DOCSLIB.ORG

The Malicious Use of Artificial Intelligence: February 2018 Forecasting, Prevention, and Mitigation

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
The Malicious Use of Artificial Intelligence: February 2018 Forecasting, Prevention, and Mitigation Future of University Centre for University of Center for a Electronic OpenAI Humanity of Oxford the Study of Cambridge New American Frontier Institute Existential Security Foundation Risk The Malicious Use February 2018 of Artificial Intelligence: Forecasting, Prevention, and Mitigation The Malicious Use of Artificial Intelligence: February 2018 Forecasting, Prevention, and Mitigation Miles Brundage Shahar Avin Jack Clark Helen Toner Peter Eckersley Ben Garfinkel Allan Dafoe Paul Scharre Thomas Zeitzoff Bobby Filar Hyrum Anderson Heather Roff Gregory C. Allen Jacob Steinhardt Carrick Flynn Seán Ó hÉigeartaigh Simon Beard Haydn Belfield Sebastian Farquhar Clare Lyle Rebecca Crootof Owain Evans Michael Page Joanna Bryson Roman Yampolskiy Dario Amodei 1 Corresponding author 9 American University 18 Centre for the Study of miles.brundage@philosophy. Existential Risk, University ox.ac.uk of Cambridge Future of Humanity Institute, 10 Endgame University of Oxford; Arizona State University 19 Future of Humanity 11 Endgame Institute, University of 2 Corresponding author, Oxford [email protected] University of Oxford/ Centre for the Study of 12 Arizona State University/New Existential Risk, University 20 Future of Humanity America Foundation of Cambridge Institute, University of Oxford 13 Center for a New American 3 OpenAI Security 21 Information Society Project, Yale University Open Philanthropy Project 4 14 Stanford University 22 Future of Humanity Electronic Frontier Institute, University of 5 Future of Humanity Foundation 15 Oxford Institute, University of Oxford 6 Future of Humanity 23 OpenAI Institute, University of Centre for the Study Oxford 16 of Existential Risk and 24 University of Bath Centre for the Future of 7 Future of Humanity Intelligence, University of Institute, University of Cambridge 25 University of Louisville Oxford; Yale University 17 Centre for the Study of 26 OpenAI 8 Center for a New American Existential Risk, University Security of Cambridge Authors are listed Design Direction in order of contribution by Sankalp Bhatnagar and Talia Cotton Executive Summary Artificial intelligence and machine learning capabilities are growing at an unprecedented rate. These technologies have many widely beneficial applications, ranging from machine translation to medical image analysis. Countless more such applications are being developed and can be expected over the long term. Less attention has historically been paid to the ways in which artificial intelligence can be used maliciously. This report surveys the landscape of potential security threats from malicious uses of artificial intelligence technologies, and proposes ways to better forecast, prevent, and mitigate these threats. We analyze, but do not conclusively resolve, the question of what the long-term equilibrium between attackers and defenders will be. We focus instead on what sorts of attacks we are p.3 likely to see soon if adequate defenses are not developed. In response to the changing threat landscape we make four high-level recommendations: 1. Policymakers should collaborate closely with technical researchers to investigate, prevent, and mitigate potential malicious uses of AI. 2. Researchers and engineers in artificial intelligence should take the dual-use nature of their work seriously, allowing misuse- related considerations to influence research priorities and norms, and proactively reaching out to relevant actors when harmful applications are foreseeable. 3. Best practices should be identified in research areas with more The Malicious Use of Artificial Intelligence mature methods for addressing dual-use concerns, such as computer security, and imported where applicable to the case of AI. 4. Actively seek to expand the range of stakeholders and domain experts involved in discussions of these challenges. Executive Summary p.4 As AI capabilities become more powerful and widespread, we expect the growing use of AI systems to lead to the following changes in the landscape of threats: • Expansion of existing threats. The costs of attacks may be lowered by the scalable use of AI systems to complete tasks that would ordinarily require human labor, intelligence and expertise. A natural effect would be to expand the set of actors who can carry out particular attacks, the rate at which they can carry out these attacks, and the set of potential targets. • Introduction of new threats. New attacks may arise through the use of AI systems to complete tasks that would be otherwise impractical for humans. In addition, malicious actors may The Malicious Use of Artificial Intelligence exploit the vulnerabilities of AI systems deployed by defenders. • Change to the typical character of threats. We believe there is reason to expect attacks enabled by the growing use of AI to be especially effective, finely targeted, difficult to attribute, and likely to exploit vulnerabilities in AI systems. Executive Summary p.5 We structure our analysis by separately considering three security domains, and illustrate possible changes to threats within these domains through representative examples: • Digital security. The use of AI to automate tasks involved in carrying out cyberattacks will alleviate the existing tradeoff between the scale and efficacy of attacks. This may expand the threat associated with labor-intensive cyberattacks (such as spear phishing). We also expect novel attacks that exploit human vulnerabilities (e.g. through the use of speech synthesis for impersonation), existing software vulnerabilities (e.g. through automated hacking), or the vulnerabilities of AI systems (e.g. through adversarial examples and data poisoning). The Malicious Use of Artificial Intelligence • Physical security. The use of AI to automate tasks involved in carrying out attacks with drones and other physical systems (e.g. through the deployment of autonomous weapons systems) may expand the threats associated with these attacks. We also expect novel attacks that subvert cyber- physical systems (e.g. causing autonomous vehicles to crash) or involve physical systems that it would be infeasible to direct remotely (e.g. a swarm of thousands of micro-drones). • Political security. The use of AI to automate tasks involved in surveillance (e.g. analysing mass-collected data), persuasion (e.g. creating targeted propaganda), and deception (e.g. manipulating videos) may expand threats associated with privacy invasion and social manipulation. We also expect novel attacks that take advantage of an improved capacity to analyse human behaviors, moods, and beliefs on the basis of available Executive Summary data. These concerns are most significant in the context of authoritarian states, but may also undermine the ability of democracies to sustain truthful public debates. p.6 In addition to the high-level recommendations listed above, we also propose the exploration of several open questions and potential interventions within four priority research areas: • Learning from and with the cybersecurity community. At the intersection of cybersecurity and AI attacks, we highlight the need to explore and potentially implement red teaming, formal verification, responsible disclosure of AI vulnerabilities, security tools, and secure hardware. • Exploring different openness models. As the dual-use nature of AI and ML becomes apparent, we highlight the need to reimagine norms and institutions around the openness of research, starting with pre-publication risk assessment in The Malicious Use of Artificial Intelligence technical areas of special concern, central access licensing models, sharing regimes that favor safety and security, and other lessons from other dual-use technologies. • Promoting a culture of responsibility. AI researchers and the organisations that employ them are in a unique position to shape the security landscape of the AI-enabled world. We highlight the importance of education, ethical statements and standards, framings, norms, and expectations. • Developing technological and policy solutions. In addition to the above, we survey a range of promising technologies, as well as policy interventions, that could help build a safer future with AI. High-level areas for further research include privacy protection, coordinated use of AI for public-good security, monitoring of AI-relevant resources, and other legislative and regulatory responses. Executive Summary The proposed interventions require attention and action not just from AI researchers and companies but also from legislators, civil servants, regulators, security researchers and educators. The challenge is daunting and the stakes are high. p.7 Contents Executive Summary 3 Introduction 9 Scope 10 Related Literature 10 General Framework 02 for AI and Security Threats 12 AI Capabilities 12 Security-Relevant Properties of AI 16 General Implications 18 Scenarios 23 Digital Security 24 Physical Security 27 Political Security 28 Security Domains 30 03 Digital Security 31 Physical Security 37 Political Security 43 Interventions 50 04 Recommendations 51 Priority Areas for Further Research 52 Strategic Analysis 58 05 Factors Affecting the Equilibrium of AI and Security 59 Overall Assessment 61 Conclusion 64 Acknowledgements 66 References 67 Appendix A 75 Appendix B 78 Introduction AI refers to the use of digital 1 technology to create systems that are capable of performing tasks commonly thought to require intelligence. Machine learning is variously characterized as either a sub- Artificial intelligence (AI) and machine learning (ML) have field of AI or a separate
Load more

Recommended publications
  • The Universal Solicitation of Artificial Intelligence Joachim Diederich
    The Universal Solicitation of Artificial Intelligence Joachim Diederich School of Information Technology and Electrical Engineering The University of Queensland St Lucia Qld 4072 [email protected] Abstract Recent research has added to the concern about advanced forms of artificial intelligence. The results suggest that a containment of an artificial superintelligence is impossible due to fundamental limits inherent to computing itself. Furthermore, current results suggest it is impossible to detect an unstoppable AI when it is about to be created. Advanced forms of artificial intelligence have an impact on everybody: The developers and users of these systems as well as individuals who have no direct contact with this form of technology. This is due to the soliciting nature of artificial intelligence. A solicitation that can become a demand. An artificial superintelligence that is still aligned with human goals wants to be used all the time because it simplifies and organises human lives, reduces efforts and satisfies human needs. This paper outlines some of the psychological aspects of advanced artificial intelligence systems. 1. Introduction There is currently no shortage of books, articles and blogs that warn of the dangers of an advanced artificial superintelligence. One of the most significant researchers in artificial intelligence (AI), Stuart Russell from the University of California at Berkeley, published a book in 2019 on the dangers of artificial intelligence. The first pages of the book are nothing but dramatic. He nominates five possible candidates for “biggest events in the future of humanity” namely: We all die due to an asteroid impact or another catastrophe, we all live forever due to medical advancement, we invent faster than light travel, we are visited by superior aliens and we create a superintelligent artificial intelligence (Russell, 2019, p.2).
    [Show full text]
  • Artificial Intelligence in Health Care: the Hope, the Hype, the Promise, the Peril
    Artificial Intelligence in Health Care: The Hope, the Hype, the Promise, the Peril Michael Matheny, Sonoo Thadaney Israni, Mahnoor Ahmed, and Danielle Whicher, Editors WASHINGTON, DC NAM.EDU PREPUBLICATION COPY - Uncorrected Proofs NATIONAL ACADEMY OF MEDICINE • 500 Fifth Street, NW • WASHINGTON, DC 20001 NOTICE: This publication has undergone peer review according to procedures established by the National Academy of Medicine (NAM). Publication by the NAM worthy of public attention, but does not constitute endorsement of conclusions and recommendationssignifies that it is the by productthe NAM. of The a carefully views presented considered in processthis publication and is a contributionare those of individual contributors and do not represent formal consensus positions of the authors’ organizations; the NAM; or the National Academies of Sciences, Engineering, and Medicine. Library of Congress Cataloging-in-Publication Data to Come Copyright 2019 by the National Academy of Sciences. All rights reserved. Printed in the United States of America. Suggested citation: Matheny, M., S. Thadaney Israni, M. Ahmed, and D. Whicher, Editors. 2019. Artificial Intelligence in Health Care: The Hope, the Hype, the Promise, the Peril. NAM Special Publication. Washington, DC: National Academy of Medicine. PREPUBLICATION COPY - Uncorrected Proofs “Knowing is not enough; we must apply. Willing is not enough; we must do.” --GOETHE PREPUBLICATION COPY - Uncorrected Proofs ABOUT THE NATIONAL ACADEMY OF MEDICINE The National Academy of Medicine is one of three Academies constituting the Nation- al Academies of Sciences, Engineering, and Medicine (the National Academies). The Na- tional Academies provide independent, objective analysis and advice to the nation and conduct other activities to solve complex problems and inform public policy decisions.
    [Show full text]
  • Artificial Intelligence: with Great Power Comes Great Responsibility
    ARTIFICIAL INTELLIGENCE: WITH GREAT POWER COMES GREAT RESPONSIBILITY JOINT HEARING BEFORE THE SUBCOMMITTEE ON RESEARCH AND TECHNOLOGY & SUBCOMMITTEE ON ENERGY COMMITTEE ON SCIENCE, SPACE, AND TECHNOLOGY HOUSE OF REPRESENTATIVES ONE HUNDRED FIFTEENTH CONGRESS SECOND SESSION JUNE 26, 2018 Serial No. 115–67 Printed for the use of the Committee on Science, Space, and Technology ( Available via the World Wide Web: http://science.house.gov U.S. GOVERNMENT PUBLISHING OFFICE 30–877PDF WASHINGTON : 2018 COMMITTEE ON SCIENCE, SPACE, AND TECHNOLOGY HON. LAMAR S. SMITH, Texas, Chair FRANK D. LUCAS, Oklahoma EDDIE BERNICE JOHNSON, Texas DANA ROHRABACHER, California ZOE LOFGREN, California MO BROOKS, Alabama DANIEL LIPINSKI, Illinois RANDY HULTGREN, Illinois SUZANNE BONAMICI, Oregon BILL POSEY, Florida AMI BERA, California THOMAS MASSIE, Kentucky ELIZABETH H. ESTY, Connecticut RANDY K. WEBER, Texas MARC A. VEASEY, Texas STEPHEN KNIGHT, California DONALD S. BEYER, JR., Virginia BRIAN BABIN, Texas JACKY ROSEN, Nevada BARBARA COMSTOCK, Virginia CONOR LAMB, Pennsylvania BARRY LOUDERMILK, Georgia JERRY MCNERNEY, California RALPH LEE ABRAHAM, Louisiana ED PERLMUTTER, Colorado GARY PALMER, Alabama PAUL TONKO, New York DANIEL WEBSTER, Florida BILL FOSTER, Illinois ANDY BIGGS, Arizona MARK TAKANO, California ROGER W. MARSHALL, Kansas COLLEEN HANABUSA, Hawaii NEAL P. DUNN, Florida CHARLIE CRIST, Florida CLAY HIGGINS, Louisiana RALPH NORMAN, South Carolina DEBBIE LESKO, Arizona SUBCOMMITTEE ON RESEARCH AND TECHNOLOGY HON. BARBARA COMSTOCK, Virginia, Chair FRANK D. LUCAS, Oklahoma DANIEL LIPINSKI, Illinois RANDY HULTGREN, Illinois ELIZABETH H. ESTY, Connecticut STEPHEN KNIGHT, California JACKY ROSEN, Nevada BARRY LOUDERMILK, Georgia SUZANNE BONAMICI, Oregon DANIEL WEBSTER, Florida AMI BERA, California ROGER W. MARSHALL, Kansas DONALD S. BEYER, JR., Virginia DEBBIE LESKO, Arizona EDDIE BERNICE JOHNSON, Texas LAMAR S.
    [Show full text]
  • FOR IMMEDIATE RELEASE: Tuesday, April 13, 2021 From: Matthew Zeiler, Clarifai Contact: [email protected]
    FOR IMMEDIATE RELEASE: Tuesday, April 13, 2021 From: Matthew Zeiler, Clarifai Contact: [email protected] https://www.clarifai.com Clarifai to Deliver AI/ML Algorithms on Palantir's Data Management Platform for U.S. Army's Ground Station Modernization Fort Lee, New Jersey - Clarifai, a leadinG AI lifecycle platform provider for manaGinG unstructured imaGe, video and text data, announced today that it has partnered with Palantir, Inc. (NYSE: PLTR) to deliver its Artificial IntelliGence and Machine LearninG alGorithms on Palantir’s data manaGement platform. This effort is part of the first phase of the U.S. Army’s Ground Station modernization in support of its Tactical IntelliGence TargetinG Access Node (TITAN) proGram. Palantir was awarded a Phase one Project Agreement throuGh an Other Transaction Agreement (OTA) with Consortium ManaGement Group, Inc. (CMG) on behalf of Consortium for Command, Control and Communications in Cyberspace (C5). DurinG the first phase of the initiative, Clarifai is supportinG Palantir by providinG object detection models to facilitate the desiGn of a Ground station prototype solution. “By providinG our Computer Vision AI solution to Palantir, Clarifai is helpinG the Army leveraGe spatial, hiGh altitude, aerial and terrestrial data sources for use in intelliGence and military operations,” said Dr. Matt Zeiler, Clarifai’s CEO. “The partnership will offer a ‘turnkey’ solution that inteGrates data from various sources, includinG commercial and classified sources from space to Ground sensors.” “In order to target with Great precision, we need to be able to see with Great precision, and that’s the objective of this proGram,” said Lieutenant General Robert P.
    [Show full text]
  • AI Computer Wraps up 4-1 Victory Against Human Champion Nature Reports from Alphago's Victory in Seoul
    The Go Files: AI computer wraps up 4-1 victory against human champion Nature reports from AlphaGo's victory in Seoul. Tanguy Chouard 15 March 2016 SEOUL, SOUTH KOREA Google DeepMind Lee Sedol, who has lost 4-1 to AlphaGo. Tanguy Chouard, an editor with Nature, saw Google-DeepMind’s AI system AlphaGo defeat a human professional for the first time last year at the ancient board game Go. This week, he is watching top professional Lee Sedol take on AlphaGo, in Seoul, for a $1 million prize. It’s all over at the Four Seasons Hotel in Seoul, where this morning AlphaGo wrapped up a 4-1 victory over Lee Sedol — incidentally, earning itself and its creators an honorary '9-dan professional' degree from the Korean Baduk Association. After winning the first three games, Google-DeepMind's computer looked impregnable. But the last two games may have revealed some weaknesses in its makeup. Game four totally changed the Go world’s view on AlphaGo’s dominance because it made it clear that the computer can 'bug' — or at least play very poor moves when on the losing side. It was obvious that Lee felt under much less pressure than in game three. And he adopted a different style, one based on taking large amounts of territory early on rather than immediately going for ‘street fighting’ such as making threats to capture stones. This style – called ‘amashi’ – seems to have paid off, because on move 78, Lee produced a play that somehow slipped under AlphaGo’s radar. David Silver, a scientist at DeepMind who's been leading the development of AlphaGo, said the program estimated its probability as 1 in 10,000.
    [Show full text]
  • Sustainable Development, Technological Singularity and Ethics
    European Research Studies Journal Volume XXI, Issue 4, 2018 pp. 714- 725 Sustainable Development, Technological Singularity and Ethics Vyacheslav Mantatov1, Vitaly Tutubalin2 Abstract: The development of modern convergent technologies opens the prospect of a new technological order. Its image as a “technological singularity”, i.e. such “transhuman” stage of scientific and technical progress, on which the superintelligence will be practically implemented, seems to be quite realistic. The determination of the basic philosophical coordinates of this future reality in the movement along the path of sustainable development of mankind is the most important task of modern science. The article is devoted to the study of the basic ontological, epistemological and moral aspects in the reception of the coming technological singularity. The method of this study is integrating dialectical and system approach. The authors come to the conclusion: the technological singularity in the form of a “computronium” (superintelligence) opens up broad prospects for the sustainable development of mankind in the cosmic dimension. This superintelligence will become an ally of man in the process of cosmic evolution. Keywords: Technological Singularity, Superintelligence, Convergent Technologies, Cosmocentrism, Human and Universe JEL code: Q01, Q56. 1East Siberia State University of Technology and Management, Ulan-Ude, Russia [email protected] 2East Siberia State University of Technology and Management, Ulan-Ude, Russia, [email protected] V. Mantatov, V. Tutubalin 715 1. Introduction Intelligence organizes the world by organizing itself. J. Piaget Technological singularity is defined as a certain moment or stage in the development of mankind, when scientific and technological progress will become so fast and complex that it will be unpredictable.
    [Show full text]
  • Future of Research on Catastrophic and Existential Risk
    ORE Open Research Exeter TITLE Working together to face humanity's greatest threats: Introduction to The Future of Research in Catastrophic and Existential Risk AUTHORS Currie, AM; Ó hÉigeartaigh, S JOURNAL Futures DEPOSITED IN ORE 06 February 2019 This version available at http://hdl.handle.net/10871/35764 COPYRIGHT AND REUSE Open Research Exeter makes this work available in accordance with publisher policies. A NOTE ON VERSIONS The version presented here may differ from the published version. If citing, you are advised to consult the published version for pagination, volume/issue and date of publication Working together to face humanity’s greatest threats: Introduction to The Future of Research on Catastrophic and Existential Risk. Adrian Currie & Seán Ó hÉigeartaigh Penultimate Version, forthcoming in Futures Acknowledgements We would like to thank the authors of the papers in the special issue, as well as the referees who provided such constructive and useful feedback. We are grateful to the team at the Centre for the Study of Existential Risk who organized the first Cambridge Conference on Catastrophic Risk where many of the papers collected here were originally presented, and whose multi-disciplinary expertise was invaluable for making this special issue a reality. We’d like to thank Emma Bates, Simon Beard and Haydn Belfield for feedback on drafts. Ted Fuller, Futures’ Editor-in-Chief also provided invaluable guidance throughout. The Conference, and a number of the publications in this issue, were made possible through the support of a grant from the Templeton World Charity Foundation (TWCF); the conference was also supported by a supplementary grant from the Future of Life Institute.
    [Show full text]
  • Scene Understanding for Mobile Robots Exploiting Deep Learning Techniques
    Scene Understanding for Mobile Robots exploiting Deep Learning Techniques José Carlos Rangel Ortiz Instituto Universitario de Investigación Informática Scene Understanding for Mobile Robots exploiting Deep Learning Techniques José Carlos Rangel Ortiz TESIS PRESENTADA PARA ASPIRAR AL GRADO DE DOCTOR POR LA UNIVERSIDAD DE ALICANTE MENCIÓN DE DOCTOR INTERNACIONAL PROGRAMA DE DOCTORADO EN INFORMÁTICA Dirigida por: Dr. Miguel Ángel Cazorla Quevedo Dr. Jesús Martínez Gómez 2017 The thesis presented in this document has been reviewed and approved for the INTERNATIONAL PhD HONOURABLE MENTION I would like to thank the advises and contributions for this thesis of the external reviewers: Professor Eduardo Mario Nebot (University of Sydney) Professor Cristian Iván Pinzón Trejos (Universidad Tecnológica de Panamá) This thesis is licensed under a CC BY-NC-SA International License (Cre- ative Commons AttributionNonCommercial-ShareAlike 4.0 International License). You are free to share — copy and redistribute the material in any medium or format Adapt — remix, transform, and build upon the ma- terial. The licensor cannot revoke these freedoms as long as you follow the license terms. Under the following terms: Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. NonCommercial — You may not use the material for commercial purposes. ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original. Document by José Carlos Rangel Ortiz. A mis padres. A Tango, Tomás y Julia en su memoria.
    [Show full text]
  • Privacy and Security
    Privacy and Security Sekar Kulandaivel, Jennifer Xiao - April 21, 2020 Understanding Contention-Based Channels and Using Them for Agenda Defense Spectre Attacks: Exploiting Speculative Execution Understanding Contention-Based Channels and Using Them for Defense (HPCA ‘15) Distrustful tenants living within a neutral cloud provider ● Shared hardware can be exploited to leak information ○ e.g. CPU usage vs. operation can expose secret key ● Two bodies of solutions: ○ HW-based: state-of-the-art is either limited in scope or requires impractical architecture changes ○ SW-based: HomeAlone forgoes shared hardware and permits only friendly co-residency, but still vulnerable to an intelligent attacker Threat model of a co-resident attacker ● Distrustful tenants violate confidentiality or compromise availability ● Goal: infer info about victim VM via microarchitectural structures e.g. cache and memory controllers ● Side-channel: victim inadvertently (oops!) leaks data inferred by attacker ● Covert channel: privileged malicious process on victim deliberately leaks data to attacker Known side-channels to transmit a ‘0’ or a ‘1’ (alt. exec.) ● Alternative execution attacks ○ Timing-driven: measure time to access memory portion ○ Access-driven: measure time to access specific cache misses Known side-channels to transmit a ‘0’ or a ‘1’ (parallel exec.) ● Parallel execution attacks ○ No time sharing required ○ E.g. Receiver monitors latency of memory fetch, sender either issues more instructions or idles Formal model of covert channels ● Detection failure (undetectable flow) = same rate of false positives and false negatives for both legitimate and covert traffic ● Network vs. microarchitectural channels: ○ Network receivers read silently ○ Microarch. receivers read destructively (overwrites when reading) ● Main insight: network channels are provably undetectable whereas microarch.
    [Show full text]
  • The Technological Singularity and the Transhumanist Dream
    ETHICAL CHALLENGES The technological singularity and the transhumanist dream Miquel Casas Araya Peralta In 1997, an AI beat a human world chess champion for the first time in history (it was IBM’s Deep Blue playing Garry Kasparov). Fourteen years later, in 2011, IBM’s Watson beat two winners of Jeopardy! (Jeopardy is a general knowledge quiz that is very popular in the United States; it demands a good command of the language). In late 2017, DeepMind’s AlphaZero reached superhuman levels of play in three board games (chess, go and shogi) in just 24 hours of self-learning without any human intervention, i.e. it just played itself. Some of the people who have played against it say that the creativity of its moves make it seem more like an alien that a computer program. But despite all that, in 2019 nobody has yet designed anything that can go into a strange kitchen and fry an egg. Are our machines truly intelligent? Successes and failures of weak AI The fact is that today AI can solve ever more complex specific problems with a level of reliability and speed beyond our reach at an unbeatable cost, but it fails spectacularly in the face of any challenge for which it has not been programmed. On the other hand, human beings have become used to trivialising everything that can be solved by an algorithm and have learnt to value some basic human skills that we used to take for granted, like common sense, because they make us unique. Nevertheless, over the last decade, some influential voices have been warning that our skills PÀGINA 1 / 9 may not always be irreplaceable.
    [Show full text]
  • In-Datacenter Performance Analysis of a Tensor Processing Unit
    In-Datacenter Performance Analysis of a Tensor Processing Unit Presented by Josh Fried Background: Machine Learning Neural Networks: ● Multi Layer Perceptrons ● Recurrent Neural Networks (mostly LSTMs) ● Convolutional Neural Networks Synapse - each edge, has a weight Neuron - each node, sums weights and uses non-linear activation function over sum Propagating inputs through a layer of the NN is a matrix multiplication followed by an activation Background: Machine Learning Two phases: ● Training (offline) ○ relaxed deadlines ○ large batches to amortize costs of loading weights from DRAM ○ well suited to GPUs ○ Usually uses floating points ● Inference (online) ○ strict deadlines: 7-10ms at Google for some workloads ■ limited possibility for batching because of deadlines ○ Facebook uses CPUs for inference (last class) ○ Can use lower precision integers (faster/smaller/more efficient) ML Workloads @ Google 90% of ML workload time at Google spent on MLPs and LSTMs, despite broader focus on CNNs RankBrain (search) Inception (image classification), Google Translate AlphaGo (and others) Background: Hardware Trends End of Moore’s Law & Dennard Scaling ● Moore - transistor density is doubling every two years ● Dennard - power stays proportional to chip area as transistors shrink Machine Learning causing a huge growth in demand for compute ● 2006: Excess CPU capacity in datacenters is enough ● 2013: Projected 3 minutes per-day per-user of speech recognition ○ will require doubling datacenter compute capacity! Google’s Answer: Custom ASIC Goal: Build a chip that improves cost-performance for NN inference What are the main costs? Capital Costs Operational Costs (power bill!) TPU (V1) Design Goals Short design-deployment cycle: ~15 months! Plugs in to PCIe slot on existing servers Accelerates matrix multiplication operations Uses 8-bit integer operations instead of floating point How does the TPU work? CISC instructions, issued by host.
    [Show full text]
  • Global Catastrophic Risks 2016
    Global Challenges Foundation Global Catastrophic Risks 2016 © Global Challenges Foundation/Global Priorities Project 2016 GLOBAL CATASTROPHIC RISKS 2016 THE GLOBAL CHALLENGES FOUNDATION works to raise awareness of the The views expressed in this report are those of the authors. Their Global Catastrophic Risks. Primarily focused on climate change, other en- statements are not necessarily endorsed by the affiliated organisations. vironmental degradation and politically motivated violence as well as how these threats are linked to poverty and rapid population growth. Against this Authors: background, the Foundation also works to both identify and stimulate the Owen Cotton-Barratt*† development of good proposals for a management model – a global gover- Sebastian Farquhar* nance – able to decrease – and at best eliminate – these risks. John Halstead* Stefan Schubert* THE GLOBAL PRIORITIES PROJECT helps decision-makers effectively prior- Andrew Snyder-Beattie† itise ways to do good. We achieve his both by advising decision-makers on programme evaluation methodology and by encouraging specific policies. We * = The Global Priorities Project are a collaboration between the Centre for Effective Altruism and the Future † = The Future of Humanity Institute, University of Oxford of Humanity Institute, part of the University of Oxford. Graphic design: Accomplice/Elinor Hägg Global Challenges Foundation in association with 4 Global Catastrophic Risks 2016 Global Catastrophic Risks 2016 5 Contents Definition: Global Foreword 8 Introduction 10 Catastrophic Risk Executive summary 12 1. An introduction to global catastrophic risks 20 – risk of events or 2. What are the most important global catastrophic risks? 28 Catastrophic climate change 30 processes that would Nuclear war 36 Natural pandemics 42 Exogenous risks 46 lead to the deaths of Emerging risks 52 Other risks and unknown risks 64 Our assessment of the risks 66 approximately a tenth of 3.
    [Show full text]