Zack Korman

AI "cyber models" aren't going to solve our security problems. "If we just patch all the vulnerabilities then we're safe" is a convenient story but it's also not true.

If you're worried about cybersecurity, the solution involves hard work, not just more tokens.

“Cybersecurity doesn’t work on AI because humans will just give AI access” is such a bad argument from doomers.

Cybersecurity isn’t the field of “never give access to anything”. If that were true, every employee onboarding would be a security failure.

AI 2027 is really just a story about AI progress, with this weird background conspiracy about the AI secretly plotting to kill us all mixed in.

They now say we're 5 months ahead of that timeline, so August 1st is now the last chance of human control. Apparently.

Observability has been so bad in cybersecurity for so long that it’s basically become “do you have the data for incident response”. That’s a problem.

Vendors often do more harm than threat actors but we show up to security conferences and talk about cool little hacks instead of the systemic issues plaguing this industry.