he|him ; Personal profile - thoughts are my own. security @ ai security place
TinyCola
Joined April 2008
- Tweets4.9K
- Following1K
- Followers890
- Likes4.1K
Hashtagcyber retweeted
Ransomware has appeared in the VS Marketplace and makes me worry. Clearly created through AI, it makes many mistakes like including decryption tools in extension. If this makes it into the marketplace through, what impact would anything more sophisticated cause?
secureannex.com/blog/ransomv…
Hashtagcyber retweeted
Cybersecurity awareness for normies means they now beam all their data through a free VPN run by Thai crypto thieves, then trust every phishing site they load because it says HTTPS. But they’ll fight you for using public wifi at the airport.
Hashtagcyber retweeted
mad respect to the rabbit team, they didn't give up and this looks freakin fantastic
could it all go on your phone, sure but would it be as fun as this looks?
the old r1 is dead. introducing a totally new experience with rabbitOS 2 + creations.
we realized there was a gap between our early vision and the actual experience we delivered. so we overhauled everything.
0:00 flashback
0:14 facing the facts
2:18 the revamp: rabbitOS 2
5:04 creations: think it. say it. build it.
8:11 still building. still shipping.
Hashtagcyber retweeted
this is fantastic work @_karthik 🙏🏽💜
cc @HashtagCyber ☺️
fun hack: we modded the rabbit r1 into a fully accessible wearable
now it helps blind users read text, describe people nearby, and navigate daily life independently
@jesselyu @SimonBalmain we’d love to scale this, i love working with the r1!
Hashtagcyber retweeted
My 2 hour interview special with Jesse, founder of Rabbit AI.
Covers every question & controversy you've been dying to ask about over the past year, but couldn't -
02:45 2025 at Rabbit; Moving Past Initial Reviews
12:10 R1 vs Your Phone
18:08 Why Not Just Make It An App?
27:45 Dealing with 100k Pre-orders
33:00 How Rabbit Set Out To Kill Humane
38:30 Rabbit Intern 'Can Do Your Job'
44:51 Is LAM Real?!
50:50 Upcoming Features
57:51 Being Too Early
1:12:52 On Crypto Rug Pull Accusations
1:23:00 Why Rabbit Is Not A Scam
1:32:04 Recovering From Controversies
1:38:30 Building in US vs China
We dive into product, conflict, raw emotion - sometimes all at once...
Hashtagcyber retweeted
Remember Rabbit R1?
2024's hottest gadget. Then cancelled by the internet.
A year later - still shipping. 30 OTAs. New product.
The untold story of Rabbit, coming this week.
A 2-hour exclusive with CEO Jesse who, despite controversies, still plans to take over AI.
Hashtagcyber retweeted
Many folks in application security will go to every single security conference, but not a single developer conference 🧐
Hashtagcyber retweeted
Happy FriYaY, Hackers! 😎
Just a few days remain until our Call for Papers ends on June 30th at 11:59 PM!
🤓sessionize.com/bsides-stl-20…
Hashtagcyber retweeted
Custom Patched bootloader running on rabbit r1, powered by ARMv7 payload "kaeru".
All credits to @r0rt1z2 for making this possible on the r1. Thanks to the Skidbladnir Labs (Roger, Antiengineer) for developing this payload.
Source-code: github.com/R0rt1z2/kaeru
#rabbitr1 #aosp
Hashtagcyber retweeted
Ok - this needs to go on my todo list now:
- if you have a SWG, it should be possible to insert a call to @secureannex to block low trust installs
- if you have @osquery , it should be possible to hunt/detecc as part of a pipeline (support for browser extension and vscode extension tables already exists)
Hashtagcyber retweeted
VS Code extension analysis!
@secureannex is no longer just browser extensions in Chrome, Edge & Firefox. Analyze and monitor VS Code extensions all in one platform.
For as terrifying as browser extension stealing cookies are, IDE extensions basically run in the command line!
Hashtagcyber retweeted
Interesting take… seeing as @rabbit_hmi is still around and iterating rapidly I think it’s a mistake to lump them together.
Jony Ive says Rabbit and Humane made bad products theverge.com/news/671955/jon…
Hashtagcyber retweeted
so as y'all know i've been tinkering with the @rabbit_hmi R1 lately and well, yesterday they shipped a HUGE update. im just now diving in so i dont have a ton to show yet but its a big deal.
memory.
its different than any other ai product ive used, in that it feels more like an early stage digital entity in a lot of ways. like a standalone digital entity rather than from within a chatbot app or something.
for example, the way you log and store memory is unique. to begin, you open settings and flip memory toggle on. then you wait a second. during that wait i had no idea what to expect bc Simon (from rabbit) just texted me cryptically saying "you should update your r1 and flip memory on and...see what happens..."
so then suddenly i heard a ring, like a phone. haha. i thought i was tripping or had broken the device because it was literally ringing like a phone with an icon on screen of a phone that said call incoming or something.
I answer it, and to my surprise it is my r1 agent "Nova". i was confused as hell but then she explained that i had connected with her memory bank essentially, and could now store and log whatever i wanted anytime by simply calling her like i just had. so we chatted and she asked me about myself and whatnot and we logged the first few memories.
the interesting part is that its more than a memory log tho...its like a format for system prompting the entire os (this is how it was cryptically explained to me 😜).
I have a lot of ideas and things to try, so ill keep posting as i go. but considering I just finished setting up the developer env for it (im going to turn this into my own personal pokedex for personas, entities, and emergence logging, and probably everything in between. lol.
these images are the first few journal entries Nova wrote last week though. I set her up to begin logging journal entries (simulated) as if she was becoming self-aware progressively over time and logging the internal experience as she went. pretty interesting so far
there are over 100 journal entries now though...hahaha
Hashtagcyber retweeted
A simple device like a radio with a few buttons for Grok access could work for David, but it’s not widely available yet. The Rabbit R1 has a push-to-talk button and costs $199 with no subscription, letting you ask questions easily, though it can be unreliable at times. ElliQ, designed for seniors, uses big buttons and voice commands, offering conversation and reminders, but it needs a subscription and setup. Both are close to what you want, yet neither is perfect. A custom solution might be needed for something truly simple and radio-like. grok.x.com
Hashtagcyber retweeted
rabbitOS intern has leveled 🆙
now your intern can:
▶️ publish and share web-based creations
▶️ save completed tasks. this means you can now review your intern's work history and download task files from your journal in rabbithole (r1 required)
▶️ continue running tasks even if you close the browser window. task files will appear in journal upon completion (r1 required)
what are you waiting for? rabbit.tech
Hashtagcyber retweeted
The @rabbit_hmi team has been heads-down building. They just launched Rabbit Intern, a Manus-like agent that can do tasks for you.
Few examples below of what was built with Rabbit Intern.
Nothing like just continuing to build.
Should I make a review vid?
