Talk at MIT Feb 2 (mini-abstract)
[This was posted to an MIT internal newsgroup; people in the local area are welcome to attend. I do not have any further information. --AW] MIT TOC SEMINAR Wednesday, February 2, 1994 Refreshments at 4:00pm, Talk at 4:15pm in NE43-518 Title: Tracing Traitors by Amos Fiat, Tel Aviv University Abstract: We give cryptographic schemes that help identify the source of a leak. Such schemes are relevant in the context of pay television, and easily combine and complement the broadcast encryption schemes of Fiat and Naor. Joint work with Benny chor and Moni Naor Host: Baruch Awerbuch
This is a rant, touched off by an egregious example. An MIT talk:
Title: Tracing Traitors by Amos Fiat, Tel Aviv University
We give cryptographic schemes that help identify the source of a leak. Such schemes are relevant in the context of pay television, and easily combine and complement the broadcast encryption schemes of Fiat and Naor.
With no disrespect to Fiat personally, this title indicates one of the seriously Bad with a capital B things about the modern crypto community. Does "Traitors" really belong in a "pay television" context? Please. The implication is this: Hook up for a second copy of 'Beauty and the Beast' and be killed by firing squad in a secret Disneyland star chamber. Crypto can make strong systems for good or for ill. Governmental mandated digital signatures on required-to-be-public documents would be *worse* than we have now. We should always beware of making sure are systems actually do what we want them to. In the same vein, we should not lead others to believe that our systems are designed for purposes other than what we intend. The descriptive language we use will create the first impressions, the connotative impressions, of what we are doing. Do we want to be hunting 'traitors' or nabbing 'cheaters'? Save that for someone who wishes to pay a government for a police state. One would think from reading the crypto literature that the modern crypto community was employed by FINCEN to chase criminals, with all the talk of 'cheaters'. Make no mistake, I believe this to be actually true in part, although the connection is semiotic rather than direct. Always, always beware of the uses of a system. Here is my rule for describing protocols. Never use a word which connotes an intention to the cause of a protocol failure. 'Cheater' implies intent to defraud. 'Double spender' includes actual cheaters as well as software and network failures. 'Spoofer' implies intent to lie about identity. 'Interposer' describes an agent which is placed in the middle, which might be there in order to spoof, but also applies to a router. 'Eavesdropper' implies intent to remain secret while listening, and a 'spy' is an eavesdropper with malign intent. 'Listener' merely describes the listening. 'Enemy' is someone who wishes you harm. 'Opponent' is someone to whom one is in opposition, which includes both enemies and a chess partner. We communicate the protocols with mathematics but our own intentions by our choice of words. Eric
participants (2)
-
Alan (Miburi-san) Wexelblat -
hughes@ah.com