Well I saw the context bomb method, with prompts which block AI attacks... So I gave those to my qwen... It not only happily crunched them, didnt even blink, but he started solving them vigorously 🫠

😁

https://haven.downisontheup.ca/80fb00156de60e33d2eb8f63b25bfaed12d443c19dc703af6e7cafc0849478d2.jpg

https://blossom.primal.net/44de86a1fa952a401e628e4fa9e850c8c7e757612cb33cfedf261a42303116af.jpg

collectivism > individualism nostr:nevent1qvzqqqqqqypzqprpljlvcnpnw3pejvkkhrc3y6wvmd7vjuad0fg2ud3dky66gaxaqyt8wumn8ghj7un9d3shjtnyd968gmewwp6kytcprpmhxue69uhhyetvv9ujuerjv4sk66t5dqh8gme0qqsv07eevjd5lkfggs05zjk37ud7tmjyhp2ruftutu48azslw3s76pcu6cnlr

would be ironic if the next attack is a backdoor from the chinese llm

There is zero chance someone isn't leaking the findings from the team to various country's intelligence and selling it for corporate espionage

You are joking, right? This is a very common term in the field.