Mike Duvos writes:

> Break a code, go to jail.  Even a silly code, like XOR. 

This is probably true.  In the current political climate, anyone who
posts "turbo-factor" on the Internet, and destroys secure communications
worldwide, can probably expect the secret tribunal followed by lethal
injection, after being smeared in the press as a traitor. 

Remember, if you're not on Shrub's bandwagon, helping him beat his little
drum, you're with the "terrorists."

> The 00's will be the Golden Age of something else.  Superintelligent AI
> perhaps. 

Opposite ends of the complexity spectrum.  Superintelligent AI can break
strong crypto.  Strong crypto means superintelligent AI requires
intractable computation.

Perhaps the complexity landscape permits only a middle ground.  Not
particularly smart AI, and not particularly strong crypto.

>> Even Rivest, Shamir, and Adleman knew essentially no number theory. 

> ... cryptography is based on faith, much like tea-leaf reading.  

A .sigfile quality observation, I'm sure.

> We have absolutely no hard mathematical evidence that factoring is any
> harder than multiplying or taking square roots, ...

I've always found it irksome that we haven't managed to move beyond
combination of congruences/homomorphism-based factoring techniques.

There has to be a simpler technique for unraveling multiplication, which,
after all, is a very simple and straightforward manipulation of bits.

> It is likely our ability to generate algorithms by a direct "grep" of all
> formulas having a specific form, and perhaps in the near future, all
> formulas under a certain length, will uncover many simple but difficult to
> directly derive formulas that do useful things.

Automated mining of reality for awesome but simple equations whose
derivations are just a bit too messy for humans to manually perform will
probably play an increasingly important role in the future of mathematics.

Ramanujan, as I recall, produced a lot of stuff which proved to be
correct, but which seemed impossible to arrive at without knowing it in
the first place.

> "Delete PGP, Win a Free Turkey,"

Har.

> Yes, folks.  It's the End of the Golden Age of Crypto.

Well, I'm not quite ready to run out and close the patent office yet.

We still have quantum cryptography and one-time pads, which, if our
current understanding is correct, are intrinsically unbreakable.

If one-way functions turn out to have been a crack-induced hallucination, 
quantum cryptography can replace public key systems for secure key
exchange.  

Some crypto-notable, I forget who, proposed putting satellites in orbit
which transmitted high bandwidth random noise, which one would XOR with
ones data before sending it.  The recipient, also receiving the satellite
signal, would know the starting bit in the random garbage, and could
decrypt.  Since it would be impractical to record the output of the
satellite over any period of time, this would preclude messages being
later decrypted, no matter how much CPU was thrown at them, as the
information to decrypt them would no longer exist.

Techniques like this, with satellite-based quantum crypto key exchange
services, would permit us to retain a reliable national crypto
infrastructure, should complexity-based systems fall apart under increased
combinatorial scrutiny.

-- 
Eric Michael Cordian 0+
O:.T:.O:. Mathematical Munitions Division
"Do What Thou Wilt Shall Be The Whole Of The Law"

Reply via email to