Replying to @HybridStaticAnimate@infosec.exchange
@HybridStaticAnimate @GrapheneOS
Bot blocked.
S.Fynn 
🔒 Manual approval|| VulnRes/EmbeddedDev (self-employed)
|| linux::gentoo | android::grapheneos | bsd::net
pfp: Photomode shot of my modded V (C2077 patch1.5)
hdr: 3 x usb sticks and an Orange Pi 5 pro
TERFs, AI/Tech Bros and NAZIs FUCK OFF
#fedi22 #cyberpunkred #cyberpunk #baltimorecity #antifa #anarchist 
#fuckice #noai #nobot #nobridge #noindex #vtm #ttrpg
- Fynn (they/them)
lesbian 
non-binary 
- Website (WIP)
- https://shortcircuit.bar verified
- Keyoxide
- https://keyoxide.org/DEAA79D2EEC35E6FFCAD3D5D6D2F3486CC92B2EC verified
- Location
- Baltimore City
Replying to @erikarn@mstdn.social
Yeah. You are a reply guy.
Replying to @GrapheneOS@grapheneos.social
Instead of wading through LLM model vomit which takes a significant amount of time, with an admittedly low signal to noise ratio, you could invest in experienced vulnerability researchers who don't use LLMs.
One could argue that the time discovering vulnerabilities with traditional or custom tools (some vuln researchers write their own tooling) could take ~same amount of time as sifting through mass produced LLM bugs to determine if those bugs are actually vulnerabilities or not.