Speaker 0: in today's digital landscape where our personal and professional lives intertwine with technology, the quest for uncompromising data security grows ever more urgent. imagine a world where sensitive information is shielded by encryption systems so advanced that it renders hacking nearly impossible. researchers are inching closer to just that with groundbreaking approach that marries artificial intelligence and holographic technology. this innovation is not only not just a fleeting trend. it could very well change how we think about data protection. so what's going on with this, john? Speaker 1: yeah. so this is cool. as you and i have talked about, like, we need to get away from rsa aes encryption because they're vulnerable, and they're just gonna get more and more vulnerable. so this is cool. this is merging holography with neural networks. so researchers led by stelios, i'm not gonna try pronouncing his last name, at the institute electronic structure and laser in greece. what they've done is they've unveiled a groundbreaking optical encryption system. and like i said, it merges holography and neural networks, and it creates, apparently, an nearly unbreakable security. how it works is the system uses high power lasers, which pass through ethanol to generate chaotic light scattering, which encodes data into scrambled holograms. so traditional decryption methods fail here because the chaotic distortion, it can't be reversed computationally, which is awesome. kinda like that cloud flare, lava lamp kind of thing. so instead for this, ai trained neural networks are mimicking the human's brain synaptic connections and learn to decode these patterns, which produce a unique decryption key only accessible to the system's owner. so, again, it's a great encryption, but it all comes down to the owner as well. but still cool. so in tests, the system achieved ninety to ninety five percent accuracy in reconstructing the encrypted images of handwritten digits, animals, and everyday objects. now you might think, why use ethanol? well, ethanol, as a nonlinear medium, what that ensures is a cost effective rapid scrambling, adapting to chaos conditions like thermal turbulence. so their future plans, what they're looking to do, now that they've achieved this, obviously, they still have to hone it in and work on it. but they're looking to integrate multifactor authentication and reduce reliance on expensive high power lasers to enable broader accessibility. so this is cool because, again, it addresses critical needs in sectors like digital currencies, health care, secure communications, where, essentially, rsa and aes are becoming vulnerable. so it's cool because it leverages laser filamentation where high power femtosecond lasers create self guided channels in air or liquid. and that chaotic process scrambles light into holograms, which is really cool. i don't know if you have a video of it, but i don't think we do. but, yeah. so ethanol's nonlinear properties, that enables rapid low cost distortion, which makes it ideal for real world applications. so this isn't just in the lab, high, like, pie in the sky kind of stuff. this has potential. so, what do you think about this, craig? do you think this is where we're gonna go next? Speaker 0: well, i think it's really cool. still wanna understand the costs related to something like this because it sounds really expensive. so that it doesn't sound like the technology i can use to encrypt my hard drive on my laptop because it's just too expensive. Speaker 1: yes. yeah. so that's true. Speaker 0: alright. so the question is this at a very high level, are we gonna get something that scales down at this level that, you know, doesn't involve, ethanol and having some lasers and chemical and all this thing, you know, around to figure it out? or, like, that's really cool for high end data centers that are cost, you know, for the us government, google, and whoever, but for the rest of us? Speaker 1: well, that's a good point. but i i think that because of where they're looking to go so, obviously, the cost, the high power laser lasers are expensive and bulky. freaking lasers. and researchers are they're exploring cost effective alternative like diode laser. so if they can do diode lasers, that's a lot more, cost effective and makes it a lot more attainable for people like you and i. so a couple other things they're looking. so, scalability. so current experiments are using small scale ethanol containers. so scaling to long distance atmospheric transmission, that's gonna require further testing. so yeah. so you raise a good point, but i think they've already thought of that, and i think that they're trying to make this to become more mainstream, if that makes sense. Speaker 0: absolutely. and i look at this. you know? the holographic technology forms three dimension security layers that confuse and deter sophisticated cyber intrusion and that the ai algorithms dynamically generate encryption keys, and combining these two introduces multifactor encryption with overlapping security dimensions. so at the heart of just that, that technology should be available to the laptop people, for lack of a better word, you know, where we can actually now come up with a couple algorithms and say, hey. this is kinda the random algorithm that we get from this laser in ethanol magic, now randomly create something similar to give me multiple layers of encryption, to do this thing. now, so we can probably, at some point, get ai to generate all of it, but the question is, will it be as good? Speaker 1: well yeah. and my question is, if they can train ai to encrypt, can ai not be trained to decrypt, or is it just too random for ai to decrypt? right? because, like, any, any, encryption can be broken so far. but think? do you think that we could teach the ai or neural network to decrypt and hack this, or is that unhackable based on what we're reading so far? Speaker 0: you know, every new encryption thing's unhackable. like, we've talked about this, and then, you know, every major company comes out with this new unhackable thing. we're encrypting this. don't worry. we've got your back. there's no way it's unhackable. we'll end. like, two days later, it's hacked. so i wanna keep that in mind that this is still kind of, not for public consumption. it's not out there that everyone can start testing against. i don't know if this will be truly unhackable. i really like the concept, but Speaker 1: yeah. Speaker 0: flipping it around, you can get ai, like you're saying, to if you can get ai to build it, you can get ai to tear it apart. but i think it's the combination of some randomness related to building holograms and multilevel encryption in it that you're probably even hiding a lot of the encryption in the hologram or whatever the thing it is. so it takes a little bit more than what you think. and you have to have the right combination in the right order, in the right you know, that people wouldn't i think it's gonna be way more complicated, especially if you use different randomization techniques for each of those things. okay. we're going we have so many different layers of encryption in each one of those encryption in this holographic thing. let's just say we have twenty layers of encryption because we can do anything. we have ai and compute power, and it's easy to do now. mhmm. so we have those, and it goes in and it encrypts all that encryption in encryption in encryption in different layers inside this holographic image. it's gonna be hard to decode that no matter what. now could quantum decode it? i mean, it depends. like, is it is quantum really tapping into the multiverse and doing, more computing power that we've ever generated since the beginning of time in, like, five seconds and generating our answer? or is it always if it's always changing because it's based on the imagery of the laser interacting with the ethanol, then, it's it's always changing. but if you're saying, let's create an ai algorithm that do it, yes. it would be hackable, but i just think it would be extremely hard. Speaker 1: yeah. and it comes down to how random is random. like, if you if you if you test it out and you run it long enough, is it actually truly random, or will there patterns that will emerge? right? Speaker 0: yeah. and we don't know yet. right? Speaker 1: yeah. oh, and and that's just in general, though. like, randomness how random is randomness? right? i think there was a paper. i don't know if we covered it. i don't think so. but there's nothing that's really random if you actually dive into and look at it in a big enough scale. Speaker 0: we talked about that actually. at a large enough scale, there's nothing random. there's always patterns. yeah. they did this, and it showed that the neural network could accurately retrieve the encoded images ninety to ninety five percent of the time. so they haven't even got it down where they can with the stuff to unencrypt it themselves. so it's Speaker 1: a it's a lot of new technology, though. right? like, they're they're still learning it. they're still training it, all that sort of stuff. so Speaker 0: absolutely. it's it's funny that, you know, it's so secure that even with all the keys, it you can't unlock it all the time. so, you know, this is very interesting. i'm not sure if this is gonna be for everyone. like, how do you encrypt? how do you decrypt? what is the cpu powered required to do all this? Speaker 1: mhmm. Speaker 0: and does it happen at real time? so every time i have to access data, we're doing that in the network application layer or whatever going back and forth depending on where the encryption happens. i assume it's at the application layer, but could be anything. Speaker 1: yeah. Speaker 0: the yeah. it's how much he like, the concept is great, but it sounds like you're gonna take quantum level of computing power to do the encryption in the first place just through all this stuff. Speaker 1: yeah. yeah. it's it it'll be interesting to see where this goes and how far it gets, to actual commercialization. Speaker 0: yeah. i'm i love the concept of better encryption for everything, but, you know, as we look to the future, one can't help but wonder what our world might look like when data breaches become a distant memory. with ongoing advancements in ai holographic technologies, we stand at a precipice of a new age in cybersecurity. will adaptive encryption systems and two factor authentications become the new standard for digital communication? while the timeline for widespread adopt adoption is still a mystery, one thing is for certain. the intersection of ai and encryption technologies offers not just hope, but tantalizing glimpse into what future might hold for our digital security landscape and, alternatively, the hacking landscape.