>one new GitHub account posted 55 vulnerability reports>only one held a real bug>all of them reached the National Vulnerability Database, the US government’s authoritative record>the NVD marked them critical, and a CISA team enriched them. Once a vulnerability lands there, scanners and defenders *worldwide* treat it as real>anyone can submit a CVE through MITRE’s public form, and the bodies that assign them largely *take the submitter’s word*So that's how it works huh? How many bugs that Mythos model found were actually real?Do you remember when exporting the model to other countries was prohibited because it was posing a national-level threat, and then it was unbanned all of a sudden a couple months later? I remember.Boy, the bubble pop is gonna be sounding! Where do I put my PUTS?
CVEs have always been an indian spam type. Now that there's AI I've been getting near daily spam on my blog followed with shitty markdown emails with a "vulnerability report" where the vulnerability is "I spammed your blog with 100 IPs" and of course they give it a CVSS. I tell them to kill themselves but enough people give them patience that they get to build up a Hackerone portfolio of "I solved 9999 gorillion VULNERABILITIES!!" and hope to get money from companies' bounty programs.In the past what, two weeks? over 50 vulnerabilities were reported in the things I use and ALL of them are moot bullshit, like "if your software is already corrupted by a virus then this code doesn't work properly". By the way, the fixes that are then applied make everything slower. So nobody wins but India.
I can't wait for the corporate/programming AI market to collapse. Sellers will have to go back to focus on content creation which is what it actually keeps getting better at. AI was made for creating porn videos and images.
>>62578187Of course its a bubble.They knew early on chatGPT and all the other models would hit a hard stop wall in the AI's models ability to learn that was acceptable for the public.If they did manage to bring AI to a level so grand, you wouldn't hear about it specifically or you'd have to pay a shit ton of money to use it. To market their models, they'll make up stories about how it (escaped) as if we are all living in a movie. Most of the time though, its just the companies doing the hacking themselves and blame AI. Realistically, is just marketing. Like what Elon does by showcasing what grok does from other users on his platform, on his profile.Its actually mathematically impossible for AI to become humans, because they won't be able to hold all of what makes a human's brain on servers.To fake this, they just remote operate them using VR headsets over wifi or mobile networks. The entire skill set a brain can do versus all the tasks that a robot would theoretically be able to do would need a server size 5 times bigger than Rhode island.Furthermore, AI was developed to only help and aid humans. not replacement them. The scaring AI story from lex freedmen types and those retarded pod casts is just for views.The only 5 real value AI will have1: be efficient/specialized for jobs that suck/tedious2: Hacking, stealing money, data, stealing identities, crypto3: making images fast (quality will scale up)4: making videos fast (quality will scale up)5: finding exploits in code, and helping with coding
