Special Report of National Commission for Intelligent Systems Subcommittee on Regulation and Ethics in Artificial Intelligence
The field of artificial intelligence has become a significant part of the technology and general industry and the proof of this is that huge companies from all over the world are investing huge sums in developments based on artificial intelligence technologies. At the same time, the field of artificial intelligence has become a very important issue in the general strategy of various countries around the world, for the reason that no country wants to be left behind and lose strategic superiority with proven technological advantages. The State of Israel has realized that it must formulate its own strategic plan in the field of artificial intelligence, enabling it to formulate a national strategy for the coming years and ensure the continued prosperity of Israel.
After a long wait, we received approval for public publication and I am pleased to share a special report by the National Subcommittee on Artificial Intelligence and Regulation. Professors Yitzhak Ben-Israel and Evitar Matanya were responsible for producing the report detailing the State of Israel’s strategy in the field of artifical intelligence. In addition, the report offers concrete recommendations for implementation. The purpose of the report is to ensure that the State of Israel remains relevant in the technological arms race and continue to be a global leader in technological innovation. The full report has not yet been approved for publication, but there is an interim report now available to the general public.
The chairman of the committee is Prof. Karine Nahon and a dignified and earnest representation of experts from the country who have devoted quite a bit of their time to producing the report.
The report before you outlines how it is built as well as the main issues that have come up in committee discussions. The aim was to provide a practical look at each issue and provide essential tools for companies and organizations that want to make sure they do not create ethical issues during the development process.
You can also get a complete snapshot of all artificial intelligence in the world through Appendix A, which provides a brief overview of what is happening in the US and Europe and how they have defined their strategy in this area. It is recommended that you read the executive summary at the beginning of the report to understand the content of the chapters and committee recommendations.
The report consists of technological issues related to ethics and regulation in the field of artificial intelligence and is written in clear and accessible language and you will find three main chapters dealing with the whole issue:
The first chapter deals with the unique characteristics of artificial intelligence technologies, as well as the question of ethical and legal disagreements as to whether the field of artificial intelligence needs special regulatory and ethical considerations.
The second chapter deals with ethics in the use of artificial intelligence based technologies. There is a serious reference that defines familiar concepts from the content world that allow us to identify places where ethical issues can arise in product development, such as fairness, transparency, safety and more. In addition, the committee has developed a tool that can help decision makers in organizations understand in which cases they may encounter and fail ethical issues and how they can be avoided using the developed tool. In order to test the tool, a number of case studies were reviewed in which targeted ethical violations were carried out on various projects and demonstrated how the tool can be used in these types of test cases. The tool can assist both decision makers and developers during the development of different artificial intelligence systems.
The third chapter deals with the different options for the artificial intelligence regulations various aspects in the fields of ethical and legal in the field of artificial intelligence and more. In this chapter you will find a table that lists the different types of regulators as well as the pros and cons of each way of action.
The entire report is here>>
AIEthicsRegulationReportHebrew
Tags: A/B testing, Abstract, academia, acceleration, Accuracy, action, activation function, active learning, AdaGrad, administrative perspectives, agent, agglomerative clustering, AI, AI methods, AI system, Algorithms, analytics, analytics space, and not the other way around, application design, applied machine learning, ar, area under the PR curve, area under the ROC curve, artificial general intelligence, artificial intelligence, artificial intelligence-based systems, association, attention, attribute, AUC (Area under the ROC Curve), Augmented Reality, automation bias, Automation of Machine-Learning, Autonomous driving, available, average precision, backpropagation, bag of words, baseline, batch, batch normalization, batch size, Bayesian, Bayesian neural network, behavior, Bellman equation, bias (ethics/fairness), bias (math), biases, Big Data Analytics, bigram, binary classification, binning, boosting, broad spectrum, broadcasting, bucketing, calibration layer, candidate generation, candidate sampling, categorical data, centroid, centroid-based clustering, checkpoint, class, class-imbalanced dataset, classification, classification model, classification threshold, clipping, Cloud TPU, clustering, co-adaptation, collaborative filtering, comments, Community, Computational neural networks, computer vision, Conference, confirmation bias, confusion matrix, content, context, Continuous delivery, continuous feature, control, convenience sampling, convergence, convex function, convex optimization, convex set, convolution, convolutional filter, convolutional layer, convolutional layers, convolutional neural network, convolutional operation, core principles, cost, counterfactual fairness, coverage bias, crash blossom, critic, Cross validation, cross-entropy, custom Estimator, Cyber, data, data analysis, data augmentation, Data Quality, Data Science, Data Scientists, data set or dataset, database, DataFrame, Dataset API (tf.data), datasets, decision boundary, decision maker, decision making, decision threshold, decision tree, decision trees, decisions, deep learning, deep model, deep neural network, Deep Q-Network (DQN), delivering, demographic parity, dense feature, dense layer, depth, depthwise separable convolutional neural network (sepCNN), DESIGN, designed, Developers, device, DevOps, Dialogue Bots, different hyperparameters, different initializations, different overall structure, dimension reduction, dimensions, discrete feature, discriminative model, discriminator, disparate impact, disparate treatment, divisive clustering, domains, downsampling, DQN, dropout regularization, dynamic model, eager execution, early stopping, economic and social obstacles, Education, effective removal, embedding space, embeddings, empirical risk minimization (ERM), engineering, ensemble, environment, equal groups, equal treatment, errors, especially, establish, ethical professional duty, Ethics of artificial intelligence, events, Evitar Matanya, Evitar Matanya Prof. Karine Nahon, excellent, experience, experimenter’s bias, Explaining of Israel, fine tuning, Fintech, focus, forget gate, formal, framework, full softmax, fully autonomous, fully connected layer, Future of AI, GAN, generalization, generalization curve, generalized linear model, generative adversarial network (GAN), generative model, generator, gradient, gradient clipping, gradient descent, Gradient descent algorithm, graph, graph execution, great, great success, greedy policy, ground truth, group attribution bias, hashing, Healthcare, helpful feedback, heuristic, hidden layer, hierarchical clustering, high-quality, highlight, hinge loss, holdout data, human, hyperparameter, hyperplane, i.i.d., ideas, image recognition, imbalanced dataset, impact, Implementation of principles, implicit bias, improvement, in every context, in-group bias, including, incompatibility of fairness metrics, increasing, increasingly, independently and identically distributed (i.i.d), individual fairness, individuals, industry, industry tracks, Inference, innovation, innovative, input function, input layer, instance, Intelligent robots, inter-rater agreement, interpretability, iot, item matrix, items, iteration, k-means, k-median, Keras, Kernel Support Vector Machines (KSVMs), label, labeled example, lambda, Large scale analytics, layer, Layers API (tf.layers), leading experts, learning rate, least squares regression, liberty, linear model, linear regression, Log Loss, log-odds, logistic regression, logits, Long Short-Term Memory (LSTM), loss, loss curve, loss surface, LSTM, Machine ethics, MACHINE LEARNING, Machine learning systems, majority class, Markov decision process (MDP), Markov property, matplotlib, matrix factorization, Mean Absolute Error (MAE), Mean Squared Error (MSE), metric, Metrics API (tf.metrics), mini-batch, mini-batch stochastic gradient descent (SGD), minimax loss, minority class, ML, MNIST, model, model function, model training, Momentum, momentum (Momentum), moral and ethical use of AI technologies, more complex math (Proximal and others), multi-class classification, multi-class logistic regression, multi-class regression, multinomial classification, N-gram, NaN trap, Natural language processing, Natural Language Understanding, negative class, neural network, neural networks, neuron, Neutrality, new moral and ethical challenges, new regulatory, NLU, node (neural network), node (TensorFlow graph), noise, non-arbitrary, non-response bias, normalization, Nowadays, numerical data, NumPy, objective, objective function, offline inference, one-hot encoding, one-shot learning, one-vs.-all, online inference, Operation (op), optimizer, organizing conference, out-group homogeneity bias, outliers, output layer, Overfitting, pandas, parameter, Parameter Server (PS), parameter update, partial derivative, participants, participation bias, partitioning strategy, perceptron, performance, perplexity, pipeline, policy, pooling, population, positive class, post-processing, powerful, PR AUC (area under the PR curve), pre-trained model, precision, precision-recall curve, prediction, prediction bias, predictive applications, predictive parity, predictive rate parity, predicts, premade Estimator, preprocessing, principles, prior belief, procedural, process data, Prof. Karine Nahon, Professors, proxy (sensitive attributes), proxy labels, public accountability, public sector, public sectors, Q-function, Q-learning, quantile, quantile bucketing, quantization, queue, random forest, random policy, rank (ordinality), rank (Tensor), rater, re-ranking, real world, real-world domains, reason, recall, recommendation system, reconstruction, Rectified Linear Unit (ReLU), recurrent neural network, Registration, regression, regression model, regularization, regularization rate, reinforcement learning, reinforcement learning (RL), replay buffer, reporting bias, representation, research and application, research innovations, research track, researchers, respect, Responsibility, results, Retail, return, reward, ridge regularization, RNN, Robot rights, robotics, ROC (receiver operating characteristic) Curve, root directory, Root Mean Squared Error (RMSE), rotational invariance, sampling bias, Satisfying certain basic universal needs, SavedModel, Saver, scalar, scaling, scikit-learn, scoring, selection bias, semi-supervised learning, sensitive attribute, sequence model, services, serving, session (tf.session), shape (Tensor), sigmoid function, significant capital, similarity measure, size invariance, sketching, society, softmax, sparse feature, sparse representation, sparse vector, sparsity, sparsity/regularization (Ftrl), spatial pooling, sponsorship, squared hinge loss, squared loss, State, state-action value function, state-of-the-art, static model, stationarity, step, step size, stochastic gradient descent (SGD), stride, structural risk minimization (SRM), subsampling, summary, supervised learning, supervised machine learning, support vector machines, synthetic feature, system, Systems for ML, tabular Q-learning, target, target network, technical assistants, technical presentations, TECHNOLOGY, temporal data, Tensor, Tensor Processing Unit (TPU), Tensor rank, Tensor shape, Tensor size, TensorBoard, TensorFlow, TensorFlow Playground, TensorFlow Serving, termination condition, test set, tf.Example, tf.keras, The conference, the need of people in AI-based systems, the rights of the individual, The Summit, their benefits, Threat to human dignity, time series analysis, timestep, to benefit, topics, tower, TPU, TPU chip, TPU device, TPU master, TPU node, TPU Pod, TPU resource, TPU slice, TPU type, TPU worker, training, training set, trajectory, transfer learning, translational invariance, transparency, trigram, true negative (TN), true positive (TP), true positive rate (TPR), tutorial, unawareness (to a sensitive attribute), underfitting, unlabeled example, unsupervised learning, unsupervised machine learning, update frequency, upweighting, user matrix, validation, validation set, vanishing gradient problem, Wasserstein loss, Weaponization of AI, weight, Weighted Alternating Least Squares (WALS), wide model, width, Yitzhak Ben-Israel
About The Author
CODESIGN.BLOG
Special Report of National Commission on Regulation and Ethics in AI
sherry 0 Data Science, Design, Engineering, Explaining of Israel, Technology,
Special Report of National Commission for Intelligent Systems Subcommittee on Regulation and Ethics in Artificial Intelligence
The field of artificial intelligence has become a significant part of the technology and general industry and the proof of this is that huge companies from all over the world are investing huge sums in developments based on artificial intelligence technologies. At the same time, the field of artificial intelligence has become a very important issue in the general strategy of various countries around the world, for the reason that no country wants to be left behind and lose strategic superiority with proven technological advantages. The State of Israel has realized that it must formulate its own strategic plan in the field of artificial intelligence, enabling it to formulate a national strategy for the coming years and ensure the continued prosperity of Israel.
After a long wait, we received approval for public publication and I am pleased to share a special report by the National Subcommittee on Artificial Intelligence and Regulation. Professors Yitzhak Ben-Israel and Evitar Matanya were responsible for producing the report detailing the State of Israel’s strategy in the field of artifical intelligence. In addition, the report offers concrete recommendations for implementation. The purpose of the report is to ensure that the State of Israel remains relevant in the technological arms race and continue to be a global leader in technological innovation. The full report has not yet been approved for publication, but there is an interim report now available to the general public.
The chairman of the committee is Prof. Karine Nahon and a dignified and earnest representation of experts from the country who have devoted quite a bit of their time to producing the report.
The report before you outlines how it is built as well as the main issues that have come up in committee discussions. The aim was to provide a practical look at each issue and provide essential tools for companies and organizations that want to make sure they do not create ethical issues during the development process.
You can also get a complete snapshot of all artificial intelligence in the world through Appendix A, which provides a brief overview of what is happening in the US and Europe and how they have defined their strategy in this area. It is recommended that you read the executive summary at the beginning of the report to understand the content of the chapters and committee recommendations.
The report consists of technological issues related to ethics and regulation in the field of artificial intelligence and is written in clear and accessible language and you will find three main chapters dealing with the whole issue:
The first chapter deals with the unique characteristics of artificial intelligence technologies, as well as the question of ethical and legal disagreements as to whether the field of artificial intelligence needs special regulatory and ethical considerations.
The second chapter deals with ethics in the use of artificial intelligence based technologies. There is a serious reference that defines familiar concepts from the content world that allow us to identify places where ethical issues can arise in product development, such as fairness, transparency, safety and more. In addition, the committee has developed a tool that can help decision makers in organizations understand in which cases they may encounter and fail ethical issues and how they can be avoided using the developed tool. In order to test the tool, a number of case studies were reviewed in which targeted ethical violations were carried out on various projects and demonstrated how the tool can be used in these types of test cases. The tool can assist both decision makers and developers during the development of different artificial intelligence systems.
The third chapter deals with the different options for the artificial intelligence regulations various aspects in the fields of ethical and legal in the field of artificial intelligence and more. In this chapter you will find a table that lists the different types of regulators as well as the pros and cons of each way of action.
The entire report is here>>
AIEthicsRegulationReportHebrewTags: A/B testing, Abstract, academia, acceleration, Accuracy, action, activation function, active learning, AdaGrad, administrative perspectives, agent, agglomerative clustering, AI, AI methods, AI system, Algorithms, analytics, analytics space, and not the other way around, application design, applied machine learning, ar, area under the PR curve, area under the ROC curve, artificial general intelligence, artificial intelligence, artificial intelligence-based systems, association, attention, attribute, AUC (Area under the ROC Curve), Augmented Reality, automation bias, Automation of Machine-Learning, Autonomous driving, available, average precision, backpropagation, bag of words, baseline, batch, batch normalization, batch size, Bayesian, Bayesian neural network, behavior, Bellman equation, bias (ethics/fairness), bias (math), biases, Big Data Analytics, bigram, binary classification, binning, boosting, broad spectrum, broadcasting, bucketing, calibration layer, candidate generation, candidate sampling, categorical data, centroid, centroid-based clustering, checkpoint, class, class-imbalanced dataset, classification, classification model, classification threshold, clipping, Cloud TPU, clustering, co-adaptation, collaborative filtering, comments, Community, Computational neural networks, computer vision, Conference, confirmation bias, confusion matrix, content, context, Continuous delivery, continuous feature, control, convenience sampling, convergence, convex function, convex optimization, convex set, convolution, convolutional filter, convolutional layer, convolutional layers, convolutional neural network, convolutional operation, core principles, cost, counterfactual fairness, coverage bias, crash blossom, critic, Cross validation, cross-entropy, custom Estimator, Cyber, data, data analysis, data augmentation, Data Quality, Data Science, Data Scientists, data set or dataset, database, DataFrame, Dataset API (tf.data), datasets, decision boundary, decision maker, decision making, decision threshold, decision tree, decision trees, decisions, deep learning, deep model, deep neural network, Deep Q-Network (DQN), delivering, demographic parity, dense feature, dense layer, depth, depthwise separable convolutional neural network (sepCNN), DESIGN, designed, Developers, device, DevOps, Dialogue Bots, different hyperparameters, different initializations, different overall structure, dimension reduction, dimensions, discrete feature, discriminative model, discriminator, disparate impact, disparate treatment, divisive clustering, domains, downsampling, DQN, dropout regularization, dynamic model, eager execution, early stopping, economic and social obstacles, Education, effective removal, embedding space, embeddings, empirical risk minimization (ERM), engineering, ensemble, environment, equal groups, equal treatment, errors, especially, establish, ethical professional duty, Ethics of artificial intelligence, events, Evitar Matanya, Evitar Matanya Prof. Karine Nahon, excellent, experience, experimenter’s bias, Explaining of Israel, fine tuning, Fintech, focus, forget gate, formal, framework, full softmax, fully autonomous, fully connected layer, Future of AI, GAN, generalization, generalization curve, generalized linear model, generative adversarial network (GAN), generative model, generator, gradient, gradient clipping, gradient descent, Gradient descent algorithm, graph, graph execution, great, great success, greedy policy, ground truth, group attribution bias, hashing, Healthcare, helpful feedback, heuristic, hidden layer, hierarchical clustering, high-quality, highlight, hinge loss, holdout data, human, hyperparameter, hyperplane, i.i.d., ideas, image recognition, imbalanced dataset, impact, Implementation of principles, implicit bias, improvement, in every context, in-group bias, including, incompatibility of fairness metrics, increasing, increasingly, independently and identically distributed (i.i.d), individual fairness, individuals, industry, industry tracks, Inference, innovation, innovative, input function, input layer, instance, Intelligent robots, inter-rater agreement, interpretability, iot, item matrix, items, iteration, k-means, k-median, Keras, Kernel Support Vector Machines (KSVMs), label, labeled example, lambda, Large scale analytics, layer, Layers API (tf.layers), leading experts, learning rate, least squares regression, liberty, linear model, linear regression, Log Loss, log-odds, logistic regression, logits, Long Short-Term Memory (LSTM), loss, loss curve, loss surface, LSTM, Machine ethics, MACHINE LEARNING, Machine learning systems, majority class, Markov decision process (MDP), Markov property, matplotlib, matrix factorization, Mean Absolute Error (MAE), Mean Squared Error (MSE), metric, Metrics API (tf.metrics), mini-batch, mini-batch stochastic gradient descent (SGD), minimax loss, minority class, ML, MNIST, model, model function, model training, Momentum, momentum (Momentum), moral and ethical use of AI technologies, more complex math (Proximal and others), multi-class classification, multi-class logistic regression, multi-class regression, multinomial classification, N-gram, NaN trap, Natural language processing, Natural Language Understanding, negative class, neural network, neural networks, neuron, Neutrality, new moral and ethical challenges, new regulatory, NLU, node (neural network), node (TensorFlow graph), noise, non-arbitrary, non-response bias, normalization, Nowadays, numerical data, NumPy, objective, objective function, offline inference, one-hot encoding, one-shot learning, one-vs.-all, online inference, Operation (op), optimizer, organizing conference, out-group homogeneity bias, outliers, output layer, Overfitting, pandas, parameter, Parameter Server (PS), parameter update, partial derivative, participants, participation bias, partitioning strategy, perceptron, performance, perplexity, pipeline, policy, pooling, population, positive class, post-processing, powerful, PR AUC (area under the PR curve), pre-trained model, precision, precision-recall curve, prediction, prediction bias, predictive applications, predictive parity, predictive rate parity, predicts, premade Estimator, preprocessing, principles, prior belief, procedural, process data, Prof. Karine Nahon, Professors, proxy (sensitive attributes), proxy labels, public accountability, public sector, public sectors, Q-function, Q-learning, quantile, quantile bucketing, quantization, queue, random forest, random policy, rank (ordinality), rank (Tensor), rater, re-ranking, real world, real-world domains, reason, recall, recommendation system, reconstruction, Rectified Linear Unit (ReLU), recurrent neural network, Registration, regression, regression model, regularization, regularization rate, reinforcement learning, reinforcement learning (RL), replay buffer, reporting bias, representation, research and application, research innovations, research track, researchers, respect, Responsibility, results, Retail, return, reward, ridge regularization, RNN, Robot rights, robotics, ROC (receiver operating characteristic) Curve, root directory, Root Mean Squared Error (RMSE), rotational invariance, sampling bias, Satisfying certain basic universal needs, SavedModel, Saver, scalar, scaling, scikit-learn, scoring, selection bias, semi-supervised learning, sensitive attribute, sequence model, services, serving, session (tf.session), shape (Tensor), sigmoid function, significant capital, similarity measure, size invariance, sketching, society, softmax, sparse feature, sparse representation, sparse vector, sparsity, sparsity/regularization (Ftrl), spatial pooling, sponsorship, squared hinge loss, squared loss, State, state-action value function, state-of-the-art, static model, stationarity, step, step size, stochastic gradient descent (SGD), stride, structural risk minimization (SRM), subsampling, summary, supervised learning, supervised machine learning, support vector machines, synthetic feature, system, Systems for ML, tabular Q-learning, target, target network, technical assistants, technical presentations, TECHNOLOGY, temporal data, Tensor, Tensor Processing Unit (TPU), Tensor rank, Tensor shape, Tensor size, TensorBoard, TensorFlow, TensorFlow Playground, TensorFlow Serving, termination condition, test set, tf.Example, tf.keras, The conference, the need of people in AI-based systems, the rights of the individual, The Summit, their benefits, Threat to human dignity, time series analysis, timestep, to benefit, topics, tower, TPU, TPU chip, TPU device, TPU master, TPU node, TPU Pod, TPU resource, TPU slice, TPU type, TPU worker, training, training set, trajectory, transfer learning, translational invariance, transparency, trigram, true negative (TN), true positive (TP), true positive rate (TPR), tutorial, unawareness (to a sensitive attribute), underfitting, unlabeled example, unsupervised learning, unsupervised machine learning, update frequency, upweighting, user matrix, validation, validation set, vanishing gradient problem, Wasserstein loss, Weaponization of AI, weight, Weighted Alternating Least Squares (WALS), wide model, width, Yitzhak Ben-Israel
Related posts
CVPR 2020 • FATE in Computer Vision
CVPR 2020 • Learning 3D Generative Models Workshop
About The Author
sherry
CODESIGN.BLOG