Post ASgEISGuVIsLcNnPoO by alex@cybervillains.com
 (DIR) More posts by alex@cybervillains.com
 (DIR) Post #ASgDTx8TzakUZYlyU4 by alex@cybervillains.com
       2023-02-14T19:30:05Z
       
       0 likes, 0 repeats
       
       The lack of any consensus on AI safety, prompt injection attacks and other adversarial manipulations make me feel like it’s 1998 again. Huge, open field for researchers and entrepreneurs, except the technology is already deployed into the lives of normies.
       
 (DIR) Post #ASgDlPZrV5YThnu7NY by invisv@ioc.exchange
       2023-02-14T19:33:10Z
       
       0 likes, 0 repeats
       
       @alex Chapman's new writing on it is well worth a read: https://betterwithout.ai/
       
 (DIR) Post #ASgDvNM4cNjLi0Cpge by alex@cybervillains.com
       2023-02-14T19:35:02Z
       
       0 likes, 0 repeats
       
       For those of you who weren't working/alive the last time we went through this, the 90's and early 2000's were full of smart individual researchers writing papers or sending mailing list posts that would reveal and whole new class of issues that affected huge amounts of shipping code, forcing the handful of companies with prodsec teams to scramble and the rest to live in ignorance.A great example of the genre: https://seclists.org/bugtraq/2000/Sep/214
       
 (DIR) Post #ASgEISGuVIsLcNnPoO by alex@cybervillains.com
       2023-02-14T19:39:19Z
       
       0 likes, 0 repeats
       
       That was the era when you could go to Black Hat, watch a talk and then spend a year just finding examples of that bug in every product you looked at. Jeremiah Grossman's CSRF talk had that kind of impact: https://www.blackhat.com/presentations/bh-usa-06/BH-US-06-Grossman.pdfThis is where we are on AI, except the stakes are already high and the financial incentives are much juicier for keeping your research private.
       
 (DIR) Post #ASgEkp9z01V8iH5Diy by Di4na@hachyderm.io
       2023-02-14T19:44:28Z
       
       0 likes, 0 repeats
       
       @alex tbf format string attacks are still a target-rich technique today...
       
 (DIR) Post #ASgFjXTHNdRJW73mNM by mikebabcock@floss.social
       2023-02-14T19:55:21Z
       
       0 likes, 0 repeats
       
       @alex oh the memories ... sigh. Thanks for that reminder lol.
       
 (DIR) Post #ASgGcTFYmCFcLspurY by drwho@hackers.town
       2023-02-14T20:05:16Z
       
       0 likes, 0 repeats
       
       @alex That was the era when you could actually get into stuff at Black Hat and Defcon.
       
 (DIR) Post #ASgGfWmk5a5uBWoaK8 by drwho@hackers.town
       2023-02-14T20:05:48Z
       
       0 likes, 0 repeats
       
       @alex I really, really, really miss those days.
       
 (DIR) Post #ASgGi2JNqTIoXeHLDk by alex@cybervillains.com
       2023-02-14T20:06:20Z
       
       0 likes, 0 repeats
       
       @drwho Although, in the case of the latter, you might have had to sit right next to the AC unit in the tent outside the Alexis Park to not die of heat stroke.
       
 (DIR) Post #ASgJ6EqN6R2SGeNQFk by rdp@notpickard.com
       2023-02-14T20:33:00Z
       
       0 likes, 0 repeats
       
       @alex the "here is a simple, generic, setuid wrapper that people could use...." response is also a great example of trying to admin a "fix"https://www.youtube.com/watch?v=dAagFKdsSNs
       
 (DIR) Post #ASgNrLjxg8fV5H6A1Q by drwho@hackers.town
       2023-02-14T21:26:19Z
       
       0 likes, 0 repeats
       
       @alex That was something I did frequently.