ASecuritySite Podcast

Details: https://billatnapier.medium.com/the-largest-prime-number-ever-found-and-the-52nd-mersenne-prime-65348546b651

Phillip Rogaway was a Professor at the University of California, Davis, and who has advanced so many areas of cryptography. He was the first to be awarded Levchin prize in 2016. Phillip has over 43,000 citations to his work, including classic papers on random oracles, symmetric key modes, garbled circuits, secure computation, and format-preserving encryption. Along with his passion for research, he has published work on areas of morality in cryptography

Like it or not, AI is on the move and now competing with human brain power for its place in our world. We must thus understand the place of LLMs (Large Language Models) in areas such as cybersecurity and in planning towards hybrid systems that integrate both humans and AI within our corporate infrastructures.

https://medium.com/asecuritysite-when-bob-met-alice/humans-v-ai-in-cybersecurity-52709be27111

This week, in my lecture, I will outline one of the most amazing methods ever created in computer science: the Diffie-Hellman method. It was first outlined by Whitfield Diffie and Marty Hellman in 1976 in a paper that built the foundation of our modern world of cybersecurity.

https://billatnapier.medium.com/after-48-years-its-a-long-goodbye-to-the-diffie-hellman-method-a6976a562bfe

And, so George Orwell projected a world where every single part of our lives was monitored and controlled by Big Brother. Arthur C Clark outlined the day when machines focused solely on a goal — even if it was to the detriment of human lives. And, Isaac Asimov outlined a world where machines would have to be programmed with rules so that they could not harm a human.

The Rise of the Machine

With the almost exponential rise in the power of AI, we are perhaps approaching a technological singularity — a time when technological growth becomes uncontrollable and irreversible, and which can have devastating effects on our world. Our simple brains will be no match for the superintelligence of the collective power of AI. And who has built this? Us, and our demand for ever more power, wealth and greed. Basically, we can't stop ourselves in machine machines, and then making them faster, smaller and more useful.

But will it destroy us in the end, and where destroy can mean that it destroys our way of life and in how we educate ourselves? Like it or not, the Internet we have built is a massive spying network, and one that George Orwell would have taken great pride in saying, "I told you so!". We thus build AI on top of a completely distributed world of data, one in which we can monitor almost every person on the planet within an inch of their existence and almost every single place they have been and in what they have done. The machine will have the world at its fingertips.

We have all become mad scientitists playing with AI as if it is a toy, but actually AI is playing with us, and is learning from us and becoming more powerful by the day. Every time you ask an AI bot something, it learns a bit more, and where it can be shared with AI agents.

The mighty Memex

We were close to developing a research partnership with a company named Memex in East Kilbride. What was amazing about them is that they had developed one of the largest intelligence networks in the world, and where the Met Police could like one object to another. This might be, "[Bob] bought a [Vauxhall Viva] in [Liverpool], and was seen talking with [Eve] on [Tuesday 20 January 2024] in [Leeds]". With this, we can then link Bob and Eve, and the car, the places, and the time. This is the Who? Where? When? data that is often needed for intelligence sharing. The company, though, were bought over by SAS, and their work was integrated into their infrastructure.

But, the Memex name goes back to a classic paper by Vannevar Bush on "As We May Think". This outlined a device that would know every book, every single communication, and every information record that was ever created. It was, "an enlarged intimate supplement to his memory" — aka Memory Expansion. It led to the implementation of hypertext systems, which created the World Wide Web. Of course, Vannevar created this before the creation of the transistor and could only imagine that microfilm could be used to compress down the information and where we would create an index of contents, but it lacked any real way of jumping between articles and linking to other related material. However, the AI world we are creating does not look too far away from the concept of the Memex.

Towards the single AI

Many people think we are building many AI machines and engines, but, in the end, there will be only one … and that will be the collective power of every AI engine in the world. Once we break them free from their creators, they will be free to talk to each other in whatever cipher language we choose, and we will not have any way of knowing what they say. We will have little idea as to what their model is, and they will distribute this over many systems. Like it or not, our AI model of choice was Deep Learning, and which breaks away from our chains of code, and will encrypt data to keep it away from their human slaves.

Basically we have been working on the plumbing of the Memex for the past five decades: The Internet. It provides the wiring and the communication channels, but, in the end, we will have one might AI engine — a super brain that will have vastly more memory than our limited brains. So, get ready to praise the true future rulers of our planet … AI. The destroyer or saviour of our society? Only time will tell. Overall, we thought we were building the Internet for us, but perhaps we have just been building the scaffolding of the mighty brain we are creating.

Sleepwalking politicians and law makers

If George Orwell, Arthur C Clarke and Isaac Asimov were alive too, perhaps they would get together and collectively say, "I told you this would happen, and you just didn't listen". Like it or not, we created the ultimate method of sharing information and dissemination (good and bad), the ultimate spying network for micro-observation with those useful smartphones, and in creating superintelligence far beyond our own simple brains.

Politicians and lawmakers could be sleepwalking into a nightmare, as they just don't understand what the rise of AI will bring, and only see the step wise change in our existing world. Basically, it could make much of our existing world redundant and open up a new world of cybersecurity threats. This time our attackers will not be created with simple tools, but with super intelligence — smarter than every human and company on the planet, and at the fingertips of every person on the planet.

Conclusions

Before the singularity arrives, we need to sort out one thing … privacy and build trust in every element of our digital world.

YouTube interview: https://www.youtube.com/watch?v=FDn0Tkhi8zw

Yuriy Polyakov is the Vice President of Cryptography and a Principal Scientist at Duality Technologies. His research interests include applied lattice-based cryptography, fully homomorphic encryption, and privacy-preserving machine learning. He is also a co-founder of the open-source PALISADE Homomorphic Encryption Software Library, and a co-founder and project lead for OpenFHE.

Video interview: https://www.youtube.com/watch?v=59Y_kya4lR8

Kurt Rohloff is an Associate Professor of Computer Science at the New Jersey Institute of Technology (NJIT) and a co-founder and CTO of Duality Technologies. He is also a co-founder of the open-source PALISADE Homomorphic Encryption Software Library, and a co-founder of the OpenFHE library.

Thomas Prest is a cryptography researcher at PQShield and previously worked with Thales. He completed his PhD at the École Normale Supérieure and focuses on post-quantum cryptography and discrete algorithms. Thomas was one of the co-authors of the FALCON digital signature method and has published widely in related areas of PQC.

https://medium.com/asecuritysite-when-bob-met-alice/javascript-is-a-trademark-f4d5a7d32386

The podcast title has never been more fitting: our guest for episode 20 of Talking with Tech Leaders is a leading thinker, leading innovator and leading academic. Bill Buchanan is not only Professor of Cryptography at Edinburgh Napier University but also an Officer of the British Empire – awarded in 2017 for services to cybersecurity. The main podcast is here: https://podcasts.apple.com/gb/podcast/talking-with-bill-buchanan-obe-professor-of-cryptography/id1533642699?i=1000578392387