DOCSLIB.ORG

2021 Tech Trends Report Is De- It, Produced Two Factions: Those Signed to Help You Confront Deep Who Wanted to Reverse Time and Uncertainty, Adapt and Thrive

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
2021 Tech Trends Report Is De- It, Produced Two Factions: Those Signed to Help You Confront Deep Who Wanted to Reverse Time and Uncertainty, Adapt and Thrive Volume 1 of 12 14th Annual Edition 2021 Tech Trends Report Artificial Strategic trends that will influence business, Intelligence government, education, media and society in the coming year. 00 01 02 03 04 05 06 07 08 09 10 11 12 Artificial Intelligence 03 Overview 19 Algorithm Marketplaces 32 Machine Image Completion 43 Nation-based Guardrails 04 Macro Forces and Emerging Trends 19 100-Year Software 32 Predictive Models Using and Regulations 06 Summary 20 Scenario: Rage Against the Machine Single Images 43 Regulating Deepfakes 08 Artificial Intelligence 22 Health, Medicine, and Science 33 Model-free Approaches to RL 43 Making AI Explain Itself 09 An Executive’s Guide to AI 22 AI Speeds Scientific Discovery 33 Real-time Machine Learning 43 New Strategic Technical Alliances 09 Machine Learning 22 AI-First Drug Discovery 33 Automated Machine Learning 43 The New Mil-Tech (AutoML) Industrial Complex 09 Deep Learning 23 AI Improves Patient Outcomes 33 Hybrid Human-Computer Vision 44 Algorithmic Warfighting 10 Weak and Strong AI 23 Deep Learning Applied to 33 Neuro-Symbolic AI 12 Enterprise Medical Imaging 33 General Reinforcement 12 The Rise of MLOps 23 NLP Algorithms Detect Virus Mutations Learning Algorithms 12 Low-Code or No-Code 23 Diagnostics Without Tests 34 Continuous Learning Machine Learning 45 China’s AI Rules 34 Proliferation of 12 Web-Scale Content Analysis 23 Protein Folding Franken-Algorithms 48 Society 12 Simulating Empathy and Emotion 23 Dream Communication 34 Proprietary, Homegrown 48 Ethics Clash 24 Thought Detection 13 Artificial Emotional Intelligence AI Languages 48 Ambient Surveillance 25 Scenario: Deep Twins in the OR 13 Serverless Computing 36 Talent 48 Marketplace Consolidation 27 Consumer 14 Expert Insight: Emotion AI Will 36 AI Brain Drain 48 Fragmentation Power the Empathy Economy, 27 Zero UIs 36 AI Universities 49 Expert Insight: AI reveals but AI Still Needs to Work 27 Consumer-grade AI Applications 37 Demand for AI Talent Growing our real-world biases 17 AI in the Cloud 27 Ubiquitous Digital Assistants 37 Corporate AI Labs 50 AI Still Has a Bias Problem 17 AI at the Edge 28 Deepfakes for Fun 37 AI for Interviews 50 Problematic Training Data 17 Advanced AI Chipsets 28 Personal Digital Twins 39 Creative 50 AI to Catch Cheaters 17 Digital Twins 30 Research 39 Assisted Creativity 50 Algorithms Targeting 17 Spotting Fakes 30 Closed-Source Code Vulnerable Populations 39 Generative Algorithms for 18 Natural Language Processing 30 Framework Consolidation Content Production 51 AI Intentionally Hiding Data for ESGs 30 Cost of Training Models 39 Generating Virtual Environments 51 Undocumented AI Accidents 18 Intelligent Optical Character 31 NLP Benchmarks from Short Videos 51 Digital Dividends Recognition 31 Machine Reading Comprehension 40 Automated Versioning 51 Prioritizing Trust 18 Robotic Process Automation 31 AI Summarizing Itself 40 Automatic Voice Cloning 52 Scenario: Bully Bots 18 Massive Translation Systems 31 No Retraining Required and Dubbing 53 Application 19 Predicting Systems and 40 Automatic Ambient Noise Dubbing Site Failures 31 Graph Neural Networks 54 Key Questions 42 Geopolitics and Defense 19 Liability Insurance for AI 31 Federated Learning 55 Sources 42 AI Nationalism 19 Manipulating AI Systems for 31 GP Models 56 Authors Competitive Advantage 31 GPT-3’s Influence 42 National AI Strategies 19 Global Rush to Fund AI 32 Vokenization 42 AI as Critical Infrastructure 00 01 02 03 04 05 06 07 08 09 10 11 12 Artificial Intelligence Overview The 1920s began in chaos. Cata- It’s difficult not to see striking ed trends. In total, we’ve analyzed clysmic disruption resulting from parallels to our modern world. A nearly 500 technology and science the first world war and the Spanish tumultuous U.S. election, extreme trends across multiple industry flu shuttered businesses and pro- weather events and Covid-19 sectors. In each volume, we discuss voked xenophobia. Technological continue to test our resolve and the disruptive forces, opportunities marvels like the radio, refrigerator, our resilience. Exponential tech- and strategies that will drive your vacuum cleaner, moving assembly nologies—artificial intelligence, organization in the near future. line and electronic power trans- synthetic biology, exascale com- Now, more than ever, your organi- mission generated new growth, puting, autonomous robots, and zation should examine the poten- even as the wealth gap widened. off-planet missions to space—are tial near and long-term impact of More than two-thirds of Ameri- challenging our assumptions about tech trends. You must factor the cans survived on wages too low to human potential. Under lockdown, trends in this report into your stra- sustain everyday living. The pace we’ve learned how to work from tegic thinking for the coming year, of scientific innovation—the dis- our kitchen tables, lead from our and adjust your planning, opera- covery of insulin, the first modern spare rooms, and support each tions and business models accord- antibiotics, and insights into theo- other from afar. But this disruption ingly. But we hope you will make retical physics and the structure of has only just begun. time for creative exploration. From atoms—forced people to reconsid- With the benefit of both hindsight chaos, a new world will come. er their cherished beliefs. and strategic foresight, we can The sheer scale of change, and the choose a path of reinvention. Our great uncertainty that came with 2021 Tech Trends Report is de- it, produced two factions: those signed to help you confront deep who wanted to reverse time and uncertainty, adapt and thrive. For Amy Webb return the world to normal, and this year’s edition, the magnitude Founder those who embraced the chaos, of new signals required us to cre- The Future Today Institute faced forward, and got busy build- ate 12 separate volumes, and each ing the future. report focuses on a cluster of relat- 03 © 2021 Future Today Institute 00 01 02 03 04 05 06 07 08 09 10 11 12 6 1 Macro Forces and Emerging Trends 2 3 4 5 For nearly two decades, the Future Today Institute has meticulously re- searched macro forces of change and the emerging trends that result. Our focus: understanding how these forces and trends will shape our futures. Our 14th annual Tech Trends Report identifies new opportunities for growth and potential collaborations in and adjacent to your business. We also highlight emerging or atypical threats across most industries, including all levels of government. For those in creative fields, you will find a wealth of new ideas that will spark your imagination. Our framework organizes nearly 500 trends into 12 clear categories. Within those categories are specific use cases and recommendations for key roles in many organizations: strategy, innovation, R&D, and risk. Each trend offers six important insights. 1. Years on the List 2. Key Insight 4. Disruptive Impact 6. Action Scale Informs Strategy We track longitudinal tech and Concise description of this trend The implications of this trend on FTI’s analysis of what action your Strong evidence and data. Longer- science trends. This measurement that can be easily understood and your business, government, or organization should take. Fields term uncertainties remain. Use it to indicates how long we have repeated to others. society. include: inform your strategic planning. followed the trend and its progression. 3. Examples 5. Emerging Players Watch Closely Act Now Real-world use cases, some of Individuals, research teams, Mounting evidence and data, but Ample evidence and data. This which should be familiar to you. startups, and other organizations more maturity is needed. Use it to trend is already mature and emerging in this space. inform your vision, planning, and requires action. research. 04 © 2021 Future Today Institute 00 01 02 03 04 05 06 07 08 09 10 11 12 Artificial Intelligence Macro Forces and Emerging Trends Scenarios Describe Plausible Outcomes You will find scenarios imagining future worlds as trends evolve and converge. Scenarios offer a fresh perspective on trends and often chal- lenge your deeply held beliefs. They prompt you to consider high-impact, high-uncertainty situations using signals available today. 1 1. Headline 2 A short description offering you a glimpse into future changes. 2. Temporal and Emotive Tags 3 A label explaining both when in the future this scenario is set and whether it is optimistic, neutral, pessimistic, or catastrophic. 3. Narrative The descriptive elements of our imagined world, including the developments leading us to this point in our future history. Scenario sources: The Future Today Institute uses a wide array of quali- tative and quantitative data to create our scenarios. Some of our typical sources include patent filings, academic preprint servers, archival re- search, policy briefings, conference papers, data sets, structured inter- views with experts, conversations with kids, critical design, and specula- tive fiction. 05 © 2021 Future Today Institute 00 01 02 03 04 05 06 07 08 09 10 11 12 Artificial Intelligence + Natural language processing is + Natural language processing an area experiencing high inter- algorithms— typically used for est, investment, and growth. text, words, and sentences—are being used to interpret genetic + No-code or low-code systems changes in viruses. are unlocking new use cases for businesses. + COVID-19 accelerated the use of AI in drug discovery last year. The + Amazon Web Services, Azure, first trial of an AI-discovered drug and Google Cloud’s low-code is underway in Japan. and no-code offerings will trickle down to everyday people, al- + AI plays key roles in synthetic Artificial lowing them to create their own biology, genetics, and medical artificial intelligence applica- imaging; predicting the spread of tions and deploy them as easily disease; and improving patient as they could a website.
Load more

Recommended publications
  • Artificial Intelligence and Big Data – Innovation Landscape Brief
    ARTIFICIAL INTELLIGENCE AND BIG DATA INNOVATION LANDSCAPE BRIEF © IRENA 2019 Unless otherwise stated, material in this publication may be freely used, shared, copied, reproduced, printed and/or stored, provided that appropriate acknowledgement is given of IRENA as the source and copyright holder. Material in this publication that is attributed to third parties may be subject to separate terms of use and restrictions, and appropriate permissions from these third parties may need to be secured before any use of such material. ISBN 978-92-9260-143-0 Citation: IRENA (2019), Innovation landscape brief: Artificial intelligence and big data, International Renewable Energy Agency, Abu Dhabi. ACKNOWLEDGEMENTS This report was prepared by the Innovation team at IRENA’s Innovation and Technology Centre (IITC) with text authored by Sean Ratka, Arina Anisie, Francisco Boshell and Elena Ocenic. This report benefited from the input and review of experts: Marc Peters (IBM), Neil Hughes (EPRI), Stephen Marland (National Grid), Stephen Woodhouse (Pöyry), Luiz Barroso (PSR) and Dongxia Zhang (SGCC), along with Emanuele Taibi, Nadeem Goussous, Javier Sesma and Paul Komor (IRENA). Report available online: www.irena.org/publications For questions or to provide feedback: [email protected] DISCLAIMER This publication and the material herein are provided “as is”. All reasonable precautions have been taken by IRENA to verify the reliability of the material in this publication. However, neither IRENA nor any of its officials, agents, data or other third- party content providers provides a warranty of any kind, either expressed or implied, and they accept no responsibility or liability for any consequence of use of the publication or material herein.
    [Show full text]
  • Machine Theory of Mind
    Machine Theory of Mind Neil C. Rabinowitz∗ Frank Perbet H. Francis Song DeepMind DeepMind DeepMind [email protected] [email protected] [email protected] Chiyuan Zhang S. M. Ali Eslami Matthew Botvinick Google Brain DeepMind DeepMind [email protected] [email protected] [email protected] Abstract 1. Introduction Theory of mind (ToM; Premack & Woodruff, For all the excitement surrounding deep learning and deep 1978) broadly refers to humans’ ability to rep- reinforcement learning at present, there is a concern from resent the mental states of others, including their some quarters that our understanding of these systems is desires, beliefs, and intentions. We propose to lagging behind. Neural networks are regularly described train a machine to build such models too. We de- as opaque, uninterpretable black-boxes. Even if we have sign a Theory of Mind neural network – a ToM- a complete description of their weights, it’s hard to get a net – which uses meta-learning to build models handle on what patterns they’re exploiting, and where they of the agents it encounters, from observations might go wrong. As artificial agents enter the human world, of their behaviour alone. Through this process, the demand that we be able to understand them is growing it acquires a strong prior model for agents’ be- louder. haviour, as well as the ability to bootstrap to Let us stop and ask: what does it actually mean to “un- richer predictions about agents’ characteristics derstand” another agent? As humans, we face this chal- and mental states using only a small number of lenge every day, as we engage with other humans whose behavioural observations.
    [Show full text]
  • Efficiently Mastering the Game of Nogo with Deep Reinforcement
    electronics Article Efficiently Mastering the Game of NoGo with Deep Reinforcement Learning Supported by Domain Knowledge Yifan Gao 1,*,† and Lezhou Wu 2,† 1 College of Medicine and Biological Information Engineering, Northeastern University, Liaoning 110819, China 2 College of Information Science and Engineering, Northeastern University, Liaoning 110819, China; [email protected] * Correspondence: [email protected] † These authors contributed equally to this work. Abstract: Computer games have been regarded as an important field of artificial intelligence (AI) for a long time. The AlphaZero structure has been successful in the game of Go, beating the top professional human players and becoming the baseline method in computer games. However, the AlphaZero training process requires tremendous computing resources, imposing additional difficulties for the AlphaZero-based AI. In this paper, we propose NoGoZero+ to improve the AlphaZero process and apply it to a game similar to Go, NoGo. NoGoZero+ employs several innovative features to improve training speed and performance, and most improvement strategies can be transferred to other nonspecific areas. This paper compares it with the original AlphaZero process, and results show that NoGoZero+ increases the training speed to about six times that of the original AlphaZero process. Moreover, in the experiment, our agent beat the original AlphaZero agent with a score of 81:19 after only being trained by 20,000 self-play games’ data (small in quantity compared with Citation: Gao, Y.; Wu, L. Efficiently 120,000 self-play games’ data consumed by the original AlphaZero). The NoGo game program based Mastering the Game of NoGo with on NoGoZero+ was the runner-up in the 2020 China Computer Game Championship (CCGC) with Deep Reinforcement Learning limited resources, defeating many AlphaZero-based programs.
    [Show full text]
  • MUSTAFA SULEYMAN, DEEPMIND INTRO Thank You Mr. President and Your Excellencies for the Invitation to Speak to You About What Ar
    MUSTAFA SULEYMAN, DEEPMIND INTRO Thank you Mr. President and your excellencies for the invitation to speak to you about what artificial intelligence can do to help achieve the Sustainable Development Goals. As a global community, we’ve made significant strides over the last few decades, taking on some of the world’s greatest tragedies like child mortality. Every single day 17,000 new lives get to be lived, by children who would have died just 25 years ago. Peaceful cooperation and relentless innovation has been the driving force of this truly spectacular progress. And, yet, in many ways inequality hasn’t improved—it’s become worse. Malnutrition and preventable disease continue to kill millions, impacting both developed and developing world healthcare systems, and the devastating threat of climate change looms, with the effects set to hit the poorest the hardest. If we are to decrease this level of suffering, then we’ll have to make sure humanity is doing everything possible to come up with bold, new solutions. But to have any chance of truly solving these problems - one of the great goals of human civilisation and I would argue ​ core to the founding purpose of this great institution - then what’s possible won’t be ​ enough. We must take on what is currently impossible, and do all we can to push back the ​ ​ limits of what humanity can achieve. These limits are real, and they cap all of our aspirations for change. Our efforts to tackle disease are capped by a desperate shortage of trained nurses and doctors — in the rich West as well as in developing nations.
    [Show full text]
  • AI in Focus - Fundamental Artificial Intelligence and Video Games
    AI in Focus - Fundamental Artificial Intelligence and Video Games April 5, 2019 By Isi Caulder and Lawrence Yu Patent filings for fundamental artificial intelligence (AI) technologies continue to rise. Led by a number of high profile technology companies, including IBM, Google, Amazon, Microsoft, Samsung, and AT&T, patent applications directed to fundamental AI technologies, such as machine learning, neural networks, natural language processing, speech processing, expert systems, robotic and machine vision, are being filed and issued in ever-increasing numbers.[1] In turn, these fundamental AI technologies are being applied to address problems in industries such as healthcare, manufacturing, and transportation. A somewhat unexpected source of fundamental AI technology development has been occurring in the field of video games. Traditional board games have long been a subject of study for AI research. In the 1990’s, IBM created an AI for playing chess, Deep Blue, which was able to defeat top-caliber human players using brute force algorithms.[2] More recently, machine learning algorithms have been developed for more complex board games, which include a larger breadth of possible moves. For example, DeepMind (since acquired by Google), recently developed the first AI capable of defeating professional Go players, AlphaGo.[3] Video games have recently garnered the interest of researchers, due to their closer similarity to the “messiness” and “continuousness” of the real world. In contrast to board games, video games typically include a greater
    [Show full text]
  • Benchmarking Safe Exploration in Deep Reinforcement Learning
    Benchmarking Safe Exploration in Deep Reinforcement Learning Alex Ray∗ Joshua Achiam∗ Dario Amodei OpenAI OpenAI OpenAI Abstract Reinforcement learning (RL) agents need to explore their environments in order to learn optimal policies by trial and error. In many environments, safety is a critical concern and certain errors are unacceptable: for example, robotics systems that interact with humans should never cause injury to the humans while exploring. While it is currently typical to train RL agents mostly or entirely in simulation, where safety concerns are minimal, we anticipate that challenges in simulating the complexities of the real world (such as human-AI interactions) will cause a shift towards training RL agents directly in the real world, where safety concerns are paramount. Consequently we take the position that safe exploration should be viewed as a critical focus area for RL research, and in this work we make three contributions to advance the study of safe exploration. First, building on a wide range of prior work on safe reinforcement learning, we propose to standardize constrained RL as the main formalism for safe exploration. Second, we present the Safety Gym benchmark suite, a new slate of high-dimensional continuous control environments for measuring research progress on constrained RL. Finally, we benchmark several constrained deep RL algorithms on Safety Gym environments to establish baselines that future work can build on. 1 Introduction Reinforcement learning is an increasingly important technology for developing highly-capable AI systems. While RL is not yet fully mature or ready to serve as an “off-the-shelf” solution, it appears to offer a viable path to solving hard sequential decision-making problems that cannot currently be solved by any other approach.
    [Show full text]
  • Arxiv:1911.05531V1 [Q-Bio.BM] 9 Nov 2019 1 Introduction
    Accurate Protein Structure Prediction by Embeddings and Deep Learning Representations Iddo Drori1;2, Darshan Thaker1, Arjun Srivatsa1, Daniel Jeong1, Yueqi Wang1, Linyong Nan1, Fan Wu1, Dimitri Leggas1, Jinhao Lei1, Weiyi Lu1, Weilong Fu1, Yuan Gao1, Sashank Karri1, Anand Kannan1, Antonio Moretti1, Mohammed AlQuraishi3, Chen Keasar4, and Itsik Pe’er1 1 Columbia University, Department of Computer Science, New York, NY, 10027 2 Cornell University, School of Operations Research and Information Engineering, Ithaca, NY 14853 3 Harvard University, Department of Systems Biology, Harvard Medical School, Boston, MA 02115 4 Ben-Gurion University, Department of Computer Science, Israel, 8410501 Abstract. Proteins are the major building blocks of life, and actuators of almost all chemical and biophysical events in living organisms. Their native structures in turn enable their biological functions which have a fun- damental role in drug design. This motivates predicting the structure of a protein from its sequence of amino acids, a fundamental problem in com- putational biology. In this work, we demonstrate state-of-the-art protein structure prediction (PSP) results using embeddings and deep learning models for prediction of backbone atom distance matrices and torsion angles. We recover 3D coordinates of backbone atoms and reconstruct full atom protein by optimization. We create a new gold standard dataset of proteins which is comprehensive and easy to use. Our dataset consists of amino acid sequences, Q8 secondary structures, position specific scoring matrices, multiple sequence alignment co-evolutionary features, backbone atom distance matrices, torsion angles, and 3D coordinates. We evaluate the quality of our structure prediction by RMSD on the latest Critical Assessment of Techniques for Protein Structure Prediction (CASP) test data and demonstrate competitive results with the winning teams and AlphaFold in CASP13 and supersede the results of the winning teams in CASP12.
    [Show full text]
  • Memory-Based Control with Recurrent Neural Networks
    Memory-based control with recurrent neural networks Nicolas Heess* Jonathan J Hunt* Timothy P Lillicrap David Silver Google Deepmind * These authors contributed equally. heess, jjhunt, countzero, davidsilver @ google.com Abstract Partially observed control problems are a challenging aspect of reinforcement learning. We extend two related, model-free algorithms for continuous control – deterministic policy gradient and stochastic value gradient – to solve partially observed domains using recurrent neural networks trained with backpropagation through time. We demonstrate that this approach, coupled with long-short term memory is able to solve a variety of physical control problems exhibiting an as- sortment of memory requirements. These include the short-term integration of in- formation from noisy sensors and the identification of system parameters, as well as long-term memory problems that require preserving information over many time steps. We also demonstrate success on a combined exploration and mem- ory problem in the form of a simplified version of the well-known Morris water maze task. Finally, we show that our approach can deal with high-dimensional observations by learning directly from pixels. We find that recurrent deterministic and stochastic policies are able to learn similarly good solutions to these tasks, including the water maze where the agent must learn effective search strategies. 1 Introduction The use of neural networks for solving continuous control problems has a long tradition. Several recent papers successfully apply model-free, direct policy search methods to the problem of learning neural network control policies for challenging continuous domains with many degrees of freedoms [2, 6, 14, 21, 22, 12]. However, all of this work assumes fully observed state.
    [Show full text]
  • Neurips 2020 Workshop Book
    NeurIPS 2020 Workshop book Schedule Highlights Generated Fri Dec 18, 2020 Workshop organizers make last-minute changes to their schedule. Download this document again to get the lastest changes, or use the NeurIPS mobile application. Dec. 10, 2020 None Topological Data Analysis and Beyond Rieck, Chazal, Krishnaswamy, Kwitt, Natesan Ramamurthy, Umeda, Wolf Dec. 11, 2020 None Privacy Preserving Machine Learning - PriML and PPML Joint Edition Balle, Bell, Bellet, Chaudhuri, Gascon, Honkela, Koskela, Meehan, Ohrimenko, Park, Raykova, Smart, Wang, Weller None Tackling Climate Change with ML Dao, Sherwin, Donti, Kuntz, Kaack, Yusuf, Rolnick, Nakalembe, Monteleoni, Bengio None Meta-Learning Wang, Vanschoren, Grant, Schwarz, Visin, Clune, Calandra None OPT2020: Optimization for Machine Learning Paquette, Schmidt, Stich, Gu, Takac None Advances and Opportunities: Machine Learning for Education Garg, Heffernan, Meyers None Differential Geometry meets Deep Learning (DiffGeo4DL) Bose, Mathieu, Le Lan, Chami, Sala, De Sa, Nickel, Ré, Hamilton None Workshop on Dataset Curation and Security Baracaldo Angel, Bisk, Blum, Curry, Dickerson, Goldblum, Goldstein, Li, Schwarzschild None Machine Learning for Health (ML4H): Advancing Healthcare for All Hyland, Schmaltz, Onu, Nosakhare, Alsentzer, Chen, McDermott, Roy, Akera, Kiyasseh, Falck, Adams, Bica, Bear Don't Walk IV, Sarkar, Pfohl, Beam, Beaulieu-Jones, Belgrave, Naumann None Learning Meaningful Representations of Life (LMRL.org) Wood, Marks, Jones, Dieng, Aspuru-Guzik, Kundaje, Engelhardt, Liu, Boyden, Lindorff-Larsen,
    [Show full text]
  • The Dawn of Artificial Intelligence”
    Statement of Greg Brockman Co-Founder and Chief Technology Officer OpenAI before the Subcommittee on Space, Science, and Competitiveness Committee on Commerce, Science, and Transportation United States Senate Hearing on “The Dawn of Artificial Intelligence” Thank you Chairman Cruz, Ranking Member Peters, distinguished members of the Subcommittee. Today’s hearing presents an important first opportunity for the members of the Senate to understand and analyze the potential impacts of artificial intelligence on our Nation and the world, and to refine thinking on the best ways in which the U.S. Government might approach AI. I'm honored to have been invited to give this testimony today. By way of introduction, I'm Greg Brockman, co-founder and Chief Technology Officer of OpenAI. OpenAI is a non-profit AI research company. Our mission is to build safe, advanced AI technology and ensure that its benefits are distributed to everyone. OpenAI is chaired by technology executives Sam Altman and Elon Musk. The US has led the way in almost all technological breakthroughs of the last hundred years, and we've reaped enormous economic rewards as a result. Currently, we have a lead, but hardly a monopoly, in AI. For instance, this year Chinese teams won the top categories in a Stanford University-led image recognition competition. South Korea has declared a billion dollar AI fund. Canada produced some technologies enabling the current boom, and recently announced an investment into key areas of AI. I'd like to share 3 key points for how we can best succeed in AI and what the U.S.
    [Show full text]
  • Deep Reinforcement Learning from Human Preferences
    Deep Reinforcement Learning from Human Preferences Paul F Christiano Jan Leike Tom B Brown OpenAI DeepMind Google Brain⇤ [email protected] [email protected] [email protected] Miljan Martic Shane Legg Dario Amodei DeepMind DeepMind OpenAI [email protected] [email protected] [email protected] Abstract For sophisticated reinforcement learning (RL) systems to interact usefully with real-world environments, we need to communicate complex goals to these systems. In this work, we explore goals defined in terms of (non-expert) human preferences between pairs of trajectory segments. We show that this approach can effectively solve complex RL tasks without access to the reward function, including Atari games and simulated robot locomotion, while providing feedback on less than 1% of our agent’s interactions with the environment. This reduces the cost of human oversight far enough that it can be practically applied to state-of-the-art RL systems. To demonstrate the flexibility of our approach, we show that we can successfully train complex novel behaviors with about an hour of human time. These behaviors and environments are considerably more complex than any which have been previously learned from human feedback. 1 Introduction Recent success in scaling reinforcement learning (RL) to large problems has been driven in domains that have a well-specified reward function (Mnih et al., 2015, 2016; Silver et al., 2016). Unfortunately, many tasks involve goals that are complex, poorly-defined, or hard to specify. Overcoming this limitation would greatly expand the possible impact of deep RL and could increase the reach of machine learning more broadly. For example, suppose that we wanted to use reinforcement learning to train a robot to clean a table or scramble an egg.
    [Show full text]
  • Deepmind's Demis Hassabis Inspires London Schoolchildren
    PRESS RELEASE DeepMind ’s Demis Hassabis inspires London schoolchildren John Saunders reports: Demis Hassabis, co-founder of the leading artificial intelligence company DeepMind, now part of Google ’s Alpha Group, paid a visit to UCL Academy, a secondary school in Camden, London, on 8 November to talk to the academy students about Artificial Intelligence and give a simul with a difference in which the opponents could consult a computer engine during play. Though now more famous for his pioneering work in the field of neuroscience-inspired AI, Demis Hassabis is also a chess player of some renown, having reached a rating of 2300 aged just 13 on FIDE ’s January 1990 list at a time when the only player of his age group rated higher than him was Judit Polgar. He also has a connection with UCL Academy in that he studied for his doctorate in cognitive neuro-science at University College London, which sponsors the school. Demis touched on the famous Kasparov-Deep Blue man versus machine encounters and how information from these legendary encounters had fed into his planning for a computer program to play Go. Reluctant though we die-hard chessers sometimes are to admit it, we have to acknowledge that this 3000-year-old game from China, though ostensibly simple in terms of the rules, is a quantum leap more complex to play than its 64-square rival. The Alpha Group at Google ’s London HQ had no illusions in the task they were taking on. A Go-playing program couldn ’t simply ape chess engines because it was impossible to build an evaluation function that would match the depth and complexity of the game – so AI was the only approach.
    [Show full text]