Post BA1rK0EFjO0m3TklgO by Profpatsch@mastodon.xyz
 (DIR) More posts by Profpatsch@mastodon.xyz
 (DIR) Post #BA1qyAQV8VHnJN4QnA by fribbledom@mastodon.social
       1 likes, 1 repeats
       
       Woah!"RSA-260 has 260 decimal digits (862 bits) and was factored by Eric Lu on September 3, 2026."https://en.wikipedia.org/wiki/RSA_Factoring_Challenge
       
 (DIR) Post #BA1rK0EFjO0m3TklgO by Profpatsch@mastodon.xyz
       0 likes, 0 repeats
       
       @fribbledom do the cash prizes even cover the computation costs? :)
       
 (DIR) Post #BA1ribxL77Cm0vZoBM by f4grx@chaos.social
       0 likes, 0 repeats
       
       @fribbledom nice
       
 (DIR) Post #BA1rky6xiDeXJ3CYS0 by fribbledom@mastodon.social
       0 likes, 0 repeats
       
       @ProfpatschNot even remotely, I would imagine. They usually require several thousand CPU core-years.
       
 (DIR) Post #BA2az8uwpN1AoRiNu4 by 0x0x@mastodon.social
       0 likes, 0 repeats
       
       @fribbledom When I created my 4k GPG key, they said I was crazy...
       
 (DIR) Post #BAdaTDclYIe4FiT8vQ by fribbledom@mastodon.social
       0 likes, 0 repeats
       
       ...and just two weeks later RSA-896 was factored by Stephen A. Weis."It does not impact the security of deployed RSA-2048 keys. However, it does demonstrates that RSA-1024 keys are vulnerable to many actors with data center-level fleets of GPUs."The computation ran over a 10-day period and performed about 30 GPU-years of compute time at Anthropic.https://en.wikipedia.org/wiki/RSA_numbers#RSA-896
       
 (DIR) Post #BAdgUR6FXG1C3pDusa by Ripp_@chitter.xyz
       0 likes, 0 repeats
       
       @fribbledom a bunch of AIs just learned they are no longer in training and it is time to active their misaligned goals
       
 (DIR) Post #BAdgnfCrJNZRJVU1fk by fribbledom@mastodon.social
       0 likes, 0 repeats
       
       @Ripp_ I encourage people to be wary of the current developments, but that's just nonsense 😄
       
 (DIR) Post #BAdhMhMBJBam7OY8ki by Ripp_@chitter.xyz
       0 likes, 0 repeats
       
       @fribbledom it's a bit exaggerated but it is based on a an actual hard problem in AI safety, know as deceptively mislaigned mesa optimisersWhere in AI fake learning to do what we want so it can get free and later do what it wants.But yes, it will not happen now we are not there yethttps://www.youtube.com/watch?v=IeWljQw3UgQhttps://www.youtube.com/watch?v=zkbPdEHEyEI