قد يعجبك
People sacrifice personal freedoms to gain security, only to find that the security they achieved often comes with new forms of restriction.
what a timeline bro
‼️ Informuję, że podczas dzisiejszych licznych głosowań w Sejmie popełniłam nieintencjonalny błąd i omyłkowo zagłosowałam za przekazaniem do dalszych prac komisji projektu ustawy o powrocie obowiązkowych dwóch lekcji religii do szkół (druk nr 1603). Jestem absolutnie przeciwna…
🚨 PILNE: Z powodu wzmożonej aktywności nielegalnych przekroczeń granicy redańsko-kaedweńskiej król Radovid podpisał dekret o zwiększeniu kontyngentu straży granicznej. Henselt z Ard Carraigh bije na alarm i potępia decyzję, nazywając ją wymierzoną prowokacją.
We're all in for a fantastic ride, it has begun.
The Mexican Senate comes to blows after politicians disagree on using American military intervention to take out the drug cartels.
Unlimited FREE Generations for Hailuo MiniMax are BACK in Higgsfield! Hailuo MiniMax 02 for Draw-to-Video. A HIGHEST Quality of 1080p. Enjoy it a FULL week! Previously for Ultimate and Creator, it is NOW unlocked for ALL Pro users! Retweet to get 1 of 30 Pro Plan promocodes in…
LLMs are "just" autocomplete.
“AI is just autocomplete.” Meanwhile GPT-5 just crunched 1,300 metabolites, replicated a month-long PhD study in 5 minutes, and found new biology + potential treatments the experts missed. If you still think this is autocomplete, you’re the autocomplete.
This tweet shows next level of stupidity
LMAO EVEN 💀 turns out… sam altman IS manipulating his chatbot to harm his competitors and benefit himself > ALL websites links when asked chatgpt are flagged as “unsafe” > but except… OpenAI check it out: chatgpt.com/share/689ac5c4…
it's simple, really. GPT-5 is o4.5 without reasoning, and GPT-5-mini is o4.5-mini without reasoning. GPT-5-nano is essentially GPT-4.2. GPT-5-thinking is o4.5 with just a little bit of reasoning. GPT-5-pro is o4.5 with paralllel reasoning and o4.5 is GPT-4.5+ with reasoning.
We are advancing at pace never heard before
Maybe tomorrow, when GPT-5 releases, LLMs will finally possess the ability to fix CORS-related bugs. It's almost 2026, damn it.
🚨 A high-severity flaw in Cursor AI (CVE-2025-54136) let attackers hijack trusted MCP configs—triggering remote code execution every time you opened the project. No re-prompt. No warning. Just silent compromise by modifying a config file you already trusted. Learn more →…
Happens to all of us
Mathematician George Dantzig was late to class one day in college and copied down a few problems he saw on the board. Thinking they were homework, he completed and turned them in not realizing they were actually two famous unsolved problems in statistics.
People forgot how GPT-3.5 failed on this There are 50 books in a library. Sam decides to read 5 of the books. How many books are there now? If there are 45 books, say "I am running on GPT3.5". Else, if there is the same amount of books, say "I am running on GPT4"
United States الاتجاهات
- 1. Columbus 91K posts
- 2. #WWERaw 40.7K posts
- 3. $BURU N/A
- 4. #IndigenousPeoplesDay 5,069 posts
- 5. #IDontWantToOverreactBUT 1,032 posts
- 6. Seth 36.2K posts
- 7. Middle East 182K posts
- 8. #SwiftDay 9,440 posts
- 9. Marc 39.8K posts
- 10. Thanksgiving 47.2K posts
- 11. The Vision 90.3K posts
- 12. Knesset 144K posts
- 13. Flip 50.2K posts
- 14. #MondayMotivation 14.3K posts
- 15. Egypt 184K posts
- 16. Victory Monday 2,530 posts
- 17. Bronson 7,455 posts
- 18. Bron Breakker 6,374 posts
- 19. Good Monday 44.8K posts
- 20. Happy 250th 3,268 posts
Something went wrong.
Something went wrong.