This week's episode focuses on the interface between physics and computing, with deep dives into how artificial intelligence (AI) is contributing to medical physics and how silicon could form the basis of a future quantum computer. First, we hear from Tami Freeman, Physics World's resident expert on medical physics, about a new positron emission tomography (PET) scanner that can image a patient's whole body much more quickly (or at higher resolutions) than is possible with current commercial scanners. We then stick with the medical theme to discuss three recent examples of how AI is being used in medicine: firstly to diagnose skin conditions (but, disturbingly, only if the patient's skin is white); secondly to help radiologists detect lung tumours in X-rays; and thirdly to develop better radiotherapy treatment plans. There are several ways of constructing the qubits, or quantum bits, that make up a quantum computer, and this week we hear from a trio of researchers – Fernando Gonzalez-Zalba, Alessandro Rossi and Tsung-Yeh Yang – who have been developing silicon-based qubits. Their work is part of a Europe-wide collaboration between universities, government laboratories and companies called MOS-Quito, and you can read more about it in their article for the Physics World Focus on Computing.

Agenda: 6:00 - 6:30 pm: Social Hour with Food, Drinks, Beer & Wine 6:30 - 6:35 pm: Introduction & Announcements 6:35 - 7:05 pm: Talk 1 Managing the Full Deployment Lifecycle of Models with the MLflow Model Registry (Databricks) 7:05 - 7:35 pm: Talk 2 MLflow on and inside Azure (Microsoft) 7:35 - 8:05 pm: Talk 3 TensorFlow(X) Data Validation: Better ML through better data (Google) 8:05 - 8:30 pm: Additional Networking Talk 1 - Title: Managing the Full Deployment Lifecycle of Models with the MLflow Model Registry Presenter: Mani Parkhe, Databricks Abstract: MLflow is an open-source platform to manage the machine learning lifecycle, including experiment tracking, reproducible runs, and model packaging. In this talk, we provide an overview of the latest component of MLflow, the Model Registry, which serves as a collaborative hub where teams can share, discuss, use, inspect, and track the lineage of models. Model Registry was introduced in MLflow 1.4 and is in Private Preview on Databricks With this addition, MLflow provides end-to-end management of the deployment lifecycle of models from experimentation to online testing and production, complete with approval and governance workflows. Bio: Mani Parkhe is an ML/AI Platform Engineer at Databricks, focusing on the customer and open-source platform initiatives, which enable data discovery, training, experimentation, and deployment of ML models on the cloud. After spending 15 years building software for semiconductor chip CAD, Mani transitioned to building big data infrastructure, distributed systems and web services, and machine learning platforms.

At this week's Intel HPC Developer Conference in Denver, Colrado, USA this week, Raja Koduri, senior vice president, chief architect, and general manager of architecture, graphics and software at Intel (pictured), introduced a new category of discrete general-purpose GPUs optimised for artificial intelligence (AI) and high performance computing (HPC) convergence. The Ponte Vecchio general purpose GPU is based on the Xe architecture. "HPC and AI workloads demand diverse architectures, ranging from CPUs, general-purpose GPUs and FPGAs, to more specialised deep-learning neural network processors (NNPs), which Intel demonstrated earlier this month," said Koduri. Ponte Vecchio is architected for HPC modeling and simulation workloads and AI training. It will be manufactured on Intel's 7nm technology and will be Intel's first Xe-based GPU optimised for HPC and AI workloads.

MOTION: This House Believes AI Will Bring More Harm Than Good This debate was run in association with IBM Research. Proposition: Project Debater Project Debater is designed by IBM research. It will deliver a speech based on over 1,100 arguments collected from Union members and others over the past week. It will not be taking points of information. Sharmila Parmanand Sharmila Parmanand is a PhD Candidate in Gender Studies at the University of Cambridge and a Gates Scholar.

Could the same computer algorithms that teach autonomous cars to drive safely help identify nearby asteroids or discover life in the universe? NASA scientists are trying to figure that out by partnering with pioneers in artificial intelligence (AI)--companies such as Intel, IBM and Google--to apply advanced computer algorithms to problems in space science. Machine learning is a type of AI. It describes the most widely used algorithms and other tools that allow computers to learn from data in order to make predictions and categorize objects much faster and more accurately than a human being can. Consequently, machine learning is widely used to help technology companies recognize faces in photos or predict what movies people would enjoy.

Of all the emerging technologies that will change our daily lives, none has more transformative potential than artificial intelligence. And AI -- the use of computers to solve problems that would normally require natural, or human, intelligence -- will also have a profound effect on the global balance of economic and military power. It will change how societies are governed and people are ruled. Debates about whether China or the U.S. will dominate the 21st century are thus necessarily debates about who will lead in AI innovation, and whether democratic or authoritarian systems are better suited to that challenge. A new report from the bipartisan National Security Commission on Artificial Intelligence contains reason for cautious optimism on that latter question, even as it reminds us that an authoritarian China will be a formidable competitor indeed.

Something you'll hear a lot is that the increasing availability of compute resources has paved the way for important advances in artificial intelligence. With access to powerful cloud computing platforms, AI researchers have been able to train larger neural networks in shorter timespans. This has enabled AI to make inroads in many fields such as computer vision, speech recognition, and natural language processing. But what you'll hear less is the darker implications of the current direction of AI research. Currently, advances in AI is mostly tied to scaling deep learning models and creating neural networks with more layers and parameters.

AI has the power to search through millions of genetic variants at high speed and identify likely ... [ ] causes of rare diseases, while also comparing what they find with the existing medical literature. This is greater than the population of the United States, yet the ominous figures don't end there. According to the Global Genes organization, eight out of ten rare diseases are caused by a faulty gene, yet it takes an average of 4.8 years to arrive at an accurate diagnosis. This is part of the reason why 30% of children with a rare disease won't live to see their fifth birthday. Neither is this situation helped by the fact that 95% of rare diseases lack an FDA-approved treatment.

As I've been focusing more and more on the Big Data and Machine Learning ecosystem, I've found Azure Databricks to be an elegant, powerful and intuitive part of the Azure Data offerings. Over my last 12 months at Slalom, I have had the incredible opportunity to travel across Canada and work hand in hand with the brilliant folks at Microsoft's Data & AI practice and Databricks experts to lead project engagements, deliver technical hands-on workshops, listen to the industry experts - the folks doing Data Science for a full time living - and absorb everything in between. There's a common theme across the industry verticals that's going to be our point of discussion today. The hot topic of 21st century tech is Machine Learning - some flavor of AI/ML is thrown into almost everything we find these days (I'm pretty sure I spotted a "genius" AI/ML toothbrush at Shoppers Drug Mart today). The reality is, the mathematical techniques that power Machine Learning models have been around for almost a century.

Rest of Europe 8.4 Asia Pacific 8.4.1