Adrian —dangerously-skip-permissions Sanabria

2K Followers
293 Following
2.6K Posts

🎙️ Enterprise Weekly Podcast
🤝 Founder @bsidesknoxville
🗣️ Faculty @IANS_Security
🕵️ Security Research
🍳 Cooking
⛰️ Hiking
🏎️ F1

"I rant with data!"

🤝 BSides Knoxvillehttps://bsidesknoxville.com
🎙️ Enterprise Security Weeklyhttps://securityweekly.com/esw
📚️ Cybersecurity Canonhttps://cybercanon.org
💣️ Destroyed by Breach Datasethttps://docs.google.com/spreadsheets/d/15CTPcgZQenWKDLDTQ2ibveUM4i7Of_n20TzdTi23xcg/edit?usp=sharing
Will MSFT at least give Github Copilot users free nights and weekends?

For anyone considering the Macbook Neo for work stuffs, I've been using it for several months now for mostly light office stuff, podcast recordings, having a million tabs open, Canva, email, a bit of Claude Code - typical stuff.

No problems. I've definitely hit its limit with a stupid amount of apps open (three separate browsers at once + a bunch of other stuff), but using it is generally indistinguishable from the Macbook Air or Pro, except that it doesn't burn your legs or make fan noise.

Battery doesn't last as long as the current Air models, but it's also half the price. I've got the TouchID model with 512GB storage because I'm not a total lunatic.

Have you vibe coded an app and then run it through an AI security review and found it had tons of issues? Or have you actually found a security issue on your own? I would love to talk to you for a story I'm writing!

In my latest post, I ponder whether Project Glasswing is just a private freemium tier, designed to hook the world’s largest software makers into fueling an endless token bonfire.

https://open.substack.com/pub/defendersinitiative/p/the-unintended-consequences-of-vulnmaxxing?r=74yjk

I knew the situation with AI companies was precarious, but I didn't realize it was THIS precarious.

https://youtu.be/zbKDmkJPVvI?si=M0noeEuewoHtGCWG

Anthopic, OpenAI Should Not Be Allowed to IPO, Says Ed Zitron

YouTube

Replaced my avatar, tired of looking at my smiling face

don't judge me

One observation - when AI hallucinates on trip time estimates, it's ALWAYS shorter.

Got me all excited one train ride would take 3.5 hours that actually takes 10.5 hours and that a car ride would take 5 hours when it takes 8.5.

Trying to work out some tricky travel logistics and every minute I attempt to use AI to help is a minute wasted

Travel estimates are all hallucinated and wrong

Half a dozen options it doesn't consider or offer

Even when it searches the web, it misses the most important and obvious options

Every now and then, it offers some advice I hadn't considered, but that's about 10% of the time

The Internet has some truly amazing travel planning resources - the data behind Google Flights and Google Maps is mind-boggling - maybe Google's AI would be doing a better job here?

But then, no way am I going to have subscriptions to 10 different AI chatbots because each has some useful specialty

I have a new game i like to call

"can I do it faster than the AI"

the answer is usually YES

I'm so tired of waiting for prompts

Coffee shops are now basically Coke Freestyle and I don't like it

I order a latte and every single time:

"you don't want any flavors?"

guuuuurl, coffee IS the flavor!

#GetOffMyGrounds #FightMe