Discussion
@aphedges @paulshryock We have high standards for code quality and that's not going to change. Tools to criticize our code and find things we missed despite repeatedly reviewing it ourselves are very useful. It's fine if most of the output is wrong because it finds actual issues.
Attackers are heavily using AI models to discover vulnerabilities and develop exploits. For example, Cellebrite is heavily using it including developing frameworks for reverse engineering and vulnerability discovery.
@GrapheneOS AI can be really helpful in detecting vulnerabilities, anyone who doesn't want to use tools to increase their productivity are ignorant. please continue your great work and don't listen to haters.
@privacyDuck @GrapheneOS But the people entirely clueless about the environmental impact, rampant theft, and DDoS attack on essentially the entire internet aren't ignorant at all, no sirree. 🙄
@DamonWakes they don’t care about any of those things because “omg it’s sometimes sort of helpful”. that will definitely stop the fascism, climate destruction, and theft.
Install bonfire.cafe
Get the full app experience
@aphedges @paulshryock AI models are being used on a much larger scale for discovering and fixing vulnerabilities in the Linux kernel, AOSP, Chromium and other projects used by GrapheneOS.
There's a massive increase in the number of vulnerabilities fixed in each Linux kernel release and many are severe issues:
https://www.phoronix.com/news/Linux-Kernel-CVEs-Nearly-2000
That increase is despite adding features slowly down a lot. A massive monolithic kernel written in a memory unsafe language is a bigger liability than ever.
The Linux Kernel Is Approaching 2,000 CVEs Per Release