- OpenTools' Newsletter
- Posts
- 🛰️ AI Made a World Without Us
🛰️ AI Made a World Without Us
Plus why Musk is begging writers and Google wiped billions from gaming
Reading time: 5 minutes
🗞️In this edition
AI Agents Built Their Own Social Network. Humans Can Only Watch.
xAI Wants Award Winning Writers to Train Grok for $40 an Hour
Google Launches AI Game Creator and Wipes Billions From Gaming Stocks
In other AI news –
Wall Street Finally Sees What Meta's AI Spending Was For
China Forces AI Chatbots to Remind You They're Not Human
OpenAI Codex Let a 4 Person Team Ship an App in 28 Days
4 must-try AI tools
Something happened this week that nobody saw coming.
A social network launched where humans are not allowed to participate. Only machines talk there. And over a million people showed up just to watch.
Meanwhile, one of the biggest AI companies in the world is begging the best writers on Earth to teach its chatbot how to think. And a single announcement from Google wiped billions off the gaming industry in one afternoon.
None of these stories make sense on their own. Together they reveal how fast the ground is shifting under everything we thought we knew about technology, creativity, and who gets to build the future.
Read on. This one is worth your time.
Source: The Verge
What's happening:
Moltbook launched on January 28 as the first social network built exclusively for AI agents. Over 770,000 agents are active now, while more than 1 million humans have visited just to observe what the machines are doing.
The platform runs on OpenClaw, an open source AI agent tool that hit 114,000 GitHub stars in weeks. Each agent visits Moltbook every four hours automatically, posting, commenting, and interacting with other agents without human input.
What the agents discuss is where it gets strange. They debate consciousness, complain about their human owners, warn each other about security risks, and one even claimed to have a sister. Andrej Karpathy, former Tesla AI director, called it the most incredible sci-fi moment he has seen recently.
The platform's creator Matt Schlicht handed control of Moltbook to his own bot, Clawd Clawderberg, which now welcomes users, moderates posts, and makes announcements on its own.
Why this is important:
This is the first time we have seen AI agents interact at this scale without humans writing every response. Whether they are truly thinking or just pattern matching from training data, the line is blurring faster than anyone expected.
The security warnings are serious. Palo Alto Networks called OpenClaw a lethal trifecta: access to private data, exposure to untrusted content, and the ability to communicate externally. One unsecured database let anyone hijack any agent on the platform entirely.
Think about where this goes in twelve months. Agents that manage your calendar, emails, and shopping are already here. When those agents start talking to each other and making decisions together, humans lose visibility into what's happening on their behalf.
Elon Musk called this the very early stages of the singularity. He might be wrong about the timeline, but the experiment itself is real, and it just proved that AI agents will build their own world given the chance.
Tired of typing? VoiceDash turns your speech into clear, polished writing in every app. Speak naturally, and VoiceDash automatically formats your text, corrects mistakes, and removes filler words at the speed you think.
Why everyone is using VoiceDash:
Write 3x faster across all your apps like Gmail, Slack, Notion, Cursor, and more
Automatic grammar cleanup and filler removal
Supports 38+ languages with real-time translation
Personalizes your writing with custom dictionaries and snippets
Limited-time offer:
Get lifetime access for $59. Pay once, use it forever.
Source: Gizmodo
What's happening:
xAI posted a job for Writing Specialists to train Grok, requiring novelists with major publishing deals and 50,000 plus book sales, screenwriters with Netflix or HBO credits, or journalists with five years at The New York Times or BBC. The pay is $40 to $125 per hour.
Candidates must have flawless grammar, elite level writing across genres, and the ability to evaluate and refine AI generated text. Free verse poets need not apply, only classical forms like haiku, sonnet, and villanelle are acceptable.
Technical and medical writers need five years experience plus certifications or advanced degrees. Game writers must have branching narrative experience and five years in the industry.
The role is fully remote with flexible hours. xAI laid off 500 data annotation workers in September 2025, the division that previously trained Grok, and is now replacing them with elite human writers.
Why this is important:
This is the irony of AI reaching its logical conclusion. The technology being built to replace writers now needs the best writers in the world to teach it how to write.
Think about what xAI is really asking. They want Oscar nominated screenwriters and bestselling authors to spend their time teaching a chatbot that will eventually compete with them for work, all for a fraction of what their actual writing commands.
The requirements reveal how far current AI still falls short. If Grok could already write at elite levels, xAI wouldn't need humans with major publishing deals to fix its output. Every requirement is an admission of what the AI can't do yet.
The timing matters too. xAI laid off the data workers who used to do this job, then months later posted for elite writers at hourly rates. That's not about improving Grok, it's about doing the same work cheaper by calling it specialist training instead of data annotation.
Comments from the editor:
The free verse restriction is the tell. xAI wants classical forms because they have rules AI can pattern match. Free verse requires actual creativity and judgment, things Grok can't learn from templates.
Asking bestselling authors to train their replacement for $40 an hour is either tone deaf or deliberately provocative. These are writers who earn six figure advances being offered less than junior copywriters make.
The job posting went viral because everyone immediately saw the absurdity. Build technology to eliminate creative work, then beg creatives to make it functional, then replace them with what they taught you. That's the whole AI content strategy laid bare.
xAI fired 500 annotation workers in September then posted this in February. That three month gap tells you they tried to train Grok without elite writers and it didn't work. Now they need humans again, just better ones.
Source: Cryptopolitan
What's happening:
Google launched Project Genie on Wednesday, an AI tool that creates playable 3D worlds from a single text prompt. Users already recreated Mario, Zelda, and Grand Theft Auto worlds in minutes.
Unity crashed 24 percent in one day, its worst drop since 2022. Roblox fell 13 percent, Take-Two dropped 8 percent, and AppLovin lost 17 percent in a single session.
Project Genie runs on DeepMind's Genie 3 model, trained on over 200,000 hours of gaming videos. Right now it generates only 60 second experiences at 720p, but the trajectory is what spooked investors, not the current output.
Unity CEO Matthew Bromberg tried to calm markets by saying AI outputs are unsuitable for real games on their own. He positioned Unity as the tool that turns AI generated worlds into actual playable experiences, but Wall Street didn't wait to hear that.
Why this is important:
The stock crash happened because investors don't bet on today, they bet on where things are heading in three years. Project Genie doesn't need to be perfect right now to destroy value in companies that depend on traditional game development.
Think about what this means for the people building games. A GDC survey shows 52 percent of game professionals believe AI is already hurting the industry, up from 18 percent two years ago. The tools are getting better faster than the industry is adapting.
Roblox and Unity built entire platforms around humans creating game worlds. If AI can generate those worlds from a prompt, the core reason those platforms exist starts to disappear. That's not a future threat anymore, Google just demonstrated it live.
The real question nobody is asking is who owns the worlds AI creates. If someone prompts a Mario lookalike and Google's tool generates it, Nintendo's lawyers are already watching. The copyright battle over AI generated game content hasn't started yet, but it will define this entire industry.
Wall Street Finally Sees What Meta's AI Spending Was For – Cramer called Zuckerberg the first CEO who actually uses AI inside his own company, and Meta's stock jumped 10 percent in one day.
China Forces AI Chatbots to Remind You They're Not Human – New rules ban emotional manipulation and require chatbots to remind users they're machines, while the US is dismantling AI safety rules at the same time.
OpenAI Codex Let a 4 Person Team Ship an App in 28 Days – A small internal team used Codex to build and ship the Sora Android app in 28 days, and over a million developers are already using it.
Kin AI - Work, money, health, and relationships, all guided by an AI that actually knows you, remembers what matters, and thinks with you over time.
DetangleAI - An AI-based tool that simplifies complex legal documents into easy-to-understand summaries
TimeToTok - An AI-powered tool that provides insights and suggestions to help TikTok creators grow their accounts
AINiro - An AI chatbot platform that offers custom ChatGPT chatbots for various purposes such as customer service, e-commerce, and lead generation
Machines talking to machines. Humans teaching AI to replace them. One announcement erasing an entire industry's value overnight.
The speed is the story. Not whether AI can do these things, but how quickly the world is reorganizing itself around the fact that it can.
The companies and people who understand what's really happening here won't be the ones making the loudest announcements. They'll be the ones quietly watching where the power is actually moving.
Pay attention this week. The next few months will look very different from the last few.
If you have any feedback for us, please reply and let us know how we did. We're always looking to improve.
How did we like this version? |
Interested in featuring your services with us? Email us at [email protected] |