Sophie Schmieg / is on Nostr

Leading cryptography (ISE Crypto) at Google.

Opinions my own.

Content usually badly explained mathematics

profession: cryptography engineer
hobby: Kerbal Space Program
hobby: Lego
hobby: Factory Sim Games

Last Notes

2024-09-09 15:56:10

When you think about it, you realize that DNA is the mother of all esoteric programming languages

2024-06-15 22:26:45

New blog post: Recovering public keys from signatures. It turns out public keys, are, in fact, public.
https://keymaterial.net/2024/06/15/reconstructing-public-keys-from-signatures/

2024-06-14 13:56:42

- reply

we barely moved the needle on our favorite atmospheric waste gas, CO2. It's still way below 1%, and those oxygenating bastards (who have since evolved to take home inside of plant cells) just consider it fertilizer, if we stopped cutting them down manually, they'd eat most of it up.

2024-06-01 07:59:25

When I talk to other software engineers, there seem to be two camps developing when it comes to LLM assisted coding. Some are over the moon happy about it, and say it improves their productivity by an order of magnitude. Others are fairly indifferent to slightly negative about it, saying it is more of a nuisance than actually helping, inventing APIs that sure would be nice if they existed etc.
Finding myself in the latter camp, I just can't deal with the unpredictability of the tools. Both in terms of reproducibility, but also in terms of continuousness. Some tasks it does well on, but it's near impossible to tell whether it will do the same for a similar, related task. To me, the one advantage and disadvantage computers have is that they do exactly as told, and a probabilistic tool breaks that and just feels buggy.
The other thing I can't deal with is that it seemingly doesn't lend itself to iterative, local approaches. Every query is a new thing, that may discard everything it had before, and you can't really hone in on the thing you want.
Add to that that it often suggests things that might be reasonable, but aren't the way I would have done it, and I like doing things the way I would have done it, so I'm going to replace the entire suggestion wholesale, and I end up with a tool that feels at best only marginally useful, if not detrimental to writing code.

2024-05-20 17:58:34

- reply

corollary: very very often, cryptography is the place where things go down to a single layer (If you have to hand out some data to a third party, they can do what they want with it, except the mathematically impossible).
Which is fine, math is powerful, but it's only as secure as your key management. Add as many layers as you can to your key management. And then add some more.

2024-05-04 22:14:08

The safe word is: GIF

2024-04-19 03:42:14

What you're hearing there is the collective sigh of relief of the entire cryptographic community.
https://eprint.iacr.org/2024/555
https://media.infosec.exchange/infosec.exchange/media_attachments/files/112/295/812/013/333/513/original/7ea1e28adc7e4bce.png

2024-04-09 14:01:31

- reply

Lattice cryptography, broken by hand in less than 30 minutes
https://media.infosec.exchange/infosec.exchange/media_attachments/files/112/241/623/760/594/912/original/ec8f8ead474f5c53.jpg

2024-04-09 14:01:31

- reply

Lattice cryptography, broken by hand in less than 30 minutes
https://media.infosec.exchange/infosec.exchange/media_attachments/files/112/241/623/760/594/912/original/ec8f8ead474f5c53.jpg

2024-04-09 13:48:59

- reply

After blowing up singularities, we start seeing the internal structure
https://media.infosec.exchange/infosec.exchange/media_attachments/files/112/241/570/916/094/121/original/0024f84979f85bca.jpg

2024-04-09 13:44:00

- reply

The puzzle creators seem to have gone with a number field lattice (the maximal order of Q[X]/(X²+1), also known as the Gaussian Integers)
While cyclotomic fields of powers of two are common in lattice cryptography, a more popular choice of base field would be Q[X]/(X²⁵⁶+1)
https://media.infosec.exchange/infosec.exchange/media_attachments/files/112/241/549/906/312/030/original/d4d7294650984bac.jpg

2024-04-09 13:40:40

- reply

As we can see here, lattice cryptography is a very secure, well trusted type of cryptography, having the highest rating of 10. That means it should have at least 128 bits of security.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/112/241/538/152/328/713/original/a213f34f1464e427.jpg

2024-04-09 13:38:46

And now it's time for some lattice cryptography
https://media.infosec.exchange/infosec.exchange/media_attachments/files/112/241/534/528/388/176/original/fc323df028c42250.jpg

2024-03-28 14:16:19

It looks like I have successfully coined the term "fancy cryptography" for cryptography beyond encryption and signatures.
You are welcome.

2024-03-28 14:16:19

It looks like I have successfully coined the term "fancy cryptography" for cryptography beyond encryption and signatures.
You are welcome.

2024-03-27 15:44:34

Public Service Announcement:
NIST also has requirements for Key Derivation Functions. You cannot just invent your own KDF slap it in front of AES-GCM, and still pretend that you're FIPS compliant.
Sadly. Because it would improve security if you could.

2024-03-26 12:59:56

- reply

Won against cipher on technical grounds (proved that ridiculously powerful attacker is powerful). Then slept.

2024-03-26 12:27:51

Didn't sleep. Attacked cipher instead.