#unsupervisedlearning; #nextAIrevolution; #MITTechnologyReview
Massachusetts (U.S.), July 28 (Canadian-Media): Yann LeCun, Facebook’s chief AI scientist, believes unsupervised learning will bring about the next AI revolution, MIT Technology Review reports said.
Artificial Intelligence Technology. Image credit: Pixaby
“Nobody tells the baby that objects are supposed to fall,” said Yann LeCun, the chief AI scientist at Facebook and a professor at NYU, during a webinar on Thursday organized by the Association for Computing Machinery, an industry body. And because babies don’t have very sophisticated motor control, he hypothesizes, “a lot of what they learn about the world is through observation.”
That theory could could lead researchers to advance the boundaries of artificial intelligence.
Deep learning, the category of AI algorithm, has increased the possibility of giving machines perceptual abilities like vision. The next step is to instill them with sophisticated reasoning.
New techniques of giving machines a kind of working memory are helping to overcome this limitation.
“Obviously we’re missing something,” LeCun said.
The answer, he thinks, is in deep-learning subcategory known as unsupervised learning. LeCun prefers the term “self-supervised learning” because it essentially uses part of the training data to predict the rest of the training data.
In recent years, such algorithms have facilitated in natural-language processing because of their ability to find the relationships between billions of words.
“Everything we learn as humans—almost everything—is learned through self-supervised learning. There’s a thin layer we learn through supervised learning, and a tiny amount we learn through reinforcement learning,” he said. “If machine learning, or AI, is a cake, the vast majority of the cake is self-supervised learning.”
What does this look like in practice? Researchers should begin by focusing on temporal prediction.
“This is kind of a simulation of what’s going on in your head, if you want,” LeCun said.
“It's a good idea to do video prediction in the context of self-driving cars because you might want to know in advance what other cars on the streets are gonna do,” he said.
Ultimately, unsupervised learning will help machines develop a model of the world that can then predict future states of the world, he said. LeCun is confident: “The next revolution of AI will not be supervised.”
#AI; #UN: #intractable problems; #UNCTAD; #SDG
United Nations, May 14 (Canadian-Media/UN): Frontier technologies such as artificial intelligence (AI) can, if properly harnessed, help solve the world’s most intractable problems, said speakers at the opening session of the United Nations (UN)’ top forum on science and technology for development on 13 May, UN reports said.
Artificial Intelligence. Image credit: Pixaby
The twenty-second annual session of the United Nations Commission on Science and Technology for Development (CSTD) started with experts discussing the pace, challenges and promises offered by rapid technological change.
“We need to ensure science, technology and innovation (STIs) contribute to sustainable and inclusive development,” said UNCTAD Deputy Secretary-General Isabelle Durant, while opening the event held at the Palais des Nations in Geneva, Switzerland.
Instead, the machines would ‘learn’ from examples and generate solutions, which would have a huge impact on many aspects of our lives, Mr. Schmidhuber said.
He said through “active AI” machines could learn just about anything, like babies do, and develop their own solutions to many problems.
Though currently expensive, AI would gradually become more affordable people, said Mr. Schmidhuber, citing the example of mobile phones, which had become cheaper over time.
“In the long run, AI will be very cheap and democratic. We are going to see the ideas and techniques that make robots smart spreading. Everyone will benefit,” Mr. Schmidhuber said.
However, in the transition period, there would be winners and losers, he warned, necessitating measures to soften the blows of rapid technological change.
The other “great mind”, Dame Wendy Hall, professor of computer science at the University of Southampton, said when the internet was created 30 years ago, greater focus was laid on its positive impact – the democratization of knowledge.
“Little did we know it would be used to launch attacks on our democracy,” said Ms. Hall, who co-founded the Web Science Trust with Sir Tim Berners Lee and others. “We have to work on how to fix the negative impacts of our technologies.”
Ms. Hall said lessons could be learned from mistakes made in the past to minimize the potential negative impacts of frontier technologies.
“This is a live experiment we have done over the last 30 years,” said Ms. Hall, who helped develop the United Kingdom’s AI strategy review.
She said discussions taking place at the Commission could help guide the development of policies to ensure such technologies are safely deployed.
“We have to look at the ethics of what we are developing. We have to lay the groundwork now,” Ms. Hall said.
For instance, she said people creating AI software needed to be diverse in terms of gender, ethnicity, culture, race, religion, age and accessibility. “If AI isn’t diverse, it is not ethical,” underlined Ms. Hall.
Equally important was developing skills in developing countries to spread the benefits of AI and other frontier technologies, Ms. Hall said.
Ms. Hall also noted the importance of maintaining a human presence in the design and deployment of frontier technologies.
“Imagine when machines go wrong, for example, in an old age home. Who fixes them?” posed Ms. Hall. “Or when we put something in our home and the company that build it goes out of business?”
New technologies also presented the means to face the climate challenge and reduce the ever-rising emissions of carbon dioxide, said Carlo Rubbia, the third “great mind.”
“Elimination of carbon dioxide emissions is of importance to the future of mankind,” said Mr. Rubbia, a former director-general of CERN, the European organization for nuclear research, and joint Nobel prize winner in physics for his contributions to particle physics, which led to the discovery of the field particles W and Z.
Renewable energy was one the most effective tools in the fight against climate change, as it could help in reducing emissions from fossil fuels, Mr. Rubbia said.
Mr. Rubbia noted that science had been become international, with countries such as China creating technological solutions in areas previously dominated by Europe and the United States.
The “great minds” discussion was moderated by award-winning journalist Didi Akinyuelure.
In a video message to the attendees, the United Nations Economic and Social Council (ECOSOC) President Inga Rhonda King said STIs were cross-cutting and pertinent to all SDGs.
The Commission is taking place from 13 to 17 May. It is a subsidiary body of the ECOSOC and provides the General Assembly and ECOSOC with high-level advice on relevant science and technology issues.
UNCTAD is responsible for the substantive servicing of the Commission.
Ms. Durant said the CSTD could be a catalyst for closer collaboration to strengthen the capacity of developing countries to address development challenges.
The role of technologies in fostering sustainable development was undeniable, but they also presented issues whose solutions required inclusive dialogue, Ms. Durant noted.
For instance, rapid technological change could exacerbate inequalities within and between countries, slowing down progress towards Sustainable Development Goals (SDGs).
“We need a global dialogue on harnessing technology while at the same time protecting our social and economic well-being,” Ms. Durant said.
Conversation with great minds
Three distinguished scientists shared their perspectives in a session entitled “A Conversation with Great Minds”, livestreamed on Facebook.
AI, which involves using technology to solve problems that used to require human intelligence, will shape our future in ways not seen before, said Jürgen Schmidhuber, one of the “great minds”.
“We will see the rise of machines that don’t slavishly imitate,” said Mr. Schmidhuber, director and professor at the Swiss AI Lab IDSIA and co-founder and chief scientist at the AI research company NNAISENCE.
#UNCSTD; #EconomicandSocialCouncil; #STI; #WorldSummitontheInformationSociety;
United Nations, May 14 (Canadian-Media): The twenty-second session of the Commission on Science and Technology for Development (CSTD) will be held at the Palais des Nations, in Geneva, Switzerland from 13 to 17 May 2019, United Nations (UN) reports said.
CSTD/Image Credit: UN
The CSTD is a subsidiary body of the Economic and Social Council (ECOSOC) and the United Nations focal point for science, technology and innovation (STI) for development, in analyzing how STI, including information and communications technologies (ICTs), serve as enablers of the 2030 Agenda.
It acts as a forum for strategic planning, sharing lessons learned and best practices, providing foresight about critical trends in STI in key sectors of the economy, the environment and society, and drawing attention to emerging and disruptive technologies.
Every year, the Commission has two priority themes. This year these are:
The twenty-second session will also review the progress made in the implementation of the outcomes of the World Summit on the Information Society (WSIS).
In addition, the Commission will hear presentations on national science, technology and innovation policy reviews (STIP Reviews).
Participants will include ministers and representatives of governments, civil society, the business community, academia and international and regional organizations. Most member States will be represented by high-level delegations.
The opening of the session will be at 10 a.m. on Monday, 13 May 2019.
The first day will consist of an opening ceremony followed by a special segment devoted to A Conversation with Great Minds – a dialogue between eminent thinkers in science and technology, which will be moderated by an award-winning journalist. It will continue with two high-level roundtables:
#UNCommissiononScienceandTechnologyforDevelopment; #inclusiveness; #SDGs; #multi-stakeholderinvolvement; #artificialintelligence;
United Nations, May 14 (Canadian-Media/UN): Today, more than ever before, a wealth of opportunities are within reach to provide solutions to many of the global challenges that have for far too long impeded sustainable pathways to development – thanks to the rapid advances made in technology, United Nations (UN) reports said.
United Nations/ facebook
The global community is faced with unprecedented possibilities to eliminate hunger, poverty, ignorance, disease, environmental degradation, and thereby accelerate progress towards the 2030 Development Agenda.
Yet the rapid and accelerating speed of technological progress sometimes outpaces the ability of societies to adapt to the ensuing social and economic changes. It’s in retrospect that we find that technological change leads to unintended consequences that we could never have anticipated.
“This is not an issue of only understanding the designs, characteristics, and functions of specific scientific and technological applications,” says Shamika Sirimanne, director of UNCTAD’s division on technology and logistics.
“Rather, it involves analysing how frontier technologies are adopted, adapted, and implicated in complex social, political, and environmental settings,” she says.
The twenty-second annual session of the UN Commission on Science and Technology for Development (CSTD), convening from 13 to 17 May in Geneva, addresses some profound questions on the implications of rapid technological change for sustainable development.
While frontier technologies offer unprecedented opportunities to transform the practice, implementation and monitoring of sustainable development, they also pose considerable normative questions.
For example, social media platforms were designed to connect us not only to those with whom we share existing affiliations but also to others whose locations, perspectives, and ideas differed from our own. Instead, these platforms are being manipulated to further divide communities and to make it difficult for different stakeholders to find common ground.
Ground-breaking advances in gene editing holds the promise to improve human, animal and plant health. But unsafe applications without regulatory oversight could unexpectedly harm the health of humanity and the environment.
And arguably one of the most transformative technological developments of our time – artificial intelligence – could be used to help identify solutions to our most intractable economic, social and environmental challenges. However, its applications in public and private sectors could counterintuitively scale inequality and make the world less secure.
As we begin to grasp the multifaceted nature of rapid technological progress, it calls into question how we can shape global mechanisms and platforms to better understand how technology evolves and impacts our world.
Technological change and innovation cannot be understood and harnessed any longer using national frameworks only. Frontier technologies are created and deployed through global networks and their implications quickly extend far beyond the regions where they originally emerge.
“We need to develop tools that all countries can use to understand how technological change is relevant for their own development and assess its economic and social benefits, consequences and feedbacks,” UNCTAD Deputy Secretary-General Isabelle Durant says.
The supranational implications of technologies have to be considered by technology assessments, and must involve multiple stakeholders, and assist national policymakers in developed and developing countries to enhance their capacities to make sense of rapid technological change.
Beyond understanding innovation and its potential impacts, the international community needs to define the shared values that define the direction in which rapid technological should proceed.
There has been a proliferation of initiatives to address the normative dimensions of various new and emerging technologies such as artificial intelligence, gene editing, and other transformative and disruptive technologies.
The principles of inclusiveness
This bottom-up emergence of values, standards and norms points to the need for a global discussion and consensus on how to promote innovation while protecting our health, the environment, and our economic, social, and political institutions.
The principles of inclusiveness and multi-stakeholder involvement must be embedded at the heart of any global discourse on technological change, to ensure coherence between multiple initiatives and their consistence with the international community’s development agenda.
Developing countries, especially least developed countries, not engaged in the development of frontier technologies but likely to be affected by their consequences, need to be part of this global discourse.
#Toronto, #World'sAICapital; #MaRSDiscoveryDistrict; #Collision; AIEcosystem; #MonaSiddiqui, HichamOudghiri, #DaphneKoller, #StephenWolfram, #Enigma; #U.S.DepartmentofHealthandHumanServices; #Coursera; #WolframResearch
Toronto, May 1 (Canadian-Media): Computing power caught up to their mathematics by a small group of researchers from Toronto led by Geoffrey Hinton worked at MaRS Discovery District (MaRSDD) and the world woke up to the potential of Artificial Intelligence (AI) and made Toronto as the worlds's AI capital, MaRSDD reports said.
Image Credit: MaRsDD/Toronto Abstract
With the booming of this industry, talent and expertise built up in this region over years have made Toronto a top destination for companies to find a competitive edge. Major investments in AI labs here have been made by Google, Microsoft, NVIDIA, Samsung, Uber, Intel, LG, General Motors and Thomson Reuters. Innovative AI-based products for sectors ranging from security and advanced manufacturing to farming and financial services are being delivered by Toronto’s thriving startup ecosystem, the fastest growing in North America.
That also makes Toronto a natural home for Collision, the fastest-growing tech conference in North America, now in its fifth year. The conference, taking place at the Enercare Centre, Toronto this May, is expected to draw 25,000 attendees and is a major event for North American startups, investors, corporates and scientists.
Citing Toronto’s flourishing AI ecosystem, Collision CEO Paddy Cosgrave has decided to stage the conference here for the next three years.
Toronto prepares to welcome the AI community’s most brilliant minds. The four noteworthy speakers would be: Mona Siddiqui, chief data officer, U.S. Department of Health and Human Services; Hicham Oudghiri, co-founder and CEO, Enigma; Daphne Koller, co-founder and co-chairman, Coursera; Stephen Wolfram, founder and CEO, Wolfram Research.