4 ms·

> Also, who is Brian Smith? Are we moving from a state that can know everything on us, but at least might be too busy, to a mob that insists on knowing everyth

> Also, who is Brian Smith? Are we moving from a state that can know everything on us, but at least might be too busy, to a mob that insists on knowing everything about us, because it's too idle? I'm pseudoanonymous on here, deleted my LinkedIn account years ago because I despise the site with an utter passion. And, fuck me, I've probably said something stupid about cryptology online because even though I love practical math, crypto is really hard, and IANAC. I could explain that I'm pseudoanonymous here because I actually am an attorney,[1, 2] and I don't want anything I accidentally say online to be associated with any of my clients. But while that sounds very serious, it's actually pretty unlikely, so no, there's another secret reason that's utterly selfish. I'm going to reveal it here for everyone. I don't use my true name all the time because I don't want my online reputation in every community tied to stupid shit I said online when I was 15. And when I'm 35,[1] I might not want to tie my online reputation to the stupid shit I'm saying right now. I like the freedom online communities provide to earn respect from nothing, without people relying on any preconceived notions based on how your name sounds, or even what you did last year. I take comfort in the realization that if it all doesn't work out, I can always walk away from a profile and start from scratch. It's social bankruptcy protection. Or maybe, the summer before your freshman year anytime you need it to be. I feel like in 2002 we wasted a lot of time trying to figure out who in the sphere of public thought was a secret terrorist. I don't think we got anything out of those conversations, and I'd hate to see us repeat the same mistake digging for secret bureaucrats. I know the recent news makes this hard, but maybe go back to "Is this guy Bruce Schneier, or someone I've never heard of? If I've never heard of him, I'll provide extra scrutiny to the proposal, not because I'm necessarily presuming malicious intent,[3] just because his reputation is not established, and it's a better use of my time to scrutinize the proposal than to scrutinize some random guy's life." You can waste a lot of cycles trying to confirm or deny if someone has a secret life, something that's probably unknowable, and none of that work actually improves the crypto. [1] Head start on the doxxing for anyone playing along at home. [2] I am not your attorney, so don't get any funny ideas. [3] Well, arguably in crypto, you should always assume malicious intent, constantly ask how any little bit could aid an adversary. But that's not all we're talking about, is it? EDIT: Formatting (mostly footnote numbering)

sillysaurus2 13y ago
Yes, my apologies. My comment was overly blunt. The reason paranoia is important in post-2013 crypto is because the NSA's knowledge of crypto is very likely ahead of academia / public sector knowledge. They've historically used that advanced knowledge to influence past standards. In one case, they seem to have secretly enhanced the security of DES.[1] But in another case, evidence suggests they may have secretly put a backdoor into their Dual_EC_DRBG proposal.[2] This raises the question: Should we worry about NSA tampering with proposals? If that's our goal, then it seems to me that our most powerful tool is trust: the fact that we can almost always trust tptacek, cperciva, the Mozilla Security team, and other established names. It's not that they're infallible --- rather, it's that they've always been genuine with their intentions, and have spent years building up that level of trust, so it would be a huge risk for them to be willing parties in NSA tampering. Therefore, if someone is proposing a new standard, scrutinizing their reputation would seem to be a necessary first step. If their knowledge is secretly ahead of the public sector's, and their proposal contains a secret flaw, then no one will be able to spot it. What else is there to examine if not our level of trust in their established reputation? So if we choose to believe that the NSA is ahead of public sector knowledge, then it seems like it's valid to be concerned about identity and reputation, because it's inherently impossible to rely on the public sector to spot any hidden influences in the proposal, unless the public sector happens to make the same sort of cryptographic breakthroughs that we presume the NSA to have made. I apologize for bringing it up in a disrespectful way. I didn't mean to dig into Brian's life. The LinkedIn profile just happened to stand out in a quick google search. For what it's worth, all I'm trying to do here is ask the community: should we be concerned about NSA tampering? If so, should we assume their knowledge is ahead of the public sector's? If we assume that, then what else can we do except scrutinize identity/trust (since if we assume advanced knowledge, then we can't rely solely on scrutinizing the content of their proposal)? I honestly don't know whether those are valid concerns going forward, or whether they should be taken so seriously. I'm hoping the community will decide. [1] http://en.wikipedia.org/wiki/Data_Encryption_Standard#NSA.27s_involvement_in_the_design http://en.wikipedia.org/wiki/Data_Encryption_Standard#NSA.27... [2] http://en.wikipedia.org/wiki/Dual_EC_DRBG http://en.wikipedia.org/wiki/Dual_EC_DRBG
- brownbat 13y ago

sillysaurus2 13y ago

Yes, my apologies. My comment was overly blunt. The reason paranoia is important in post-2013 crypto is because the NSA's knowledge of crypto is very likely ahead of academia / public sector knowledge. They've historically used that advanced knowledge to influence past standards. In one case, they seem to have secretly enhanced the security of DES.[1] But in another case, evidence suggests they may have secretly put a backdoor into their Dual_EC_DRBG proposal.[2] This raises the question: Should we worry about NSA tampering with proposals? If that's our goal, then it seems to me that our most powerful tool is trust: the fact that we can almost always trust tptacek, cperciva, the Mozilla Security team, and other established names. It's not that they're infallible --- rather, it's that they've always been genuine with their intentions, and have spent years building up that level of trust, so it would be a huge risk for them to be willing parties in NSA tampering. Therefore, if someone is proposing a new standard, scrutinizing their reputation would seem to be a necessary first step. If their knowledge is secretly ahead of the public sector's, and their proposal contains a secret flaw, then no one will be able to spot it. What else is there to examine if not our level of trust in their established reputation? So if we choose to believe that the NSA is ahead of public sector knowledge, then it seems like it's valid to be concerned about identity and reputation, because it's inherently impossible to rely on the public sector to spot any hidden influences in the proposal, unless the public sector happens to make the same sort of cryptographic breakthroughs that we presume the NSA to have made. I apologize for bringing it up in a disrespectful way. I didn't mean to dig into Brian's life. The LinkedIn profile just happened to stand out in a quick google search. For what it's worth, all I'm trying to do here is ask the community: should we be concerned about NSA tampering? If so, should we assume their knowledge is ahead of the public sector's? If we assume that, then what else can we do except scrutinize identity/trust (since if we assume advanced knowledge, then we can't rely solely on scrutinizing the content of their proposal)? I honestly don't know whether those are valid concerns going forward, or whether they should be taken so seriously. I'm hoping the community will decide. [1] http://en.wikipedia.org/wiki/Data_Encryption_Standard#NSA.27s_involvement_in_the_design http://en.wikipedia.org/wiki/Data_Encryption_Standard#NSA.27... [2] http://en.wikipedia.org/wiki/Dual_EC_DRBG http://en.wikipedia.org/wiki/Dual_EC_DRBG

brownbat 13y ago