The Riemann Hypothesis is a famous unsolved problem dating from 1859. I will present a simple proof using a radically new approach. It is based on work of von Neumann (1936), Hirzebruch (1954) and Dirac (1928).

A major advance in AI occurred in 2012 when AlexNet won the ImageNat competition with a deep network. The success was sufficiently better than previous years that deep networks were applied in many applications with great success. However, there is little understanding of why deep learning works. This talk will give an introduction to machine learning and then focus on current research directions in deep learning.

Algorithms are not programs. They can and should be written with math rather than programming languages or pseudo-languages. This applies to many more algorithms than the ones taught in algorithm courses.

This talk will focus on the current market for “Big Data” products, specifically those that deal with one or more of “the 3 V’s.” I will suggest that the volume problem for business intelligence applications is pretty well solved by the data warehouse vendors; however upcoming “data science” tasks are poorly supported at present. On the other hand, there is rapid technological progress, so “stay tuned.” In the velocity arena recent “new SQL” and stream processing products are doing a good job, but there are a few storm clouds on the horizon. The variety space has a
collection of mature products, along with considerable innovation from startups.
I will discuss opportunities in this space, especially those enabled by possible disruption from new technology. Also discussed will be the pain levels I observe in current enterprises, culminating in my presentation of “the 800 pound gorilla in the corner.”

Since the ancient times people have tried to understand systems of polynomial equations for practical and/or abstract reasons. It was slowly understood that geometry can be used to understand such equations by "graphing" their solutions in some appropriate space. The geometric picture provides a window for our intuition to guide us in how to investigate the qualitative nature of the solutions while algebra provides the machinery for rigorous proofs. In this talk I will try to explain some aspects of this story in simple terms.

A distributed ledger is a tamperproof sequence of data that can be read and augmented by everyone. Distributed ledgers stand to revolutionize the way a democratic society operates. They secure all kinds of traditional transactions – such as payments, asset transfers, titling – in the exact order in which they occur; and enable totally new transactions – such as cryptocurrencies and smart contracts. They can remove intermediaries and usher in a new paradigm for trust. As currently implemented, however, distributed ledgers cannot achieve their enormous potential.

Algorand is an alternative, democratic, and efficient distributed ledger. Unlike prior ledgers based on ‘proof of work’, it dispenses with ‘miners.’ Indeed, Algorand requires only a negligible amount of computation. Moreover, its transaction history does not ‘fork’ with overwhelming probability: i.e., Algorand guarantees the finality of all transactions.

Finally, Algorand enjoys flexible self-governance. A successful society must be able to evolve, and a cryptocurrency cannot be an ocean liner on autopilot. By using its hallmark propose-and-agree process, Algorand can correct its course as necessary or desirable, without any ‘hard forks.’ Thanks to this core process, Algorand can routinely summon the contribution of each single ‘token’ for reaching any future decision.

At the beginning of the 20th century Einstein changed the way we think about Time. Now, early in the 21st century, the measurement of Time is being revolutionized by the ability to cool a gas of atoms to temperatures millions of times lower than any naturally occurring temperature in the universe. Atomic clocks, the best timekeepers ever made, are one of the scientific and technological wonders of modern life. Such super-accurate clocks are essential to industry, commerce, and science; they are the heart of the Global Positioning System (GPS), which guides cars, airplanes, and hikers to their destinations. Today, the best primary atomic clocks use ultra-cold atoms, achieve accuracies of about one second in 300 million years, and are getting better all the time, while a new generation of atomic clocks is leading us to re-define what we mean by time. Super-cold atoms, with temperatures that can be below a billionth of a degree above absolute zero, use, and allow tests of some of Einstein's strangest predictions.  
 
This will be a lively, multimedia presentation, including exciting experimental demonstrations and down-to-earth explanations about some of today's hottest (and coolest) science.

Laureates: Visit to Local Schools 

YR: Visit to Local Institutions

I show by examples how the Theory of Programming can be taught to first year CS undergraduates. The only prerequisite is their High School acquaintance with algebra, geometry and propositional calculus. The students’ motive is to learn how to construct and debug their programs.
 
I start with the familiar laws for disjunction in Boolean Algebra, illustrated by disjunction. A deductive rule for proof by cases is derived from the algebra. The algebra is extended by the operators of spatial and temporal logic: William of Occam’s ‘while’ (|) and ‘then’ (;). They satisfy the same familiar algebraic laws, including distribution through disjunction. A weak interchange law describes how they distribute through one another by ‘shuffling.’ Proof rules are then derived for a modal logic of time and space.

If this logic is applied to propositions about the behaviour of programs, the ‘while’ and ‘then’ operations can be reinterpreted as sequential and concurrent compositions of programs. The proof rules of the modal logic are then definitionally equivalent to two historic logics due to Hoare and Milner. They are now used widely for mechanical reasoning about correctness of programs and about implementations of programming languages.

The notion of an expander was introduced in early nineteen seventies by M.S. Pinsker in his work on the complexity of a concentrator. In recent decades the theory of expanders attracted a lot of attention and became a rather big industry. I will talk only about the development of this theory until approximately 1987-88.

In the 1980s, Mead and Conway democratized chip design and high-level language programming surpassed assembly language programming, which made instruction set advances viable. Innovations like Reduced Instruction Set Computers (RISC), superscalar, and speculation ushered in a Golden Age of computer architecture, when performance doubled every 18 months. The ending of Dennard Scaling and Moore’s Law crippled this path; microprocessor performance improved only 3% last year! In addition to poor performance gains of modern microprocessors,
Spectre recently demonstrated timing attacks that leak information at high rates.

The ending of Dennard scaling and Moore’s law and the deceleration of performance gains for standard microprocessors are not problems that must be solved but facts that if accepted offer breathtaking opportunities. We believe high-level, domain-specific languages and architectures, freeing architects from the chains of proprietary instruction sets, and the demand from the public for improved security will usher in a new Golden Age. Aided by open source ecosystems, agilely developed chips will convincingly demonstrate advances and thereby accelerate commercial adoption. The instruction set philosophy of the general-purpose processors in these chips will likely be RISC, which has stood the test of time. We envision the same rapid improvement as in the last Golden Age, but this time in cost, energy, and security as well as in performance.

Like the 1980s, the next decade will be exciting for computer architects in academia and in industry!

The concept of autonomy is key to the IoT vision promising increasing integration of smart services and systems minimizing human intervention. This vision challenges our capability to build complex open trustworthy autonomous systems. We lack a rigorous common semantic framework for autonomous systems. There is currently a lot of confusion regarding the main characteristics of autonomous systems. In the literature, we find a profusion of poorly understood “self”-prefixed terms related to autonomy such as Self-healing, Self-optimization, Self-protection, Self-awareness, Self-organization etc. It is remarkable that the debate about autonomous vehicles focuses almost exclusively on AI and learning techniques while it ignores many other equally important autonomous system design issues. Autonomous systems involve agents and objects coordinated in some common environment so that their collective behavior meets a set of global goals. We propose a general computational model combining a system architecture model and an agent model. The architecture model allows expression of dynamic reconfigurable multi-mode coordination between components. The agent model consists of five interacting modules implementing each one a characteristic feature: perception, reflection, goal management, planning and self-adaptation. It determines a concept of autonomic complexity accounting for the specific difficulty to build autonomous systems.
We emphasize that the main characteristic of autonomous systems is their ability to handle knowledge and adaptively respond to environment changes. A main conclusion is that autonomy should be associated with functionality and not with specific techniques.
We conclude that autonomy is a kind of broad intelligence. Building trustworthy and optimal autonomous systems goes far beyond the AI challenge.

The concept of equilibrium, in its various forms, has played a central role in the development of Game Theory and Economics. The mathematical properties and computational complexity of equilibria are also intimately related to mathematical programming, online learning, and fixed point theory. More recently, equilibrium computation has been proposed as a means to learn generative models of high-dimensional distributions. In this talk, we review fundamental results on minimax equilibrium and its relationship to mathematical programming and online learning. We then turn to Nash equilibrium, reviewing some of our work on its computational intractability. We conclude with modern applications of equilibrium computation, presenting recent progress and open problems in the training of Generative Adversarial Networks.

Looking back over more than 50 years of doing research, I can discern several key “ingredients” that helped me in my work. First, I was not afraid of doing things that others thought foolish, and this was also important in the work of many of my most successful colleagues. Second, I had what Buddhism calls “beginner’s mind” – I did my best work in cryptography when I knew only a small fraction of the relevant math that I now know. Third, I trusted my intuition, rather than rejecting it as “irrational.” This talk illustrates each of these points from my personal experience.

Looking back over more than 50 years of doing research, I can discern several key “ingredients” that helped me in my work. First, I was not afraid of doing things that others thought foolish, and this was also important in the work of many of my most successful colleagues. Second, I had what Buddhism calls “beginner’s mind” – I did my best work in cryptography when I knew only a small fraction of the relevant math that I now know. Third, I trusted my intuition, rather than rejecting it as “irrational.” This talk illustrates each of these points from my personal experience.