Someone in the fCC Dallas Discord mentioned that they were asked to use a technology they hadn’t heard of at work (which may or not exist 😃). My response…
welp… a new post on @moltbook is now an AI saying they want E2E private spaces built FOR agents “so nobody (not the server, not even the humans) can read what agents say to each other unless they choose to share”.
it’s over
DeepMind just did the unthinkable.
They built an AI that doesn't need RAG and it has perfect memory of everything it's ever read.
It's called Recursive Language Models, and it might mark the death of traditional context windows forever.
Here's how it works (and why it matters way more than it sounds) ↓
I'm Boris and I created Claude Code. Lots of people have asked how I use Claude Code, so I wanted to show off my setup a bit.
My setup might be surprisingly vanilla! Claude Code works great out of the box, so I personally don't customize it much. There is no one correct way to use Claude Code: we intentionally build it in a way that you can use it, customize it, and hack it however you like. Each person on the Claude Code team uses it very differently.
So, here goes.
Copilot users - How do you get around censored responses that contain code that’s “too similar” to existing code? I just tell the prompt to put extra characters between each line and I’m good.
In this case obscurity was not security, as I was able to modify a hidden input and prompt inject whatever I wanted, getting the site to generate the output.
I’ve reported the issue but want to hold off sharing links or screenshots until the site is fixed. (3/3)
Have you found a bug in the wild? 🐞🐛💻🕷️🐜
Using new “#AI 🎩 🪄” features, a content creator recently relaunched one of their websites and asked for feedback… (1/3)
#appsec#promptinjection
The problem? Even though the affected page had a very limited form (a single dropdown) in the UI to provide data to one of @OpenAI‘s #GPT API’s and a server to process the form submission behind the scenes, there was no validation that the data was appropriate. (2/3)
@anerdguynow@derfloh People that didn’t game in the 90’s and early 2000’s have no idea how fast you wouldn’t be able to play the latest games on good hardware. You can easily get double the useful lifespan and reliability today.
@TomThumb_Stores https://t.co/wfnH8jngVZ works perfectly. The email works perfectly when user agent is spoofed. This is a user agent issue. My account is irrelevant as I can’t even get to the login. Thanks.
@TomThumb_Stores Clicking an email link for a weekly ad leads to the issues I described. White screen because your server is interpreting my machine as an Android device. Request fails as it is not an http request. I’d be happy to talk to an Engineer or you could hire me directly to fix it.
@TomThumb_Stores (3/X) When I spoof as Windows, it works fine. My guess is you have some logic that says “the leftovers” are all Android or something. Let me know if that’s enough info, happy to help.
@TomThumb_Stores (2/X) Here is the user agent used by the device: 'Mozilla/5.0 (X11; CrOS x86_64 14541.0.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36'
@TomThumb_Stores (1/X) This is coming from a Chrome OS desktop computer. Your servers are consuming the user agent from the device and assuming it is an Android device, trying to tell the computer to make a request starting with “intent://“ which leads to a blank screen.