Pinned Tweet
I reverse-engineered OpenClaw into a 24/7 AI employee.
Not a toy. Not an experiment. A product that generates children's books, learning materials, and digital products while you sleep.
Here's how I turned an open-source framework into a money-printing machine 🧵
I don't know why but it feels like SWE-2 Max is better than fusion. Fusion fails on the most simple of tasks @cognition keep cooking your model its really good!
Grok @bot should natively support Grok Imagine. Would be super cool for experimental ad poster design @ericzakariasson please implement.
With the current pricing Codex OpenAI is the new Anthropic and Grok is the new Codex 😂
Only the real ones will know what this means.
Wait a minute w̴̛̛̝̝̰̙̤̹̣̮̲̳͉̥̺̲̪̣̬̜͙̘̙̠͙͕̭̤͙̰͗͂͗̒̊̇̎̿̈́̌̇̒̂̂̈́̈̎̈͗̒͂͗̑̏̄̀̃̅̉͒̾̅̂̚̚̕͜͝͠͝ͅţ̶̼͍̩̜̞͖̺̘̝͔̘̪̯̦͇̳̥͉̬̲̯͖̰̗̮͇̏̎͂̅̐́͛̄̈́̈́͐͗̊͊̋̇̆͌͗͗͋̈̓̆͋̚̚̚͜͝͠f̶͙͔͇̚͝f̶̨̛̦͙̙̳̬̻͔̠͙̖̠͇͍̱͓̞̯͐̂͛̐́̚f̷̛͖͉̩͍̘̖̱̙̲̜̠͔̽̅͒̀̍̎̽͌́̅̇̓͋̎̽͆̈́͐̉̚͝͠͝͝ i so what is everyone doing to have that extension?
Just tried DSV4 flash on @opencode after subscribing first prompt and....😢
Wow SWE 1.7 is a beast @cognition cooked on this one, slightly more intelligent than GLM 5.2.
If this model had vision I think it would easily match Opus 4.6. or even exceed it, well done folks.
Hi @thsottiaux feature request, purchase of banked resets, cool right right? 😉
Devin Desktop keeps crashing when using GLM 5.2 @cognition @devindesktop @DevinAI
Without touching anything today in @windsurf i have hit the usage limit 😂😂😂😂. Lol
Replying to @devindevelopers @windsurf
Probably another unwanted feature like arena. Do you ever listen to users?
Made the mistake of installing Cursor glass and you can tell it was vibe coded, crashes more than cars.
Codex app is powerful but that inner sense of right and wrong embedded in it causes it to make decisions for you that you haven't explicitly asked it lol @OpenAI roles are changing the AI is prompting humans.
Tipslo retweeted
Want to see OpenClaw in action?
@Tips_Singer captured the whole demonstration:
xitter.cf/Tips_Singer/status/202…
This is what 24/7 AI production looks like.
#AIEmployee
Traditional vs AI Employee
Traditional:
→ 2-3 products/month
→ $2,000+ production costs
→ 40+ hours of work
AI Employee:
→ 48 products/month
→ $42 production costs
→ 8.5 hours of work
The difference? Infrastructure. attilahun.online
@attitithehul #AIEmployee
What vision @OfficialLoganK 🤌
Quick poll for creators 👇
If you had a system that could create a complete digital product (book, course, workbook) in 45 minutes...
How many could you launch per month?
🔄 5-10 (building a catalog)
❤️ 15-25 (dominating a niche)
💬 30+ (empire mode)
I'm building something that makes any of these possible.
xitter.cf/Tips_Singer/status/202…









