Great video. I'd love to see more about encryption. Maybe even something that's useful without a computer? One time pads and dice can do the trick but you already mentioned the downside of that.
Ahh, I remember RIPA in about 2000... I wrote to my MP about the implications of this (in English, not tech and not conspiracy babbling).... who ignored everything I said.
@llorTA toN Quite. The law can punish non-compliance but it actually can't compel. Though that is somewhat splitting hairs, I suppose. On the other hand , it seems to have worked out okay for Apple and Whatsapp so far.
Very clearly. So do you know about the NXP recommend upgrade DESFire EV2 to EV3 card that is the encryption AES128 problem or another issue? I'm thinking about the ev2 encryption is not security enough and they refer to EV3 with higher encry....can share with me what do you thing?
In the case of Diffie-Hellman key exchange, you use a secret key that can be created from two sub-keys in two different ways. If you combine your private sub-key with your partner's public sub-key, and they combine their private sub-key with your public sub-key, you'll both get the same secret key, while any eavesdropper will only get the two public keys which cannot be combined into the secret key.
As others have said, we use public key cryptography to do this today, but a very large responsibility of the cryptographic arms of intelligence services used to be distribution of key material. Obviously this was vulnerable to interception and betrayal, as well as the quadratic way in which the amount of key material required increases with the number of participants. It still is the case that some key material is hand-distributed, often for tactical radios in a battlefield environment, using something called a 'fill device'.
It sounds, naively, that prepending a verification block to the ciphertext is a weakness, but given a suitably large key and a sufficiently strong cipher, it's actually not. There are other methods of verification as well. Notably OpenSSL in CBC mode uses the PKCS#7 unpadding to determine correctness (although it's not completely robust) and modern cipher modes like GCM handle message authentication and integrity as part of the process.
It is a weakness, and here's why: Recall Turing's dilemma where he couldn't be sure that the messages he was getting out of his de-enigma contraption were correct unless he knew what was in the messages. The breakthrough came when he noticed that the human operators sending encrypted messages were appending the words "hail leader" (or similar) to their cleartext. So that became his checksum. Attackers know what keys they're using when trying to break encryption. If an algorithm is prepending the cleartext with the key, then an attacker need only decrypt a message to the length of the key. If the output doesn't match the key, then they can skip the effort to decrypt the rest of the message and move on to the next attempt. One of the goals of cryptography should be to maximize the cost to the attacker to decrypt messages. I'm sure there are obvious ways to do that. But prepending your key to your message doesn't seem to be one of them.
What?? That was the most ridiculously confusing word salad I’ve heard in a long time, and I actually write programs dealing with cryptography, CSPRNGs, entropy, symmetrical/asymmetrical algorithms, hashes, etc. and have calculated entropy many times. Wow!!🤪
As with other computer scientists, he spoke many words that amounted to not including the check bits that he said he wanted. Now you can't brute force the key, unless you know some properties fileformat that is encrypted.
I don't think it's just that there are no check bits, but that the encryption key is the same length as the message. Probably. I definitely agree with the 'spoke many words' part - everything was technically correct, but the actual explanation still pretty confusing. XD
This is my field. I just started the video. Will this be the first video (besides mine) to ever get cryptographic entropy right? [update] he started in the right direction, but didn't finish.
@@davidjohnston4240 well there’s one thing that I know. I’m much, much smarter than you. Orders of magnitude smarter. But you’re more handsome. To be clear I’m just trolling you because of the tenor of your other comments. But that doesn’t mean you’re wrong. You sure you made your videos public?
Great video. I'd love to see more about encryption. Maybe even something that's useful without a computer? One time pads and dice can do the trick but you already mentioned the downside of that.
OTPs are still so fascinating to me.
Love these last information theory videos, would be cool to see something about soft vs hard decoding
I love the way you use magic markers. Please be sponsored and always use it
Learned so much from the hand waving 😄
The weakness in encrytion is not the algorithms, it is that you can compelled to hand the keys over by law.
*be
Social Engineering is every human-involved system's biggest weakness
Ahh, I remember RIPA in about 2000... I wrote to my MP about the implications of this (in English, not tech and not conspiracy babbling).... who ignored everything I said.
"sorry i lost the keys, too bad"
@llorTA toN Quite. The law can punish non-compliance but it actually can't compel. Though that is somewhat splitting hairs, I suppose. On the other hand , it seems to have worked out okay for Apple and Whatsapp so far.
In future, may I recommend, when you're filming, to get a pen that works?
I always wondered why the vernam cipher was considered 100% secure
but impractical
@@An.Individual The WW2-era Soviets (and postwar agents) used to use one time pads. More than a few messages were broken because pads were reused.
@@davidgillies620 Yes. I think you’re referring to the Venona decryptions declassified in 1995
The marker displaying entropy as he writes about entropy lol.
speaking of entropy, can u cover physical unclonable functions? something on the hardware side? its pretty intriguing
That's a more arcane topic, as in fewer people really understand them. I might cover them in the future, since I do design PUFs as part of my day job.
@@davidjohnston4240 do you have any suggestions for where I could learn about them in the mean time?
His way of explaining things is at par with his marker quality
Very clearly. So do you know about the NXP recommend upgrade DESFire EV2 to EV3 card that is the encryption AES128 problem or another issue? I'm thinking about the ev2 encryption is not security enough and they refer to EV3 with higher encry....can share with me what do you thing?
I cannot stand how dry his marker is.
They r green friendly...
I still don't understand how you transfer the initial secret keys without having the encryption keys, and without anyone else grabbing those keys.
In the case of Diffie-Hellman key exchange, you use a secret key that can be created from two sub-keys in two different ways. If you combine your private sub-key with your partner's public sub-key, and they combine their private sub-key with your public sub-key, you'll both get the same secret key, while any eavesdropper will only get the two public keys which cannot be combined into the secret key.
For this, you should watch Computerphile's video on Diffie-Hellman
As others have said, we use public key cryptography to do this today, but a very large responsibility of the cryptographic arms of intelligence services used to be distribution of key material. Obviously this was vulnerable to interception and betrayal, as well as the quadratic way in which the amount of key material required increases with the number of participants. It still is the case that some key material is hand-distributed, often for tactical radios in a battlefield environment, using something called a 'fill device'.
The channel “art of the problem” does a brilliant job of explaining both Diffie Hellman and RSA as well.
A wrench, any size, using percussive maintenance, will decrypt that using the right rhythm
your bank account too
0:24 Dude, you gotta stop telling everyone the secret to life, the universe, and everything! 😂
What the deal with 42 ? It comes more often than one might expect.
It’s from Douglas Adams’ book “The Hitchhiker’s Guide to the Galaxy” and given as the answer to the meaning of life.
@@F16_viper_pilot Thanks mate
It sounds, naively, that prepending a verification block to the ciphertext is a weakness, but given a suitably large key and a sufficiently strong cipher, it's actually not. There are other methods of verification as well. Notably OpenSSL in CBC mode uses the PKCS#7 unpadding to determine correctness (although it's not completely robust) and modern cipher modes like GCM handle message authentication and integrity as part of the process.
It is a weakness, and here's why:
Recall Turing's dilemma where he couldn't be sure that the messages he was getting out of his de-enigma contraption were correct unless he knew what was in the messages. The breakthrough came when he noticed that the human operators sending encrypted messages were appending the words "hail leader" (or similar) to their cleartext. So that became his checksum.
Attackers know what keys they're using when trying to break encryption. If an algorithm is prepending the cleartext with the key, then an attacker need only decrypt a message to the length of the key. If the output doesn't match the key, then they can skip the effort to decrypt the rest of the message and move on to the next attempt.
One of the goals of cryptography should be to maximize the cost to the attacker to decrypt messages. I'm sure there are obvious ways to do that. But prepending your key to your message doesn't seem to be one of them.
This could only mean one thing! But I don't know what that is..
20+42 = 2042
I see he’s using Javascript🤣
What?? That was the most ridiculously confusing word salad I’ve heard in a long time, and I actually write programs dealing with cryptography, CSPRNGs, entropy, symmetrical/asymmetrical algorithms, hashes, etc. and have calculated entropy many times. Wow!!🤪
What just happened?
The 1's in the font in the animation is so weird.
Magnetic ink character recognition code font.
"It's always 42"
Brady knows what's up
As with other computer scientists, he spoke many words that amounted to not including the check bits that he said he wanted. Now you can't brute force the key, unless you know some properties fileformat that is encrypted.
I don't think it's just that there are no check bits, but that the encryption key is the same length as the message. Probably. I definitely agree with the 'spoke many words' part - everything was technically correct, but the actual explanation still pretty confusing. XD
This is my field. I just started the video. Will this be the first video (besides mine) to ever get cryptographic entropy right? [update] he started in the right direction, but didn't finish.
You have no videos posted to your channel though
Who are you?
where's your video
That's 13 videos so far.
@@davidjohnston4240 well there’s one thing that I know. I’m much, much smarter than you. Orders of magnitude smarter. But you’re more handsome.
To be clear I’m just trolling you because of the tenor of your other comments. But that doesn’t mean you’re wrong. You sure you made your videos public?
Pad it out. Problem solved.
✌️
peace & love
Algorithm.
First!
First