L4D Learning for Democracy

 

https://www.youtube.com/watch?v=p8y01XpbKgE&index=4&list=PLpoIPNlF8P2PFPZFhYVaUsZrlxcQr6Bhx

 

 

 

etc etc for the full report click the link at the bottom "download this"

some more info with youtube videos

UN International Telecommunication Union Secretary General Houlin Zhao Rappler

Houlin Zhao, Secretary-General, ITU - Opening speech at GET - 2016 ITU

WTPF 2009: Interview with Houlin ZHAO ITU

AI for Good INTERVIEWS: IRAKLI BERIDZE, Senior Strategy and Policy Advisor, UNICRI​ ITU

AI FOR GOOD INTERVIEWS: NICHOLAS DAVIS, WEF ITU

AI FOR GOOD INTERVIEWS: DR DAVID BENRIMOH, Content Expert, AIFRED ITU

ITU INTERVIEWS @ WTDC-17: Houlin Zhao Secretary - General, ITU ITU

Mr Houlin Zhao, Secretary-General, ITU Inauguration Speech ITU

HIGHLIGHTS Video: AI FOR GOOD Global Summit - DAY 1 ITU

AI FOR GOOD INTERVIEWS: DAVID HANSON, Founder and CEO, Hanson Robotics ITU

AI for Good INTERVIEWS: CLARA PALAU MONTAVA, Technology Team Lead, UNICEF Innovation ITU

AI for Good INTERVIEWS: PABLO RODRIGUEZ, CEO, Telefonica Alpha ITU

Interview: ITU Secretary-General Houlin Zhao on State of Broadband 2017 Report ITU

Houlin Zhao Secretary - General ITU, Opening Remarks at GSR-17 ITU

AI for GOOD INTERVIEWS: MARGARET CHAN, Director General, WHO ITU

Houlin Zhao, Secretary - General, ITU - Opening Speech ITU Telecom World 2015 ITU

AI FOR GOOD INTERVIEWS: CHAESUB LEE, Director, TSB, ITU ITU

GGC Profile : Houlin Zhao, Secretary-General, ITU ITU  

AI for Good Global Summit - 2017 Report

1. 1. Artificial Intelligence can help solve humanity’s greatest challenges AI for GOOD GLOBAL SUMMIT Hosted at ITU in Geneva, Switzerland 7-9 June 2017 REPORT #AIforGood
2. 2. UN partners Gold sponsors Corporate sponsor Investing in rural people United Nations Educational, Scientific and Cultural Organization UNIDIRUNITED NATIONS INSTITUTE FOR DISARMAMENT RESEARCH PARTNERS AND SPONSORS
3. 3. 01 Background p.6 02 What Exactly is Artificial Intelligence (AI)? p.13 03 Developing World and Prioritization p.15 04 Designing The Future – The Promise & the Peril of AI p.17 05 The Ethical Considerations of AI p.35 06 Social Challenges p.38 07 Next Steps p.40 08 Interdisciplinary Collaboration p.42 09 Summary of 16 Breakthrough Groups p.45 TABLE OF CONTENTS >< page 3/87
4. 4. The AI for Good Global Summit in June 2017 discussed how Artificial Intelligence (AI) could follow a development course able to assist the achievement of the United Nations’ Sustainable Development Goals. Co-organized by ITU and the XPRIZE Foundation, the event convened 500 representatives of government, industry, academic and research institutes, United Nations agencies and civil society to explore the potential of AI to accelerate progress in fields such as health, education, energy and the protection of our environment. ITU and XPRIZE share a firm belief that AI will help to solve some of the most pressing challenges to our planet and its people. We believe in AI’s promise of a better life for all. We also believe that this will only be achieved if government, industry and civil society work together to develop the positive aspects of the technology and manage related risks. FOREWORD >< page 4/87
5. 5. The AI for Good Global Summit brought together leading minds in AI and humanitarian action, establishing an historic milestone as the first event to launch inclusive global dialogue on the potential of AI to benefit humanity. The event hosted discussions aimed at building a common understanding of the applications and capabilities of emerging AI technologies. Thousands of people worldwide followed the discussions via webcasts offered by event partners ACM, IEEE, IFIP, CSTA and Digital Africa. This report summarizes these discussions and proposes actions and strategies to ensure the continued collaboration necessary to ensure that we take full advantage of emerging AI technologies. Sincerely, FOREWORD Houlin Zhao Secretary-General, ITU Marcus Shingles CEO, XPRIZE Foundation >< page 5/87
6. 6. 01. BACKGROUND The“AIforGoodGlobalSummit”tookplaceatITUinGeneva,Switzerland, on 7-9 June 2017, organized by ITU and the XPRIZE Foundation, in partnership with twenty UN agencies. Artificial Intelligence (AI) will be central to the achievement of the Sustainable Development Goals (SDGs) and could help solve humanity’s grand challenges by capitalizing on the unprecedented quantities of data now generated on sentient behavior, human health, commerce, communications, migration and more. >< page 6/87
7. 7. The meeting aimed to accelerate and advance the development and democratization of Artificial Intelligence (AI) solutions that can address specific global challenges related to poverty, hunger, health, equality (including gender equality), education, the environment and others. SUSTAINABLE DEVELOPMENT GOALS (SDGs) AI has great potential to advance the sustainable development goals and help to solve our greatest challenges. >< page 7/87
8. 8. Thank you for coming together to discuss a challenge of growing global importance. As someone with a background in engineering, I am deeply interested in the remarkable phenomenon of Artificial Intelligence. And, as Secretary-General of the UN, I am strongly committed to promoting global cooperation on this emerging issue. AI is advancing dramatically. It is already transforming our world, socially, economically and politically. We face a new frontier, with advances moving at warp speed. Artificial Intelligence can help analyze enormous volumes of data, which in turn can improve predictions, prevent crimes and help governments better serve people. But there are also serious challenges, and ethical issues at stake. There are real concerns about cybersecurity, human rights and privacy, not to mention the obvious and significant impact on the labor markets. The implications for development are enormous. Developing countries can gain from the benefits of AI, but they also face the highest risk of being left behind. This Summit can help ensure that AI charts a course that benefits humanity and bolsters our shared values. The UN stands ready to be a universal platform for discussion. Together, let us make sure we use Artificial Intelligence to enhance human dignity and serve the global good, thank you. Opening Address by António Guterres, Secretary-General of the United Nations >< page 8/87
9. 9. PARTICIPATION AND MEDIA COVERAGE >< page 9/87
10. 10. The AI for Good Summit attracted over 500participants and 5000webcast connections from all corners of the globe. Government Industry UN Agencies Civil Society International Organizations Academia A MULTI-STAKEHOLDER APPROACH… AI experts have said themselves that we cannot leave AI to just the experts – a multi-stakeholder approach is needed. The Summit hosted over 500 participants onsite, including many leading researchers and authorities on AI. One-third of participants and speakers were women. As such, this Summit brought together a unique mixture of many disciplines, with some of the top AI researchers (Stanford, UC Berkeley, Cambridge, etc.), top AI industry executives (Microsoft, Google, Facebook, etc.), and heads of UN agencies (WHO, UNODA, UNESCO). Delegates appreciated the ample networking opportunities that having such diverse participants brought. “It was great to see so many women attending the AI for Good Summit but we still need many more, including from the techno-political field.” Anja Kaspersen, ICRC Click play for a full HD recording of the webcast page 10/87 ><
11. 11. AI has captured the imagination of the public worldwide. The Summit was attended by over 45 journalists from a wide range of mainstream and technical international media such as the BBC, Reuters, Euronews and Wired. The global, multi-lingual coverage reached over 100 million people and the tone was overwhelmingly positive. AI has captured the imagination of the public worldwide media coverage The global multi-lingual coverage reached over 100 million people >< page 11/87
12. 12. AI for Good Summit goes viral with over 3.1millionviewson Facebook. watch the video www.facebook.com/pg/ITU/videos/?ref=page_internal THE EVENT ALSO GENERATED RECORD-BREAKING SOCIAL MEDIA COVERAGE WITH THE FACEBOOK LIVE SEGMENT GOING VIRAL AND BEING VIEWED BY OVER 3 MILLION PEOPLE. page 12/87 ><
13. 13. 02. WHAT EXACTLY IS ARTIFICIAL INTELLIGENCE (AI)? According to Eric Horvitz (Microsoft Research), “AI is not really any single thing – it is a set of rich sub-disciplines and methods, vision, perception, speech and dialogue, decisions and planning, robotics and so on. We have to consider all these different disciplinesandmethodsinseeking true solutions in delivering value to human beings and organizations”. Chaesub Lee, Director of ITU’s Standardization Bureau, observed “AI is quite old, and is made up of many individual parts, including protocols and common elements such as data. We have a situation of vertical development in AI and horizontal common elements”. At the Summit, AI was viewed as a set of associated technologies and techniques that can be used to complement traditional approaches, human intelligence and analytics and/or other techniques. >< page 13/87
14. 14. While AI is not “general magic,” Peter Diamandis (XPRIZE Foundation) claimed that “AI is probably the most important technology we have ever created.” Marcus Shingles (XPRIZE Foundation) suggested (we are now putting in place core tools to put in place systems to deal with hunger and other problems). Prof. Gary Marcus (NYU) sounded a note of caution about AI’s potential, observing, “we are not nearly as close to strong Artificial Intelligence as many believe.” Prof. Gary Marcus also argued that there has been little real progress in “true AI” (from Eliza to Alexa and Siri), where, according to him, progress has flatlined. In his estimation, machines can currently perform tasks taking a human being approximately one second of thought. Current AI developments work using an approach called deep learning — Prof. Yoshua Bengio (University of Montreal) described deep learning as a way of enabling computers to figure out good representations by considering how knowledge or information can be represented — representation has been central to AI from the beginning, but computers are now learning and discovering by practice using data how to represent information. However AI is defined, participants offered a range of perspectives on rates of progress in digitization. According to Marcus Shingles (XPRIZE), digitization is creating rapid and exponential advances in progress, data traffic and data storage. However, Prof. Yoshua Bengio (University of Montreal) suggested we are “…very, very far from ‘human-level’ AI… we have a lot of research due before we get there. Pretty much all of the major breakthroughs and amazing applications we have today of AI and machine learning are based on supervised learning.” In fact, progress in AI can proceed in stepwise increments (e.g. Google Translate’s significant improvement in November 2016, according to Prof. Jürgen Schmidhuber, Swiss AI Lab, IDSIA; Prof. of AI, USI & SUPSI, Switzerland; President of NNAISENSE). Prof. Joseph Konstan (University of Minnesota) pointed out that word processing systems are not many million times more efficient today than they were thirty years ago, but they are more useable and accessible than they ever have been. Perceived rates of progress therefore clearly depend on which aspects of AI are under discussion. page 14/87 ><
15. 15. 03. DEVELOPING WORLD & PRIORITIZATION Developing countries may have the most to gain from AI, but unless we are vigilant, they may also have the most to lose. In order to reap the benefits of AI, vast amounts of data are needed, which are only available through mass digitization – an area where developing countries lag far behind. There can be no mass digitization without universal and affordable access to broadband, which is central to ITU’s mission. We need to avoid a deepening of the digital divide, so the benefits of AI can be distributed equitably. It is vital that the needs of a diverse range of people, including the most vulnerable, guide the design and development of AI systems. Those who are furthest behind in terms of social and economic development, are at the centre of the SDGs and need to be at the centre of design and application of technologies such as AI. There was extensive discussion of the problem of significant (and exacerbated) inequality around AI. According to Salil Shetty (Amnesty International), inequality is at the highest level in 50 years in OECD countries, up seven-fold from 25 years ago, while income inequality is nine times higher. There was broad consensus that AI will create growing page 15/87 ><
16. 16. inequalities between developed and developing countries, between those who own and/or create the technology, and those who use it, as well as those people whose jobs may be replaced by AI. This was also a recurring theme in the breakthrough sessions on health and prosperity. World-changing AI needs massive interdisciplinary collaboration. Owning networks, security and verification can’t be done out of one lab. Robert Kirkpatrick (UN Global Pulse) suggested that this could include a Global Fund for AI for Social Good established to invest in cutting-edge research and promoting AI-based tools. The Fund would address the most pressing challenges of the SDGs for the most vulnerable populations and scale proven solutions through open source tools, to help define a common agenda for where the highest priority investment should go. BENEFITS OF AI CAN BE DISTRIBUTED EQUITABLY. page 16/87 ><
17. 17. 04.– DESIGNING THE FUTURE – THE PROMISE & THE PERIL OF AI 4.1. THE PROMISE & OPPORTUNITIES OF AI A number of speakers raised the considerable promise and opportunities of AI. According to Marcus Shingles (XPRIZE), those include gaining insights from “the sleeping giants of data,” improving decision-making and “harnessing the collective wisdom of the crowd.” AI is being used in a growing number of domains, in new and unexpected ways, many of which can improve human lives. Vicki Hanson (ACM) defined the measure of success for AI apps as the value they create for human lives, according to the 100-Year Study on AI (AI-100). AI can increase the productivity of arable land, detect early signs of cancer on scans, help find out how molecules interact to help find a cure for cancer and/ >< page 17/87
18. 18. or design personalized treatments for patients. Mr. Houlin Zhao, ITU Secretary-General, suggested that AI gives innovators the tools and capacity to innovate and to use AI to help solve some of our greatest challenges, starting with the SDGs. This is why it is so important that both developed and developing economies, multinationals, SMEs, universities and start-ups join forces to drive innovation forward. According to Rupert Stadler, CEO & Chairman of the Board of Management, AUDI AG, “90% of all car accidents are caused by human error. AI and autonomous driving will reduce the number of accidents significantly and will save lives. But of course, we also have to reshape the frameworks in our society – the first countries are issuing frameworks for autonomous driving, including Germany.” AI could be used to help Illiterate people online. According to Anders Sandberg (University of Oxford) AI is getting increasingly good at doing text to voice translation as well as voice to text translation. Even if you are illiterate, you can scan a text although you cannot read the letters, so you become partially literate and can interact with the formal world. According to Vicki Hanson (ACM), AI can help people with special needs in numerous ways (e.g. autonomous driving for visually impaired people, speech recognition software for people with hearing impairments or loss), potentially helping improve the lives of the estimated one billion people worldwide with disabilities. There is a need to make sure these technologies are empowering and franchising not only across disabilities, but across literacy levels, language and different levels of education as well. page 18/87 ><
19. 19. Prof. Fei-Fei Li (Google & Stanford University) warned that, although we have been progressing rapidly, vision technologies are still just at the very brink of making important contributions to our society. AI could significantly help the visually impaired; however, although we have many cameras from space, airplanes and drones overseeing Earth, we have still not yet given sight to most visually impaired people. Advances in AI also require advances in a number of other companion technologies as well. For example, visual recognition breaks down into object segmentation, object detection, 3D recognition, scene parsing, according to Prof. Fei-Fei Li (Stanford University). Thomas Wiegand (the Heinrich Hertz Institute), believes that AI, machine learning and communications are converging, with work in video compression standards, non-stationary and back projections using neural networks to tell us what in the input picture led to the final classification result. The ultimate aim of AI is towards self-learning software and computers or ‘transfer learning.’ Information that has been learned can be transferred between different robots or in the software systems and can be applied wherever various computing processes or connection to computing facilities are available, according to Anders Sandberg (University of Oxford). This can help humans by pushing the frontiers of innovation and driving forward progress in human development. page 19/87 ><
20. 20. 4.2. AI APPLICATIONSCITED DURING THE SUMMIT NUCLEAR TEST PROF. STUART RUSSELL (UC BERKELEY) described how AI is being used for monitoring verification of the nuclear test treaty to distinguish between natural seismic tremors and shocks triggered by nuclear tests. GLOBAL FISHING WATCH PAUL BUNJE (XPRIZE), described the Global Fishing Watch which uses simple machine learning to utilize the data in fishing vessels and applies machine learning algorithms to identify where the vessels have been and type of activities they are engaged in. ELECTRICITY SUPPLY PETER MARX (GENERAL ELECTRIC) described how General Electric is using machine learning for a number of different purposes, including using drones for inspecting power lines and in manufacturing. IMPROVING HEALTH According to ERIC HORVITZ (MICROSOFT), predictive modelling of cholera outbreaks can now be developed in advance based on powerful algorithms that can be used to distribute fresh water, or supply vaccines. >< page 20/87
21. 21. AI APPLICATIONS CITED DURING THE SUMMIT JACK, THE SELF-DRIVING CAR Rupert Stadler, CEO & Chairman of Board of Management, AUDI, suggested that reliable self-driving cars will be in place by the end of the next decade, 2030, at the latest, when the car will drive all by itself. GUIDE ROBOT Prof. Manuela Veloso (Carnegie Mellon University) has developed “cobots” or collaborative robots to guide people around a building, according to a pre- programmed schedule, and also to serve coffee, and describe (“verbalization”) what they did with different levels of abstraction. >< page 21/87
22. 22. AI APPLICATIONS CITED DURING THE SUMMIT TRANSLATOR Google Translate replaced their old system with Long Short Term Memory (LSTM) in November 2016, and it got noticeably better, around 50% better than previous performance. Jürgen Schmidhuber (Swiss AI Lab, IDSIA) stated that Amazon Echo is a LSTM network which has learned to sound like a woman. It is not a prerecorded woman’s voice, it is a computer that has learned to sound like a woman. SKYPE Skype speaks nine languages, and does real-time translation through machine learning. Training an English-trained system with Portuguese, Mandarin and French produces better results in the original language and enables shortcuts in the training process to obtain very high learning power, according to Peter Lee (Microsoft). >< page 22/87
23. 23. AI APPLICATIONS CITED DURING THE SUMMIT MONITORING NUTRITION Chris Fabian (UNICEF Innovation) described the use of simple AI in mid- upper arm bands (MUACs) to monitor the nutritional status of children, which are valuable tools for famine response. Companies are developing similar tools using face recognition. MONITORING HYGIENE AI algorithms can provide continuous tracking of clinicians’ movements in hospital, without revealing who they are. Sensors in hallways close to hand- hygiene dispensers with a deep learning recognition system monitor clinicians’ hygiene practices, with a performance better than many of the state-of-the- art systems, reducing hospital-acquired infection rates – Prof. Fei-Fei Li (Google & Stanford University). >< page 23/87
24. 24. AI APPLICATIONS CITED DURING THE SUMMIT ANALYSIS OF TUMORS With regards to computer analysis of tumors, Peter Lee (Microsoft) acknowledged that medical imaging is not a ‘replacement’ for radiologists, but offers tremendous improvements in augmenting what radiologists are able to do. Radiologists look at scans slice by slice, computers can work in 3D. CANCER CELLSINCOLONPOLYPS Detection algorithms for identifying cancer cells in colon polyps – these detection algorithms can do better than (very rare) specialist doctors and much better than regular doctors, according to Prof. Yoshua Bengio (University of Montreal). Prof. Lynne Parker (University of Tennessee-Knoxville) provided statistics for medical pathology showing the top human doctors make 3.5% erroneous decisions and the top AI technology has an error rate of 7.5%, but combined together, the error rate can drop to 0.5%. >< page 24/87
25. 25. AI APPLICATIONS CITED DURING THE SUMMIT TRICORDERS Basil Harris (Basil Leaf Technologies) designed a tricorder that can diagnose 13 diseases (pneumonia, urinary tract infection, diabetes, acute diseases, chronic conditions, etc), with AI built into the device in the decision-making control. >< page 25/87
26. 26. AI APPLICATIONS CITED DURING THE SUMMIT MIT’S OPEN AGRICULTURE Labs are embedded within food containers to control temperature, food supply, humidity etc. and use AI black box optimization to evolve a process that gets to the best ingredients to obtain the very best basil, quantity and quality-wise, according to Antoine Blondeau (Sentient Technologies). USING SATELLITES Andrew Zolli (Planet Labs) described how Planet Lab is using satellite imagery to monitor planting, agricultural indicators in Kenya, a flood event in Sri Lanka, the growth of Dar es Salaam city and an Internally Displaced Persons (IDP) camp in Uganda. >< page 26/87
27. 27. AI APPLICATIONS CITED DURING THE SUMMIT AGRICULTURE IBM’s Project Lucy, where IBM committed a hundred million dollars to enhance African infrastructure, uses AI to help in agriculture so farmers can improve crop yields with less water, to improve education and to enhance access to healthcare services, according to Neil Sahota. MIT’S CELL C PROJECT Predicts the onset of sepsis and blood infections in hospitals. The biggest single symptom of a blood infection is a massive drop in blood pressure, but today there is no way to predict it. You can just react to it. According to Antoine Blondeau (Sentient Technologies), an AI project across 6,000 patients was able to predict the onset of sepsis 30 minutes ahead of time, which helped preempt and prevent it in 91% of cases. “META” INITIATIVE Sam Molyneux (Chan Zuckerberg Initiative) described the “Meta” initiative within the Chan Zuckerberg Initiative, which is thinking about how AI can be used to solve or power scientific knowledge production, and is in the process of making available a “global scientific awareness system” that can accelerate advances in sciences and make advances available for free, open to scientists, researchers, funders and policy-makers. >< page 27/87
28. 28. AI APPLICATIONS CITED DURING THE SUMMIT LEARNING CAPABILITY IBM Watson can be used to see the learning capability of each student, move at their pace and help them develop confidence to master subjects. Watson can recommend things the student might like or be interested in learning, based on their specific personality traits. Watson can adjust the way it interacts with the student based on their age, competency level, vocabulary and emotional state, according to Neil Sahota from IBM. ALS Francesca Rossi (IBM, University of Padova) described the use of IBM Watson for combatting Amyotrophic Lateral Sclerosis (ALS), by using AI to examine genes to see whether they are correlated with ALS. 8/10 top ten gene candidates were found to be correlated with ALS, five of which were not known before. >< page 28/87
29. 29. AI APPLICATIONS CITED DURING THE SUMMIT “THE LOVING AI PROJECT” Talking, cloud-based robots and “the loving AI project” to understand the human condition, the value of life, liberty and the value of humans as well as machines. Hanson Robotics has developed a series of robots, which can conduct independent conversations and learn from their interactions with humans. REPLICATING A CENSUS The 200 most populated cities in America are using deep learning to recognize and identify cars (make and model for 2657 types of cars manufactured after 1990) in 50 million Google Streetview images. In additional work, Prof. Fei-Fei Li (Google & Stanford University) observed that car models are correlated with GDP, average income, socio-economic indicators and voting patterns. >< page 29/87
30. 30. AI APPLICATIONS CITED DURING THE SUMMIT FORMS OF INFORMATION Robert Kirkpatrick (UN Global Pulse) described how big data, together with the power of AI and high performance computing, are generating new forms of information and insight with tremendous value for answering business questions and for fighting hunger, poverty and disease. Labs are using big data and deep learning to map discrimination against refugees in Europe, recognize rescue attempts from shipping data in the Mediterranean, detect fires in Indonesian rainforests, predict food prices from Twitter, fight bottlenecks in sanitation services in refugee camps and generate socioeconomic indicators from satellite imagery and flows of postal traffic. >< page 30/87
31. 31. 4.3. POTENTIAL PERILS OF AI What will AI be used for? Prof. Stuart Russell (UC Berkeley) suggested that we have to be careful with regards to the problem of specifying the purpose correctly – rather like King Midas, we have to be careful what we ask for. He suggested that intelligent machines will be incredibly good at carrying out what we ask them to carry out. If we ask them to carry out the wrong thing we shall be sorry (like King Midas), and we have a long history of failing to specify the purpose correctly. He called for “human-compatible AI.” Indeed, Prof. Urs Gasser (Harvard University) pointed out there is a risk we might start to delegate our questions and our very judgments to the machine and autonomous systems, which could then effectively end up determining which are the important questions. Izumi Nakamitsu (UNODA) stated that autonomous weapons systems are already in place to select and engage targets independently. AI will increase the speed and precision of action in future battlefields, which could have massive damaging consequences, given their inherent unpredictability. The growing separation between users and targets lowers thresholds for violence, and could strain protection for civilians. “There are currently no multilateral standards or regulations covering military AI applications, and Member States have not systematically considered the challenges posed by AI technology through the UN or otherwise,” including any treatment of current or future AI systems. With regards to the use of AI technologies with malicious intent, Cindy Smith of UNICRI argued that we must think about the new threats to international peace and security that will emerge from the advances in AI technologies – “If you build it, criminals will use it. In fact, some of these technologies have already been used fo.........

                                 Some other references about +AI for GOOD" these last months ===>

IFIP AI Expert Stephen Ibaraki Chairs United Nations AI Briefing

...d who also found