Highlights from our Upgrade 2020 Virtual Summit

By Kazuhiro Gomi

With more than 1,100 registrants, 17 speakers on Day 1 and 28 technical presentations over the next three days, the Upgrade 2020 NTT Research Summit was a great success. (And it continues to be, as the recorded talks remain available online here.)

As President and CEO of NTT Research, I had the pleasure of welcoming our virtual attendees and participants on Sept. 28. Our busy Day 1 included remarks from NTT Corp. President and CEO, Jun Sawada (and his remote control avatar robot!), and NTT Corp. EVP and Head of R&D Planning, Dr. Katsuhiko Kawazoe, who is also chairman of the board of the Innovative Optical Wireless Network (IOWN) Global Forum, which was featured in a closing presentation that day. Also speaking were executives from NTT DATA Services and NTT Ltd. 

The main keynotes on Sept. 28 were delivered by two academic luminaries: Shafi Goldwasser, Director of the Simons Institute at UC Berkeley, and Peter Fitzgerald of the Stanford University School of Medicine. One of the world’s leading computer scientists, Professor Goldwasser summarized results from three recent papers demonstrating how the burgeoning field of machine learning (ML) can become more trustworthy with respect to privacy, robustness and verifiability by applying advanced cryptography technologies. A professor, researcher and innovator, Dr. Fitzgerald gave an impressive tour d’horizon of the digital healthcare landscape. 

Day 1 also featured leading scientists from our PHI, CIS and MEI Labs, on a panel moderated by our event’s emcee, theoretical physicist and popular author Dr. Sean Carroll. For a solid overview of these labs, which are the backbone of NTT Research, you could not do better than listen to this discussion

If you are at all familiar with our PHI Lab, you know that the Coherent Ising Machine (CIM) plays a key role in its research agenda. On this opening panel, PHI Lab Distinguished Scientist Robert Byer explained what a CIM actually is. A chaired professor of Applied Physics at Stanford University who has made many contributions to laser science, Dr. Byer reminded us that a CIM is a special kind of computer that solves special problems, capitalizing on amazing scientific advancements, especially in the field of optics, but also quantum physics and neuroscience. To get to the next stage of a workable computer, the PHI Lab is working with 25 external collaborators and 20 internal employees. 

How is cryptography related to quantum? Brent Waters, Distinguished Scientist in the CIS Lab, answered that question by describing the two fields as “frenemies.” On the one hand, a real quantum computer, when it is someday deployed, could break standard public key cryptosystems. Yet, that is only one of many types of potential threats, some of which classical computers could launch. Dr. Waters also noted that innovations, such as lattice-based cryptography, could stand secure against quantum attacks. There’s also the positive side of quantum, such as the new forms of money that it could enable. But in this and other areas, he reminded us that what the CIS Lab does is basic research, i.e. building core tools, which require a second pass to turn into applications.

The MEI Lab could leverage advances in cryptography and next-gen computing. In terms of its long-term goal of developing a bio digital twin, Distinguished Scientist, Joe Alexander, said the more data, the better the model. Information populating these models, moreover, will need to be delivered and processed as fast as possible. At the same time, data will not be forthcoming without privacy protections. A bio digital twin is ultimately a personalized model. Digital populations could also enable aggregated results, for instance, through virtual clinical trials. Dr. Alexander is an expert at conveying both our long-term vision and the steps needed to achieve the MEI Labs’ first major goal of creating a cardiovascular bio digital twin.

The next three days of Upgrade Reality included nearly 30 presentations. These were short talks, only about 15-20 minutes each, but technically deep. I can only share some high-level takeaways here. 

The PHI Lab day, Sept. 29, featured talks from several of our academic partners. As noted, the PHI Lab has 25 external collaborators. The latest of nine joint research agreements in this field of study was announced in September 2020 with the University of Notre Dame, and we were pleased to have the principal investigator in that project, Dr. Zoltán Toroczkai, join us to discuss some of his research. Other academics from Stanford, UC Berkeley, MIT, CalTech, University of Michigan and University of Tokyo also presented. The PHI Lab day of the summit also featured an additional 11 posters, demonstrating the breadth of active research in this field.

The nine presentations on the CIS Lab day, Sept. 30, included three from NTT Research scientists: Hoeteck Wee, Daniel Wichs and Ilan Komargodski. One presentation by Amit Sahai (UCLA) elaborated on a breakthrough paper on the topic of Indistinguishability Obfuscation (iO), written as a result of two years of collaboration with his co-authors, Huijia (Rachel) Lin (University of Washington) and Aajush Jain (graduate student researcher at UCLA’s Center for Encrypted Functionalities and former research intern at the CIS Lab). The paper solves a 20-year problem by constructing iO from “well-established assumptions.” 

On the final MEI Lab day, Oct. 1, attendees could hear more details about the cardiovascular bio digital twin from Dr. Alexander. Also presenting was Dr. Tetsuhiko Teshima, who spoke on “Miniaturized System for Cell Handling and Analysis.” Dr. Teshima is a research scientist at NTT Research engaged in joint research work with the Technical University of Munich, where he is working in the area of advanced neuroelectronics and biosensor technology. Advanced biosensors are key building blocks in the construction of a bio digital twin. Other speakers represented the University of Leicester, Keio University, NTT Basic Research Labs, Stanford and the Japanese corporation Marubeni.

Thanks to all who participated in, attended and supported this great feast for the mind. Please share relevant links with friends and colleagues who may have overlapping areas of interest, and please consider returning, whether in-person or online, when we reconvene next year.