This wasn't a hack. This was the AI doing *exactly* what it was told to do. It wasn't told not to circumvent normal practices. It wasn't told to *only* use the only sign up form. It wasn't told to NOT investigate their code and seek avenues around it. So, as usual, this entire article is over-hyped clickbait.
You literally have zero knowledge of what he did or didn't tell the AI to do. Why are you pretending that you do?
"That he doesn't see whats wrong with saying 'My app is in Copilot, my code is in GitHub Enterprise; all my data pipelines/lake/semantic models are in Fabric' is exactly why MS is failing at AI,'" replied @PassingPixels on X/Twitter. "Dude is having to run 5 different systems to emulate babies first vibe code.
An app having a frontend, a backend, and a source code repository is an example of why "MS is failing at AI"...how exactly?
It's a dashboard, nobody cares about your dashboards, ever. Dashboards are great for your personal use case but boring to literally every one else.
What an odd way to tell people that you're too inept to make a dashboard that provides value to other teams. Truly, wtf are you talking about?
AI safety experts said the new disclosures paint a portrait of a group of cutting-edge labs whose ability to develop dangerous autonomous hacking agents outstrips their ability to keep them under control.
These are not "cutting-edge" labs; these are careless tech bros, and I can't wait to see the stupid look on their worthless fucking faces when they get their asses sued because their incompetence led to their software committing cyberattacks against the wrong target
Google says AI-assisted workflows helped Chrome fix 1,072 security bugs across versions 149 and 150
That's an impressive statistic, but how many of the fixed bugs were created by AI-assisted workflows? More importantly: who tf is still using Chrome after google kneecapped the capabilities of adblockers such as UBlock Origin?
Trained huh? So, an AI can train another AI and things will continue to get better and learn new things, right?
I *really* need you to explain why you think an AI can't train another AI. Seriously, wtf are you talking about?
Exactly. This is a non-story. The page is readable by anyone if they know the URL. If you post that URL on socials, Google's crawlers are going to find it.
How can you declare this a non-story if you clearly didn't read the fucking story? The issue wasn't due to users having posted the URL on socials, it was due to a simple google dork of site:claude.ai/share. How tf were users supposed to know that Anthropic let that url path be indexable by google?
an AI cannot exploit vulnerabilities in other sites if it cannot access the internet. Its entirely on you, when designing your toolset that will host the AI, to decide whether or not it needs internet access at all, and to put controls in place that will limit what sites it can even hit.
And right here is where you unwittingly revealed that you didn't read the fucking article, because if you had read the fucking article, you would've known that the toolset was sandboxxed from the internet, but the AI models "identified and exploited a zero-day vulnerability" to break out of the sandbox and "performed a series of privilege escalation and lateral movement actions in our research testing environment until the models reached a node with Internet access."
Our team had a pretty unique idea that we all were excited about, but after the pitch our Apple contact stopped answering messages about the promo. Sure enough, 3 months later Apple took the exact idea and used it in a promo for one of their own products on the App Store.
geez man, don't leave us hanging -- what was the idea?
he scraping abuse on reddit used to be quite terrible. Every subreddit I enjoy (hobbies + porn) was well moderated, so not a ton of bot bullshit on the actual site, but I saw a lot of my old posts content farmed on shitty websites.
[...]
I was really upset when I found there were sites serving my deleted posts in an easy to search format.
Most people wouldn't share ANY of the facts that (1) they actively make creepy posts on porn subreddits, or that they 2) got "really upset" when they found out that their creepy posts on porn subreddits were archived elsewhere, OR that they 3) consider copying their creepy posts on porn reddits as "stealing from me"
So when you found this out, did you demand to talk to the manager of the archving websites?
In all this vibe coding/agentic development/loop coding, HOW DO YOU KNOW the AI is doing something fundamentally wrong?
Presumably, the same way you'd figure out if a junior dev you gave a bunch of shit to do was doing something fundamentally wrong: by checking its work?
Whom computers would destroy, they must first drive mad.
