannot provide unquestionable proof that an individual student broke the rules.
PRESENTED BY HUBSPOT
Want to get the most out of ChatGPT?
ChatGPT is a superpower if you know how to use it correctly.
Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.
Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.
AI SOURCES FROM AI FIRE
1. FREE: Claude’s Most Annoying Habits? These 5 Free Plugins Fix Them Right Away! These 5 free plugins help fix common problems, improve Claude’s behavior, and give you a smoother way to get better results without paying for extra tools.
2. I Tested 3 Top Powerful AI Game Builders With the Same Prompt. Here’s the Winner. This AI Game Builder test compares Base44, Replit, and Lovable across a board game, a platformer, and a 3D shooter to name the strongest result overall.
3. Video: I Built A Mini SaaS With ChatGPT Sites In Under 10 Minutes. What if you could turn a free public API into a real working web app with only ChatGPT Sites? I’ll show you how to go from idea → MVP → deployed mini SaaS, and how to expand your first AI-built product without overcomplicating it.
CREATE 4 PERSONAL AI AGENTS
I know Claude Code. I know Codex. I’ve tested both Kimi K3 and OpenClaw.
These models are insanely powerful, but there’s still one thing they need from you: constant, like constant direction (or too hard for beginners).
In this post, I’ll show you 4 AI agents that can handle your repetitive work, manage daily tasks, and help you automate a huge part of your life.
AI Agent | What It Can Help You Do |
|---|---|
Coordination Agent | Organize emails, meetings, deadlines, and daily priorities |
Creativity Agent | Turn rough ideas and notes into articles, slides, proposals, or other content |
Clarity Agent | Research topics, review long documents, and explain difficult information |
Coaching Agent | Practice interviews, presentations, negotiations, and important conversations |
You’ll also see verified examples that show how people are using these ideas in practice, so you can decide which type of agent is worth exploring first.
TODAY IN AI
AI HIGHLIGHTS
🚨 I nearly believed some of those satellite images too (like these photos). Google Earth’s new AI feature was abused almost instantly and removed within 24 hours.
🎓 Google and Kaggle’s free 353,000-person vibe coding course with thousands of real AI agent projects, lessons inside. All lessons are still available. Don't miss it.
👤 TIME-style headshot is going viral. Try this prompt to turn your regular LinkedIn selfies/photos into polished editorial portraits inspired by Marco Grob’s iconic covers.
🖥️ Kimi just dropped its first Kimi Work tutorial (demo video). It now researches, structures, designs charts, and exports fully editable presentations from scratch.
🏖️ OpenAI’s first luxury influencer trip backfired badly. Some were accused of selling their souls for a $5,000-a-night stay, even one deleted her video. You need to see this.
🤖 After making $1.1 billion in quarterly profit, Palantir CEO Alex Karp calls frontier AI labs ‘Marxist’ for trying to control companies’ data and workflows. Read his warning.
⚡ DeepSeek V4 Flash just got 14 new Hugging Face quants from Atomic with standout quality. It is the best for testing on 128GB hardware. Full release here.
💰 Big AI Fundraising: Horizon3 raised $250 million at a $2 billion valuation as AI attacks rise. Its NodeZero platform has run 310,000 security tests with zero disruptions across 7,200+ customers.
NEW EMPOWERED AI TOOLS
📱 Replay QA lets you tests any web app from just a live URL, finds real bugs, records every full browser session, explains the root cause, and suggests fixes immediately. Test your app for Free.
☁️ AgentSky lets you launch cloud-hosted AI agents like Claude Code or Codex, then access them from chat apps, the web, or APIs.
🤖 Qwen3.8-Max is Alibaba’s most powerful AI model yet, built for coding, research, long-context reasoning, and complex tasks.
🥽 Ctruh Studio lets you build interactive 3D experiences without coding, using AI to create and customize immersive web content.
📈 Airtop for Google Ads lets you create, optimize, and monitor Google Ads campaigns through simple conversations for everyone.
AI BREAKTHROUGH
A new survey found that managers are already using AI to help decide who gets laid off, fired, or replaced. The survey covered 1,000 U.S. managers who use AI at work. Among them:
59% use AI when making layoff decisions.
58% use AI when deciding who to fire.
1 in 4 use AI for layoffs often or all the time.
17% frequently let AI make these decisions with little or no supervision.
38% have never received training on the ethical use of AI in HR.
Managers said they ask AI to weigh factors such as performance, productivity, attendance, salary, tenure, paid time off, sick leave, medical leave, and age.
The survey also found that 58% of managers could not confirm whether their company had tested its AI systems for bias, while 23% said no testing had taken place.
Another concerning finding: 34% of managers using AI for layoffs had asked whether an employee’s job could be performed by AI instead.
Key takeaway: AI is already influencing high-stakes employment decisions, even though many companies lack proper training, bias testing, and clear human oversight.
We read your emails, comments, and poll replies daily
How would you rate today’s newsletter?Your feedback helps us create the best newsletter possible |
Hit reply and say Hello – we'd love to hear from you!
Like what you're reading? Forward it to friends, and they can sign up here.
Cheers,
The AI Fire Team
Reply