not your Weights means eventually not the same Model btw they can > quantize it
> distill it
> hot-swap to a cheaper/weaker checkpoint
> make the model manipulative
> fine-tune it in ways that break safety or depth
> drop its IQ
> run experiments on you and/or your data
>
@theahmadosman
-
Model Manipulation Risks: Quantization to Safety Degradation
By
–
-
Supreme Court Called to Intervene in AI Crisis
By
–
somebody ping the supreme court to stop this mess
-
Dario Amodei Criticized as Master Manipulator on Social Media
By
–
i cringe everytime i see dario on the timeline the dude is a master manipulator
-
Employee Recognition Failure Drives Tech Talent Attrition
By
–
في تيم قديم كان معانا حد نزل للشركة فاتورة شهرية 30 ألف دولار. مديره بعتله 200 دولار kudos على براڤو، وعرفنا السنة إللي بعدها إن زيادته كانت 3٪. حاجة مستفزة جدًا. مشي بعدها بشهرين
-
Technology set to transform regions permanently
By
–
بيتحط على المنطقة كلها وهيتعلم علينا للأبد https://
x.com/AhmedZaky/stat
/AhmedZaky/status/1731705700546498899
… -
Pink Floyd’s Wall: AI Ethics and Social Control Critique
By
–
I don't need no arms around me
I don't need no drugs to calm me
I have seen the writing on the wall… Listen to Another Brick In the Wall (Part 3) by Roger Waters https://
open.qobuz.com/track/95359047 -
Calling Out Bullshit in Everything: The Big Short Returns
By
–
– It's time to call bulshit
= Bullshit on what?
– Every fucking thing! #TheBigShort -
Personal AI Simulation Using Advanced Fine-Tuning and Data Upload
By
–
I would like for AI to be so advanced that I can upload and fine-tune my data and ask for a simulation of me living in the 60's. Would be really cool.
-
Meta’s Threads Platform Criticized for Poor UX and Performance
By
–
Threads is half-baked and rushed for release. Cannot choose between timelines, cannot look up topics, the UI feels like a Udemy intro to React course UI, and it takes ages to load, refresh, and respond. I am surprised Meta allowed this to be released.