17th Annual EmTech @ Media Lab, MIT – November 7 – 8, 2017, Cambridge, MA – This Year’s Themes, Speakers and Agenda
MIT Media Lab
Building E14
75 Amherst Street (Corner of Ames and Amherst)
Themes:
- Business Impact
- Connectivity
- Intelligent Machines
- Rewriting Life
- Sustainable Energy
- Meet the Innovators Under 35
Leaders in Pharmaceutical Business Intelligence (LPBI) Group, Boston
will cover in REAL TIME
The 17th annual EmTech MIT – A Place of Inspiration, November 7 – 8, 2017, Cambridge, MA
MIT Technology Review’s EmTech conference
In attendance, covering LIVE using Social Media
Aviva Lev-Ari, PhD, RN
Editor-in-Chief
http://pharmaceuticalintelligence.com
@pharma_BI
@AVIVA1950
#emtechmit
@techreview
https://events.technologyreview.com/emtech/17/?utm_medium=email&utm_source=press_list&utm_campaign=emtech2017&utm_term=conference&utm_content=press_credentials&discount=MEDIAM172B#section-about
AGENDA FOR TUESDAY, NOVEMBER 7, 2017
-
8:00Registration & Breakfast
- In Media Lab – MIT and MIT Technology Review was established in 1899
- EmTech 1999 – 100 years to MIT Technology Review
- Innovations and pushing the boundaries
- AI – potential and limitations
- Climate change requires new technologies
- Brain Technologies: Biology Vision
- Tomorrow: emerging technologies: Cybercrime, role of technology
- Automation and future of work
- Partners: GE, Lamburghini
- Lemelson-MIT
- MITTR – Whova on AppleStore
- Trends in AI – AI is the new Electricity
- Deep Learning & Neural Networks (NN):
- Input a picture –>> output: Is it You?
- loan application outcome: will you repay (%)
- picture from car – Output GPS address –>> Supervised Learning
- doing act in <1 sec of thinking
- training SMALL, Medium size very large NN
- Algorithm innovations:
Supervised Learning algorithm types:
- Transfer Learning
- Unsupervised learning
- Reinforcement learning – hunger for data: i.e., robotic applications
Importance of Data accumulation for launch a Product – Users — data growth
- Shopping Mall + website is not equal an Internet company
- Internet company:
- push data to CEOs
- A–B Testing
- Short cycle time
- Decisions made by PM and ERP
AI era
traditional company + NN not equal AI company
- Strategic data acquisition
- Unified data warehouse
- NEW JOB DESCRIPTIONS
- Precision automation
- ORGANIZATION CHART to interface in a matrix with AI Teams – hire Ai in the Business Units
- Scarce talent of AI
Discussion
- Children MUST learn to code
- Human-Computer communication will be by writing code
- Future of work
- warranted reliant digital connectivity
Tomasso Poggio, MIT, CSAIL, BCS
- Deep learning – next step
- Bet on Center Brain Mind Machines (CBMM)
- Josh Tennenbaum at MIT
- Autonomous Driving – Amnon Shaashua, MobilEye
- 20 years @ MI AI: Dailmer and MIT — detection of pedestrans
- Powerful computers and algorithms – Reinforcement Learning Networks (Brain Science), models of Vision and Deep Learning Networks – WHEN they work?
- Building Jarvis – a buttler application in AI built by Marc Zukenberg
- NeuroScience – MobilEye, AlphaGo
- CBMM – NSF $50 Million in AI funding – Science of Intelligence and Engineering of Intelligence
- MIT & Harvard plus several organization
- Business Partners: MS Soft, Google bought MobilEye,
- Center for Visual Gaze – 200 msec of visual processing
- ERGO SUM: toward symbols, Cognitive core, visual system, Brain OS – running routines
- Breakthroughs: Theory: under which conditions,
- Learning theory
- optimization Approximation Theory: Deep vs Shallow networks
- Intelligence is greatest problem to solve it is like LIFE, Tomasso Poggio, MIT, CSAIL, BCS
- machine can help human to think better, long time horizon is needed,
Kris Hammond, Prof. Northtwestern University
- Data analytics and Ideas
- words vs language – past, present, future – uniquely HUMAN, now machine language is Human Partner
- language vs Ideas
- machines knows a lot
- facts, dat move to narrative
- Language is understanding
- FIN information: Decisions about allocations,
- Turbidity data on the beach in Chicago: Which Beach is the cleanest vs the dirtiest
- NARRATIVE ANALYTICS: data that machine can tell us what it has as a story and presented as intelligent language,
Cognitive Science application to autonomous driving – Yibio Zhau, Tennenbaum Lab @MIT, ISEE.AI, Computer vision, Cognitive Science
- interpulate and extrapulate data needed for autonomous driving
- reasoning beyond the system: Human intelligence , intentional reasoning, pattern recognition,
- Ali baba – funding building of a Robot for autonomous driving – understanding by imagining – causes for behavior by others
- ISEE – Next generation of AI — driving drivessless ly for thousands of miles
- Car to car communication is a sensoring issue, negotiation need to be taught to machines
Young Scientists 35 years old or less
Austin Olson, Luminar – object detection 99% accuracy,
Angella Schoellig — Roborts, Prof. University of Toronto, robots in predictable environments
Lorenz Meier — Vertical Technologies – Drones and safety – DB of flights
Lee Krevat, Sempra –
owns Wind Farms- managing a Grid with renewable energy. Variable – Wind technology wind is variable – if wind blows too much switch to diesel. 100% renewable for one hour on Islands
- 20 cents diesel, wind is 10 cents help the enviroment
mainland, not yet used, price diesel vs wind
Solar wind generation – next biggest Technology in Energy
Alex Tepper,Avetars
- RAIL – Predict derailments
- OIL & GAS – corrosion is the enemy — knowledge of corosion progression – using AI algorithms
Growth area: Aviation
John Holdren – Harvard University – Government Role in ENERGY and Climate Change – Obama’s advisor Presidential CSO on Climate and energy
- mitigation
- adaptation
- suffering – shortcomings of mitigation and adaptation
- harm of business as usual
- Efficiency standards during Obama Administration, assistance to other countries led to the Agreement in Paris 195 countirs — agreement to reduce emission. China and US declare cooperation on emission of gases into the environment.
- PRESIDENT TRUMPS CALLED CLIMATE CHANGE A HOAX – proposed to cut energy R&D
- All executive orders by Obama – were reverted by Trump
- Innovations: Electricity from Solar increase and wind as well and batteries
- Carbon capture and storage – technological challenge
- Biofuel processing, liquid bio fuel
- Nuclear innovations to nuclear waste
- 2100 – 5% on defense and 2% on the environment – model under estimate the contribution of innovations for the long run.
- 1000 businesses in deployment of technologies
Evelyn wang, MIT – Material Science – Sustainable energy – nano
- material properties: superior properties of LOW DENSITIES
- Light manipulation
- membrane
- CO2 capture
- Technologies: Nnao, Thermoelectronics, energy and water
- Solar 6% and wind 21%, biomass 5%voltaic
- SOlar eneconversion
- PHOVOLTAIC: SCALBALE, SOLID STATE, INTERMITTENT, PARTIAL SOLAR SPECTRUMrgy
- Nanophotonics: Solar energy conversion: photo
- Nano absorber – area ratio; Emitter: silicon and silicon – spectral approach
- potential STPVs
- Transportation using energy with emission
- Power consumed by HVAC
- Thermal Battery for Electric Vehicle: Adsorption Heating and Cooling
- Desorption vs Adsoption: cooling vs Heating mode
- High capacity adsobents – Zeolite MOF enhancing capacity heat and mass transport
- Tmal Battery Prototype: Hybrid, electric, stationary domestic HVAC.
- Water harvesting from Air – metal organic Frameworks: Adsorption – harvest water without need of additional electricity
- Opportunities for Advanced Materials
Prof. David Keith, Harvard University
- technologies to stop global changing
- research program
- stratospheric aerosol cool planet – pollution masking global warming
- solar geo-engineering, vs emission cut 3x BAU vs business as usual
- Annual maximal Temperatures, extreme precipitation,
- carbon emission worm up vs climate risk in Time
- use of technology for climate change mitigation: carbon removal
- Solar engineering is the solution
Next Generation Brain Interfaces
Andrew Schwartz, University of Pittsburg
- Causality is obscure
AGENDA FOR WEDNESDAY, NOVEMBER 8, 2017
-
8:00Registration & Breakfast
Daniela Rus, CSAIL, MIT – Robots: drones, 3D Printing
hosted by David Rotman, MIT TR
- supply chain and transportation – city will benefit from a different business model
- autonomous driving deployed in Singapore
- all vehicles on wheels can be made autonomous
- blind – camera on a belt assists in navigation
- ML: Patterns and predictions
- AI – reasoning
- robots: motion
- Machine read entire libraries
- Radiology: Read by machines vs by Radiology: AI + Human — 0.5% error
- Rural area medicine
- Machines – Better Lawyers: NLP – read precedents to cases, machines can’t write a briefing or defend a plaintif
- Factory and Automation: Robots roles – enable mass OPTIMIZATION not only mass production
- Machines do not have common sense and do not have ability to reason
- crunching data vs analysis
- JOB Categories:Tasks vs Professions: Routine data processing and labor task — are ready for automation
- NEW jobs: User experience designer, GPS enable taxi drivers to drive and drove pay scale down
- GDP – decreased 1966 – 2016
- KY school to train coal miners to do data processing to become CODERS
- JFK – new machines brings man back to jos – new jobs
- AI supports NEW jobs: CS/AI part of literacy
- people and machines – in collaboration
discussion
- Who to make the transition?
- CODING is key – people must be active in keeping up and continue to train
- make it easy to make machines, interactions Man-Machine easier,
- YOU ARE WRONG SIGNAL IS recognized by EEG
- AI and Future of Work Conference at MIT – anxiety related to job changing due to technology
- Technology can’t solve all problems, Technology helps, Technology implications on Policy – technology as a unifier societal force not a dividers
- Transportation as Utility
9:30 – 10:00 AI and the Future of Work
Iyad Rahwan, MIT Media Lab, Introduction by Elizabeth Woyke, TR
- Physical Therapist — will not be replaced by computerisation
- Probability of computerisation: Skilled cities are better at economics shocks
- Adam Smith – simple operations
- Differential Impact from Automation on Cities – the larger the city more resilient to automation
- City size vs clusters of occupations — cluster grow with city size
- Impact on Middle Class vs Lower and Upper: low paying jobs, middle and high
- Skills in Occupations: mapping SkillScape correlations with Education
- Skills in demand
discussion
- Urbanization took place – 80% live in cities around the World
- Outliers in CIties by size and Skills: Boulder, CO – small size very skilled labor, politics support start ups and high tech
10AM – 10:30AM
Meet the Innovators Under 35
- Tracy Chou – ProjectInclude – diversity
- All about data
- Olga Russakovsky – Princeton University – Computer vision
- AI for education of under privileged high school
- IM-GENET – Data sets encode human biases
- AI is powered by Data
- AI learns societal biases
- Researchers shape AI
-
10:30
Break & Networking
- 11:00 – 11:30 What is Social Media Doing to Society?
- 300 million reach of Ads posted by Google in the Internet
- Fake news
- Network shape
- Veracity and popularity personalized
- e 11:30 – 11:45 Meet the Innovators under 35
- Phillipa Gill – UMass CS – Project of Network measurement on censorship measurement platforms
- Joshua Browder – DoNotPay
11:45 – 12:00 The Emerging Threat of Cybercriminal AI
Shuman Ghosemajumder, Shape Security
Hosted by Martin Gile
- CyberCrime is evolving using AI – Imitation Game – Turing Test restricted Turing Tests
- Computer vision, Solving CAPUTRE – Copletely Automated Turing Tests
- CAPTCHA by Google
- Credential Stuffing Accounts Attacks – SONY was hacked and 93,000 Passwords stolen
- Clip Farms at Google
- BLACKFISH – identify Credential Stuffing Accounts Attacks, all invalid password are not valid to be used by cyber attackers again – that authentication is no longer valid
- Multi Factors Authentications vs ease of use to Log In
- Knowledge Basis – Probabilistic SYmbols – BlackFISH – technological advantage – iPhone stores a math formulation of characteristics of the finger print not the image of the fingure
-
12:00Lunch & Networking – Lamborghini -super sport car
- Killer Platforme ==>Killer Interaction ==>Killer application
- Reactive ==> Proactive
- Brain Computer Interaction (BCI) – maximum Privacy no voice involved like in SPeech
- Voice, Motion Tracking, eye tracking
- Human intentionality – a World without limitations
-
NASA is a client
- consol technology for navigation, typing,
- Problems: Add to glasses or as an Ear piece
- the signal is ACTION POTENTIAL
- latency differences between individuals
- Non-invasive to invasive to capture signals
- Using functional MRI technology for a NEW device to scan emotions rather than medical diseases
- HOLOGRAPHY of the Brain – liquid crystal display is like transistors on a chip
- OPTICS – DISCONTINUITY of Moore’s law – high resolution like functional MRI
- Holographic LCD – scattering material VOXEL detector – measure intensity of light, no resolution, consumer camera speed OK Inexpensive
- Human body scattering
- HAT and Bandage
- Tell the Truth
- Media Lab — a Non-disciplinary place
- Universities play a role in Social Justice
- FEAR of AI:
- For profit will own it all
- stupid AI will govern
- displace work
- espionage
- catalytic institute that will make a contribution to OPENNESS vs technological dominance
Joi Ito, MIT Media Lab: AI problems –
- MUST be democratized – Now it is in the hands of very FEW
- RISK SCORES can’t be contested in court because they are IP of for profit companies
- Joi Ito, MIT Media Lab at MIT do good to Society vs make the most of money which the majority are doing
- AUTONOMICH vs autonomous agents, said Joi Ito, MIT Media Lab – Hoffman: Design goals more symbiotic: Scaling, more productive, Season 2 launched today
- Design principle – LEARNING vs EDUCATION, Joi Ito, MIT Media Lab
Hoffman on AI Technologies
- shaping it to avoid catastrophic negatives
- provide a public good via participation
- Tweets and News, Washington Post – Tracking tweets from US on Politics related to the Elections
- National memory on Guns, Immigrations
- Debate brief from tweets and News rooms
- topic classifier, Campaign finance, SHARE OF COVERAGE IN NEWS, SHARED OF VOICE ON TWITTER
- deep neural network training algorithms
- Passion Gap: cut data on Twitter – Trump supporters exhibited x2 fold energy vs the Democratic candidate
- How does Media flow: Sanders, Clinton, Trump – each is a Media Source
- Truth, Trust, Attention – Fact checking
- If Trust the source then I believe it is True
- Public Opinion: The Politics of Resentment in Rural WI – Katherine Cramer
- Listening Networks: Human- Human Interaction: Media sharing network – change week by week – the MOST innovative methodology developed to date for Public Opinion – presentation by
Deb Roy, MIT media Lab – using deep Neural network training
- main stream
- conservative
- liberal activist
- Health Indicators:
- Shared attention
- Shared Reality
- Varied Perspective – surface under-heard voices
- Valve – development of Tissues, biochemical properties
- signaling molecules
- mechanical strength – physiological
- Attrach stem cells – proper matrix formation
- Functional implants
- Cancer Precision medicine – Liquid biopsy – tumor mutations
- entire Cancer Genome – from blood biopsy
- Scaling: Broad Institute 100 collaborators – 3,000 blood sample genomical analysis
- Easy Taxi
- Fighting inequality
- 15Billion – Beauty Market
- IBM Research Africa – 300 million adults – lack of access to financial services
- Univesities, Government – start ups to scale ideas
-
5:002017 Innovator Under 35 Awards & Reception
- Speakers
-
-
Viktor
AdalsteinssonGroup Leader, Broad Institute of MIT and Harvard
2017 Innovator Under 35
Gene
BerdichevskyCEO, Sila Nano
2017 Innovator Under 35
- rechargeable battery
- new class of materials charge and discharge in battery
- store more energy
- more better designed electronics: electrified flight, solar, car: Hybrid and electric
- 21st Century belongs to electrification vs combustion in the 20th century,
Gene
BerdichevskyCEO, Sila Nano
-
Tracy
ChouFounding Advisor, Project Include
2017 Innovator Under 35
-
Adrienne
FeltSoftware Engineer, Google
2017 Innovator Under 35
-
Phillipa
GillAssistant Professor, University of Massachusetts, Amherst
2017 Innovator Under 35
-
Tallis
GomesCEO, Singu
2017 Innovator Under 35
Kathy
GongCEO, WafaGames
2017 Innovator Under 35
- GAMING SWARD OF GLORY – EPIC NEW RTS EXPERIENCE – WAFA GAMES IN CHINA
-
Ian
GoodfellowStaff Research Scientist, Google Brain, development occurred at OpenAI
GAN’s – Generative Adversarial Network – from AI Optimization to Game Theory
2017 Innovator Under 35
-
Yasmin
GreenDirector of Research and Development, Jigsaw at Google
Addressing Online Threats to Global Security
-
Kris
HammondChief Scientist and Cofounder, Narrative Science
AI’s Language Problem
-
Svenja
HindererScientist, Fraunhofer IGB
2017 Innovator Under 35
-
Reid
HoffmanCofounder, LinkedIn; Partner, Greylock Partners
The Future of Work
-
John
HoldrenProfessor, Harvard University
Climate Disruption: Technical Approaches to Mitigation and Adaptation
-
Joi
ItoDirector, MIT Media Lab
The Future of Work
-
Mary Lou
JepsenFounder, Openwater
Capturing Our Imagination: The Evolution of Brain-Machine Interfaces
-
David
KeithProfessor, Harvard University; Founder, Carbon Engineering
The Growing Case for Geoengineering
-
Neha
NarkhedeCofounder and CTO, Confluent
2017 Innovator Under 35
-
Andrew
NgFounder, Deeplearning.ai; Adjunct Professor, Stanford University
The State of AI
-
Tomaso
PoggioInvestigator, McGovern Institute; Eugene McDermott Professor, Brain and Cognitive Sciences, MIT
Understanding Intelligence
-
Olga
RussakovskyAssistant Professor, Princeton University
2017 Innovator Under 35
Michael
SalibaMarie Curie Fellow, EPFL
2017 Innovator Under 35
- disruptive technology in the energy space
-
Gang
WangChief Scientist, Alibaba
2017 Innovator Under 35
-
Jianxiong
XiaoChief Executive Officer, AutoX, Inc.
2017 Innovator Under 35
CAMERA-first solution affordable self-driving
-
Leave a Reply