EmTech MIT 2021 – Day 2: Leading with innovation, Virtual MIT Technology Review’s flagship event, September 28-30, 2021 | Leaders in Pharmaceutical Business Intelligence (LPBI) Group

Artificial intelligence’s superhuman data processing capabilities have far-reaching implications. Unpack the oncoming effects on industry and society and explore the issues of ethics, inequity, and more that we’ll face as we strive to maintain control of algorithms that are controlling us.

11:35 a.m.Building a Better AI

To date, every major milestone in AI has been achieved by deep learning, but whether this approach will lead us to artificial general intelligence remains to be seen. To have true AI, must we first understand the brain? This session explores a neuroscience-based approach to AI that may lead to true machine intelligence.

Jeff Hawkins Cofounder and Chief Scientist, Numenta

>>> Model of the World, learning through movement, Current AI is pretty dump, machines that are truly intelligent, Road map for AI, his Thousand Brains Book, Not build the Brain, Two elements of new modeling: (1) fastest NN (2) Sparsity – existing NeuroNets and make them xxx times faster, learn new information Integration of these theoretical concept into companies building products. Add the thousands Brain model concept into AI to get Intelligent systems. The concept of competition, Structure of the Brain is known speculation of building AI to be human’s like in terms of Intelligence. The Goal is not Human’s Like bur Intelligent machine will act in the World, not have to be like Humans. Computing models can be smart not human but helping humans in Pattern recognition. Humans is the only species knowing about time, the universe, preserve and propagate, peer reviewed was also censored,

AI as a Global Disruptor

>>> Image processing and Computer Vision, finished PhD. Negative impact of AI, animate tools used for can be socially Bias. While at Google questioning AI experiences isolation. Labor rights and anti-discrimination rights, censoring research sounds like propaganda, outside the Tech company to impact the industry, to keep Tech companies accountable from the inside silencing. labor protection laws needed. Create own institute AI technologies to be built, critiquing AI. Ideas coming from the Researchers. AI to benefit Humanity vs Tech companies Profiting from Researchers work – Labor laws. Apples treatment of workers oppressive. Coalition of people around you: Scholars. Look at the marginalized group vs dominate group. Technology is built by both.

Virginia Smith Carnegie Mellon UniversityAssistant Professor

>>> Federated Learning beneficial – Distributed AI Learning Model. Where does the Data comes from Two canonical approaches in Federated Design: Centralized Learning vs Federated Learning – training at the Edge. Federate Learning (FL): (1) Cross Silo [high availability] vs (2) Cross device Federate Learning [communication bottleneck] – Distributive model [Privacy] different devices interoperability [heterogeniety}. Personalized FT: One for each device and among the devices improves Accuracy. Bias in Data. Privacy is Key. Google deploys Federated Design. FT design for Autonomous Vehicles construction behavior, rad conditions AI – work to be done for deployment. FT is not production ready. Centralized learning vs Federated Learning. Hospitals: FT is more expansive with greater benefits.

Challenge: Privacy [Criptographic technique] and secure the learned Model

Ani Kembhavi Allen Institute for AI Research Manager

Single Purpose Model vs Multi Purpose Model (1) General Models (2) Understand (3) Question and get answers. Learning Skills, Learning Concepts, Search Engine Data to teach after learning skills and concepts using richer vocabularies using Visual and Text data. Learn to interact, navigate. A game of Cache: Hide & Seek. AI agents, CEREBRA – Cognitive Rudiments for building AI Models – multi skills models simulators, robots, based on physical principles Multi modal AI Visual, Audio, Text. Building models that understand TEMPORAL behavior.

1:15 PM EDT

MIT INSIDE TRACK Programming DNA
1:15 PM – 1:40 PM EDT
Charlotte Jee MIT Technology Review Reporter, News
Ron Weiss MIT Professor

>>> Synthetic Biology – complexity biology: Pathways and Disease state (1) Sensing (2) Logic Processing (3) Therapeutics development. Controlling Stem cell differentiation. Programming a cell development for drug development via organelle development for building organelles for replacement: Liver vascularization dysfunction, Pancreas function by design mature Organoids Cyp3A4 – for druggability. Cancer immunotherapy will be first to benefit numeric synthetic Biology for therapeutic intervention to improve precision.

2:00 p.m.Can We Trust Tech to Police AI?

We are all subject to AI, even if it’s faulty, beyond our control, and biased. Massive AI models are being developed, but how do we ensure fair systems are created? Whether you’re building your own AI or working with vendors, learn the essential elements of fair and equitable AI.

Timnit Gebru Cofounder, Black in AI; Formerly, Google

12:25 p.m.AI Learning Models: Distributed vs Centralized

More secure methods of processing and storing massive volumes of data for AI are needed to alleviate privacy concerns. Is federated learning the best option? Examine how distributed learning works for AI and the potential benefits and risks for your organization.

Virginia Smith Assistant Professor, Carnegie Mellon University

12:45 p.m.New Advances in Multi-Skilled AI

AI robotics still struggle to match the skill level of a child. Human intelligence emerges from our combination of senses and language abilities; the same might be true for artificial intelligence. Is combining vision, audio, and language processing into a single AI system possible—and will it solve the problem? Explore the implications for AI and its potential use cases. As featured in the 10 Breakthrough Technologies 2021.

Ani Kembhavi Research Manager, Allen Institute for AI

MIT Inside Track (1:00 p.m. – 1:40 p.m.)

Join the Inside Track sessions to engage more deeply with our content, speakers, and your fellow attendees during mainstage programming breaks.

1:00 p.m.Networking Break

1:15 p.m.Programming DNA

Today with synthetic biology, we can assemble DNA for the purpose of modifying individual cells. But we’ve arrived at an age where it’s now possible to write DNA programs, analogous to writing to software, that impact cell information and behavior in a given sequence for even greater impact. We’ll talk about possibilities, limitations, and timelines for this amazing advancement.

Ron Weiss Professorr, Massachusetts Institute of Technology

AI as a Global Disruptor Cont’d (1:40 p.m. – 2:05 p.m.)

1:40 p.m.Trustworthy AI in High-Stakes Environments – Presented by Raytheon Intelligence & Space

The high-stakes environment of intelligence and space is overflowing with data, where signals lurk in a sea of noise, best discoverable with AI. With no tolerance for error, engineering teams must drive trust and explainability in AI decisions so that human-machine teams, working in areas from synthetic biology to next-generation GPS, find the right solutions, every time, at speed.

Roy Azevedo President, Raytheon Intelligence & Space

Elizabeth Bramson-Boudreau CEO and Publisher, MIT Technology Review

>>> High-stake environments: National Security, Logistics, Cyber Satellite, Weather Storms, Deploy AI to explain the recommendation for trust into the decision making. AI is distrusted as it nears autonomy, edge or situation is unique, AI-ML recommendation require explainability for system automomy. Knowhow Satellite system for Weather prediction system data run through scenarios ethics applied before using AI-ML algorithm latency is not affordable, operator make decision at the edge Trust & Verify Modeling & Simulation perform.Education and Training 37,000 employees 5,000 were hired during the Pandemic. Three innovation new engine work, radio frequencies Radars, cyber technology

Biotech and Biothreats (2:05 p.m. – 3:20 p.m.)

Pandemic shutdowns focused a spotlight on biotech. Glimpse advances from drug development to gene editing, and which ones will help us through the next crisis.

2:05 p.m.Mitigating the Impact of Biological Threats

Our global society has become hyper-aware of biological pathogens and threats. We investigate some of the current and recent threat types, dispelling myths and confirming facts, while considering what comes next in the way of prevention, detection, and response for the next bio threat on the horizon.

Christina Rudzinski Assistant Division Head, Lincoln Laboratory, MIT

>>> Reducing Biological threats. COVID19 is a Global Pandemic by a pathogen, diagnostics deployed, genome of the virus gaps remain. Future pathogen will cause infections. Novel to genetically engineered pathogens, infectious agents, pathones evolve, can be used and have been used maliciously Infection progression: Pre-exposure Human transmission incubation symptoms onset illness early environmental detection population surveilence Bio-signal data for detection of host’s response to infection. Priority is both detection in advance and the vaccine capability virus detected as pathogen Active biological weapons Nation states as adversary Lab escape virus is a possibility Lab Survelience systems.

2:30 p.m.Turning CRISPR off

Genetic therapeutics has advanced to the point where we can turn genes off and on without altering DNA. Innovations like CRISPRoff affect cutting-edge research on viruses and the fight against diseases and other genetic disorders. Explore the possibilities and questions on how to manage ethical concerns and unintended consequences.

Jonathan Weissman Professor, MIT; Investigator, Howard Hughes Medical Institute

>>> CRISPR 2.0 under DARPA Chemical and BioChem funding. CRISPR gene editing correct the underlining genetics, where to cut the DNA for changing the sequence Cas9 – complicated technology. Turn up and down -Silence a gene, an existing gene programmable Epigenetic memory engineering (15 month) new opportunity in Medicine CRISP off Variant 1 vs CRISPR off Variant 2 only the targeted gene precision editing. Memorizing Gene Silencing. Silencing then reverse not permanent silencing. Germ line engineering. Motorneuron disease are good indications for Memorizing Gene Silencing

Innovators Under 35 (2:50 p.m. – 3:20 p.m.)

2:50 p.m.Scale Innovation and Ideas – Presented by JPMorgan Chase

Innovation drives results and creates value, and it must be supported by vision, leadership, purpose, and a clear path to scale. The right ecosystem is essential. From accelerator labs to automation, dynamic process transformation is built on collective intelligence put toward a common objective: to achieve real digital transformation for employees, customers, partners, and suppliers alike.

Lori Beer Global CIO, JPMorgan Chase

2:55 p.m.Next-Generation Disease Detection

New advances in CRISPR platforms are pushing biotechnology to the next level of disease detection and treatment. What does that mean for reducing human intervention, increasing diagnostics, and scalability? As featured in the 2021 Innovators Under 35.

Janice Chen Cofounder & CTO, Mammoth Biosciences

>>> On Demand DIagnostic Tool based on CRISPR: read, detect, Protein Discovery – Metagenomics – proteins Cas14, delivery advantages. Target proteins for diagnostics: detect DNA and RNA the exact sequence Future of CRISPR Diagnostics delivered to Mobile. Molecular lab accuracy in the mobile device for the results TEST to TREATMENT. CRISPR is a platform for Diagnostics it is also a therapeutics target via gene editing vs medicinal chemistry.

3:10 p.m.Microscopic Robots that Move

Programmable, autonomous, microscopic robots are coming, and now they can move. These tiny bots have the potential to revolutionize engineering new materials, rid crops of pests, act as cellular-level surgeons, and more. Get an early look at this emerging tech. As featured in the 2021 Innovators Under 35.

Marc Miskin Assistant Professor, University of Pennsylvania

>>> application of microelectronic CMOS for design of robots microorganism size not visible to eye laser spot is a control function parallel design of one robot allows deployment of an army of robots like microorganisms Repair of nerves. A factor of 10 in size power low, cost low,

MIT Inside Track (3:20 p.m. – 3:55 p.m.)

Join the Inside Track sessions to engage more deeply with our content, speakers, and your fellow attendees during mainstage programming breaks.

3:20 p.m.Networking Break

3:30 p.m.From a Mouse to a Bird

The ubiquitous computer mouse allows us to manipulate our 2D desktops. The jump to virtual worlds will require new tools to help us interact with 3D objects. Get an early look at the grasping technology called Bird.

Aubrey Simonson Graduate Student, MIT Media Lab

Andy Lippman Associate Director, MIT Media Lab

Extending the Workplace (3:55 p.m. – 5:15 p.m.)

The events of 2020 forced us to reimagine the workplace. Explore extended reality (XR) technology projects, including augmented reality (AR), virtual reality (VR), and mixed reality (MR), poised to change the way we work.

3:55 p.m.Planning for an Immersive Workplace

Extended reality (XR) technologies are transforming workspaces. Bridging the virtual and physical worlds with AR/VR/MR enables immersive experiences for team collaboration, staff training, and customer experience. Get industry insight on how these technologies are impacting work and generating surprising results.

Timoni West Vice President, XR Tools, Unity

4:20 p.m.Preparing for a New Reality

Immersion technologies are estimated to become a $57 billion industry by 2027. Get a look under the hood at how these systems work, what kind of infrastructure is needed, and tips for integrating immersion tech into existing workflow systems.

Urho Konttori Cofounder and CTO, Varjo

>>> VR and XR headsets: True Telepresence, device will be size of goggles. In 2025 standardization of the Industry.Human communication iwll be come Teleportation

4:40 p.m.Extended Reality Use Cases

Extended reality tools are already being incorporated into standard business operations in enterprise-level systems. Industry experts walkthrough real-world applications in health care, training, and construction to demonstrate the value immersion technologies can bring to the workplace.

Amitai Ziv Director, Extended Reality Hub, Sheba’s Innovation Center

>>>> Patients education with VR, Chemo Treatment VR can take patients to Paris or London. Used in Surgery CT overlaid on the Spine surgery site. Moving organ during surgery Heart, abdomen. Expectation. Assist device, not the only one. Simulation is not reality.

Will Adams Emerging Technologies Developer, M.A. Mortenson Company

>>> Rendering images of construction sites, technology is cool

Gordon Cooke Director, Research and Strategy, The U.S. Military Academy at West Point

>>> Training, new content creation is an issue.

Last Call with the Editors (5:15 p.m. – 6:20 p.m.)

Attendees and speakers are invited to attend our online EmTech MIT reception in virtual reality, which kicks off with a short talk on creating and navigating immersive spaces. Join us under the MIT Dome from your web browser or your VR headset as we mix and mingle in this unique networking session and hands on demo.

Before we send everyone home for the night, join our last call with all of our editors to get their analysis on the day’s topics, themes, and guests.

5:20 p.m.EmTech MIT VR Reception Under the MIT Dome