Digital inclusion involves the activities necessary to ensure equitable access to and use of information and communications technologies for participation in social and economic life including for education, social services, health, social, and community participation. Digital inclusion includes access to affordable broadband Internet services, Internet-enabled devices, access to digital literacy training, quality technical support, and applications and online content designed to enable and encourage self-sufficiency, participation, and collaboration. Related concepts include digital divide, digital exclusion and digital inequality, however digital inclusion focuses more on the strategies, policies, and programs required to address the digital divide. As many services have moved online and with the increasing use of telehealth to deliver primary care, particularly during the COVID-19 pandemic in 2020, digital inclusion, including digital literacy and internet access is increasingly regarded as a social determinant of health. Accessibility, relevance, and impact have been identified as essential elements of digital inclusion as it pertains to health information systems. "Digital inclusion is broadly defined as different strategies designed to ensure that all people have equal access, opportunities and skills to benefit from digital technologies and systems" (ITU, 2019, as cited in Carmi and Yates, 2020). Since 2020, there have been many technology companies that have begun implementing different features or roles within their companies to support breaking down the digital divide. For example, HP has announced the digital divide accelerator. This accelerator will support nonprofits in Greece, Indonesia, Nigeria, and Spain. The goal for this role is to help equip children and other people within the community to understand the skills needed to become a part of the digital community. == Background == With the increasing use of computers and the Internet in the 1990s and early 2000s concerns rose around digital equality, however this primarily focused on the physical access to technology. This gave rise to the concept of the digital divide which was originally developed to describe the growing disparity in Internet access between rural and urban areas of the United States of America. This gradually expanded to considerations of digital access between countries in what is termed the global digital divide, which mirrors many of the disparities seen within countries but on an international scale. However, with the adoption of digital technologies across most sectors of society, and the increasing diversity of technologies and programs, access and use of ICT became more complex and essential for many aspects of daily life. This led to new terminology and a second wave of research on digital inequality which has been identified as the (1) usage gap, (2) second level digital divide, (3) emerging digital differentiation, and (4) digital inclusion. == Strategies for digital inclusion == A review of the literature in 2019 found that while physical access to digital technologies and the internet continues to be a barrier to digital inclusion, digital ability and attitude were also potential barriers. Key strategies identified for improving digital inclusion are social support, direct user experience and collaborative learning/design. Education is a key aspect of digital inclusion as digital technologies have become a key means of engaging with all levels of the education system, requiring levels of digital competence for successful engagement with the curriculum. In addition lifelong learning is required as technologies, services and systems are changing constantly. Public libraries and community service providers play a key role in supporting digital inclusion through access to computers, internet connection and expertise and training. Designing for digital inclusion may also help with poor written literacy, which remains a barrier for 10% of the world's population. UNESCO has developed Guidelines for designing digital technologies in ways that could assist those who are illiterate. == Indigenous digital inclusion == Digital inclusion is a critical issue for many Indigenous communities across the globe, many of whom lack access to adequate resources. The Australian Government has set a National Closing the Gap target for Aboriginal and Torres Strait Islander people to have equal levels of digital inclusion by 2026. Many people on tribal land and in Native Hawaiian land struggle with the technology gap. The Native Entities Capacity and Planning Grant Program has $45.3 million available to help address these challenges and empower Indian Tribes, Alaska Native entities, and Native Hawaiian organizations. Some of the impacts so far are in the education and workforce development and healthcare access through telehealth.[13] == Measuring digital inclusion == The Australian Digital Inclusion Index (ADII) is a research project which has been tracking digital inclusion throughout Australia since 2016. It uses survey data to measure digital inclusion across three dimensions of access, affordability and digital ability. == The Future of Digital Inclusion == On February 16, 2021, a global dialogue within the United Nations (UN) took a look at the future of digital inclusion. Through the adoption of the 2030 UN Agenda for Sustainable Development, Member States made a commitment. They pledged to "leave no one behind." By 2030, the UN's goal is to close the digital divide by providing access to the Internet and mobile technologies for all nations and peoples and for all segments of society. The UN sees the crisis of too many people in our global society still living unconnected and how the digital divide remains a challenge that must be addressed. == Gaming == The Xbox Adaptive Controller is a groundbreaking example of digital inclusion, designed to make gaming more accessible to people with limited mobility. Developed by Microsoft, it features large programmable buttons and ports that connect to a wide range of external devices like switches, joysticks, and mounts, allowing users to customize their gaming experience based on their unique needs. By removing traditional physical barriers to gameplay, the Xbox Adaptive Controller empowers more people to participate in digital entertainment, promotes equal access to technology, and fosters a more inclusive gaming community. == Digital inclusion advocacy groups == Australian Digital Inclusion Alliance National Digital Inclusion Alliance (US)
Content Threat Removal
Content Threat Removal (CTR) is a cybersecurity technology intended to defeat the threat posed by handling digital content in the cyberspace. Unlike other defenses, including antivirus software and sandboxed execution, CTR does not rely on being able to detect threats. Similar to Content Disarm and Reconstruction, CTR is designed to remove the threat without knowing whether it has done so and acts without knowing if data contains a threat or not. Detection strategies work by detecting unsafe content, and then blocking or removing that content. Content that is deemed safe is delivered to its destination. In contrast, Content Threat Removal assumes all data is hostile and delivers none of it to the destination, regardless of whether it is actually hostile. Although no data is delivered, the business information carried by the data is delivered using new data created for the purpose. == Threat == Advanced attacks continuously defeat defenses that are based on detection. These are often referred to as zero-day attacks, because as soon as they are discovered attack detection mechanisms must be updated to identify and neutralize the attack, and until they are, all systems are unprotected. These attacks succeed because attackers find new ways of evading detection. Polymorphic code can be used to evade the detection of known unsafe data and sandbox detection allows attacks to evade dynamic analysis. == Method == A Content Threat Removal defence works by intercepting data on its way to its destination. The business information carried by the data is extracted and the data is discarded. Then entirely new, clean and safe data is built to carry the information to its destination. The effect of building new data to carry the business information is that any unsafe elements of the original data are left behind and discarded. This includes executable data, macros, scripts and malformed data that trigger vulnerabilities in applications. While CTR is a form of content transformation, not all transformations provide a complete defence against the content threat. == Applicability == CTR is applicable to user-to-user traffic, such as email and chat, and machine-to-machine traffic, such as web services. Data transfers can be intercepted by in-line application layer proxies and these can transform the way information content is delivered to remove any threat. CTR works by extracting business information from data and it is not possible to extract information from executable code. This means CTR is not directly applicable to web browsing, since most web pages are code. It can, however, be applied to content that is downloaded from, and uploaded to, websites. Although most web pages cannot be transformed to render them safe, web browsing can be isolated and the remote access protocols used to reach the isolated environment can be subjected to CTR. CTR provides a solution to the problem of stegware. It naturally removes detectable steganography and eliminates symbiotic and permutation steganography through normalisation.
Domain adaptation
Domain adaptation is a field associated with machine learning and transfer learning. It addresses the challenge of training a model on one data distribution (the source domain) and applying it to a related but different data distribution (the target domain). A common example is spam filtering, where a model trained on emails from one user (source domain) is adapted to handle emails for another user with significantly different patterns (target domain). Domain adaptation techniques can also leverage unrelated data sources to improve learning. When multiple source distributions are involved, the problem extends to multi-source domain adaptation. Domain adaptation is a specific type of transfer learning. According to the taxonomy laid out by Pan and Yang (2010), it falls into the category of transductive transfer learning. In this setting, the source and target tasks are the same (e.g., both are object recognition), but the domains differ (different marginal distributions). This distinguishes it from inductive transfer learning (where labeled data is available for the target task) and unsupervised transfer learning (where labels are unavailable in both domains). == Classification of domain adaptation problems == Domain adaptation setups are classified in two different ways: according to the distribution shift between the domains, and according to the available data from the target domain. === Distribution shifts === Common distribution shifts are classified as follows: Covariate Shift occurs when the input distributions of the source and destination change, but the relationship between inputs and labels remains unchanged. The above-mentioned spam filtering example typically falls in this category. Namely, the distributions (patterns) of emails may differ between the domains, but emails labeled as spam in the one domain should similarly be labeled in another. Prior Shift (Label Shift) occurs when the label distribution differs between the source and target datasets, while the conditional distribution of features given labels remains the same. An example is a classifier of hair color in images from Italy (source domain) and Norway (target domain). The proportions of hair colors (labels) differ, but images within classes like blond and black-haired populations remain consistent across domains. A classifier for the Norway population can exploit this prior knowledge of class proportions to improve its estimates. Concept Shift (Conditional Shift) refers to changes in the relationship between features and labels, even if the input distribution remains the same. For instance, in medical diagnosis, the same symptoms (inputs) may indicate entirely different diseases (labels) in different populations (domains). === Data available during training === Domain adaptation problems typically assume that some data from the target domain is available during training. Problems can be classified according to the type of this available data: Unsupervised: Unlabeled data from the target domain is available, but no labeled data. In the above-mentioned example of spam filtering, this corresponds to the case where emails from the target domain (user) are available, but they are not labeled as spam. Domain adaptation methods can benefit from such unlabeled data, by comparing its distribution (patterns) with the labeled source domain data. Semi-supervised: Most data that is available from the target domain is unlabelled, but some labeled data is also available. In the above-mentioned case of spam filter design, this corresponds to the case that the target user has labeled some emails as being spam or not. Supervised: All data that is available from the target domain is labeled. In this case, domain adaptation reduces to refinement of the source domain predictor. In the above-mentioned example classification of hair-color from images, this could correspond to the refinement of a network already trained on a large dataset of labeled images from Italy, using newly available labeled images from Norway. == Formalization == Let X {\displaystyle X} be the input space (or description space) and let Y {\displaystyle Y} be the output space (or label space). The objective of a machine learning algorithm is to learn a mathematical model (a hypothesis) h : X → Y {\displaystyle h:X\to Y} able to attach a label from Y {\displaystyle Y} to an example from X {\displaystyle X} . This model is learned from a learning sample S = { ( x i , y i ) ∈ ( X × Y ) } i = 1 m {\displaystyle S=\{(x_{i},y_{i})\in (X\times Y)\}_{i=1}^{m}} . Usually in supervised learning (without domain adaptation), we suppose that the examples ( x i , y i ) ∈ S {\displaystyle (x_{i},y_{i})\in S} are drawn i.i.d. from a distribution D S {\displaystyle D_{S}} of support X × Y {\displaystyle X\times Y} (unknown and fixed). The objective is then to learn h {\displaystyle h} (from S {\displaystyle S} ) such that it commits the least error possible for labelling new examples coming from the distribution D S {\displaystyle D_{S}} . The main difference between supervised learning and domain adaptation is that in the latter situation we study two different (but related) distributions D S {\displaystyle D_{S}} and D T {\displaystyle D_{T}} on X × Y {\displaystyle X\times Y} . The domain adaptation task then consists of the transfer of knowledge from the source domain D S {\displaystyle D_{S}} to the target one D T {\displaystyle D_{T}} . The goal is then to learn h {\displaystyle h} (from labeled or unlabelled samples coming from the two domains) such that it commits as little error as possible on the target domain D T {\displaystyle D_{T}} . The major issue is the following: if a model is learned from a source domain, what is its capacity to correctly label data coming from the target domain? == Four algorithmic principles == === Reweighting algorithms === The objective is to reweight the source labeled sample such that it "looks like" the target sample (in terms of the error measure considered). === Iterative algorithms === A method for adapting consists in iteratively "auto-labeling" the target examples. The principle is simple: a model h {\displaystyle h} is learned from the labeled examples; h {\displaystyle h} automatically labels some target examples; a new model is learned from the new labeled examples. Note that there exist other iterative approaches, but they usually need target labeled examples. === Search of a common representation space === The goal is to find or construct a common representation space for the two domains. The objective is to obtain a space in which the domains are close to each other while keeping good performances on the source labeling task. This can be achieved through the use of Adversarial machine learning techniques where feature representations from samples in different domains are encouraged to be indistinguishable. === Hierarchical Bayesian Model === The goal is to construct a Bayesian hierarchical model p ( n ) {\displaystyle p(n)} , which is essentially a factorization model for counts n {\displaystyle n} , to derive domain-dependent latent representations allowing both domain-specific and globally shared latent factors. == Software packages == Several compilations of domain adaptation and transfer learning algorithms have been implemented over the past decades: SKADA (Python) ADAPT (Python) TLlib (Python) Domain-Adaptation-Toolbox (MATLAB)
Outline of deep learning
The following outline is provided as an overview of, and topical guide to, deep learning: Deep learning is a subfield of machine learning and artificial intelligence based on artificial neural networks with multiple processing layers. It emphasizes representation learning and is widely used in areas such as computer vision, natural language processing, speech recognition, recommender systems, robotics, and generative artificial intelligence. == Ways to categorize deep learning == A field of study A branch of artificial intelligence A subfield of machine learning A subfield of computer science A form of representation learning A class of methods based on artificial neural networks An approach used in computational statistics == History == === Precursors === Cybernetics Perceptron Connectionism Neocognitron Backpropagation === Milestones === LeNet Long short-term memory Deep belief network AlexNet Sequence to sequence learning Generative adversarial network Residual neural network Transformer BERT Generative pre-trained transformer Diffusion model === Related histories === History of artificial intelligence History of machine learning Timeline of machine learning == Core concepts == == Learning settings == Supervised learning Unsupervised learning Self-supervised learning Semi-supervised learning Reinforcement learning Transfer learning Multitask learning Multimodal learning Online machine learning Continual learning == Common tasks == Image classification Object detection Image segmentation Automatic speech recognition Neural machine translation Question answering Automatic summarization Text-to-image model Protein structure prediction == Architectures == === Feedforward and convolutional architectures === Feedforward neural network Multilayer perceptron Convolutional neural network Radial basis function network Residual neural network U-Net === Recurrent and sequence architectures === Recurrent neural network Long short-term memory Gated recurrent unit Sequence to sequence learning Recursive neural network === Representation-learning architectures === Autoencoder Denoising autoencoder Sparse autoencoder Variational autoencoder Restricted Boltzmann machine Deep belief network === Attention and transformer architectures === Attention (machine learning) Transformer BERT Generative pre-trained transformer Vision transformer === Generative and probabilistic architectures === Autoregressive model Diffusion model Energy-based model Generative adversarial network Mixture of experts === Graph and memory architectures === Graph neural network Graph convolutional network Siamese network Neural Turing machine Memory network Echo state network Capsule neural network == Neural network components and techniques == Artificial neuron Activation function Rectified linear unit Sigmoid function Softmax function Embedding Convolution Pooling layer Attention Batch normalization Layer normalization Residual connections == Training and optimization == Backpropagation Gradient descent Stochastic gradient descent Adam optimization Learning rate Loss function Cross-entropy Mean squared error Regularization Dropout Early stopping Batch normalization Data augmentation Transfer learning Knowledge distillation Ensemble learning Curriculum learning == Datasets and benchmarks == CIFAR-10 ImageNet MNIST database Common Objects in Context (COCO) General Language Understanding Evaluation (GLUE) benchmark LibriSpeech SQuAD == Applications == === Computer vision === Computer vision Facial recognition system Image classification Image segmentation Medical imaging Object detection Optical character recognition === Natural language processing === Automatic summarization Chatbot Information retrieval Large language model Natural language processing Neural machine translation Question answering Sentiment analysis === Speech and audio === Automatic speech recognition Music information retrieval Speaker recognition Speech synthesis === Science and medicine === Bioinformatics Computational biology Drug discovery Medical diagnosis Protein structure prediction === Robotics and control === Autonomous car Computer game bot Control theory Robotics === Recommendation, search, and forecasting === Anomaly detection Forecasting Fraud detection Recommender system Search engine === Generative artificial intelligence === Deepfake Generative artificial intelligence Large language model Speech synthesis Text-to-image model === Computer graphics and video games === Deep Learning Anti-Aliasing (DLAA) Deep Learning Super Sampling (DLSS) == Hardware == AMD Instinct AMD XDNA Application-specific integrated circuit Deep learning processor, Neural processing unit (NPU), or Neural Engine Field-programmable gate array General-purpose computing on graphics processing units (GPGPU) Graphics processing unit NVIDIA Deep Learning Accelerator (NVDLA) Tensor processing unit Vision processing unit Wafer-scale integration === Supporting software platforms === CUDA Metal ROCm == Software == === Open-source frameworks and libraries === === Neural network software === EDLUT Emergent Encog JOONE Neuroph NeuroSolutions OpenNN Peltarion Synapse SNNS === Platforms, tools, and deployment === Amazon SageMaker Google Colab Hugging Face Kaggle Kubeflow MLflow ONNX OpenVINO TensorFlow Hub == Algorithms for deep learning and neural networks == Backpropagation Conjugate gradient method Generalized Hebbian algorithm Gradient descent Levenberg–Marquardt algorithm Perceptron Quasi-Newton method Wake-sleep algorithm == Methods and related topics == === Representation and metric learning === Contrastive learning Embedding Feature learning Manifold learning Metric learning === Generative modeling === Autoregressive model Diffusion model Generative adversarial network Generative model Variational inference === Efficient and scalable deep learning === Knowledge distillation Low-rank approximation Mixture of experts Quantization Sparsity === Reliability, safety, and interpretability === Adversarial machine learning AI alignment Algorithmic bias Catastrophic forgetting Differential privacy Explainable artificial intelligence Federated learning Hallucination (artificial intelligence) == Conferences and workshops == Annual Meeting of the Association for Computational Linguistics Conference on Computer Vision and Pattern Recognition Conference on Neural Information Processing Systems International Conference on Computer Vision International Conference on Learning Representations International Conference on Machine Learning == Organizations == === Research laboratories and institutions === Allen Institute for AI Alberta Machine Intelligence Institute European Laboratory for Learning and Intelligent Systems Google DeepMind Meta AI Mila Microsoft Research Vector Institute === Companies === Anthropic Cerebras Cohere DeepSeek Mistral AI OpenAI Stability AI xAI == Publications == === Books === Deep Learning – Ian Goodfellow and Yoshua Bengio Neural Networks and Deep Learning – Michael Nielsen Perceptrons – Marvin Minsky and Seymour Papert === Journals === IEEE Transactions on Neural Networks and Learning Systems Neural Networks Neural Computation == Influential persons ==
Intelligent automation
Intelligent automation (IA), or intelligent process automation, is a software term that refers to a combination of artificial intelligence (AI) and robotic process automation (RPA). Companies use intelligent automation to cut costs and streamline tasks by using artificial-intelligence-powered robotic software to mitigate repetitive tasks. As it accumulates data, the system learns in an effort to improve its efficiency. Intelligent automation applications consist of, but are not limited to, pattern analysis, data assembly, and classification. The term is similar to hyperautomation, a concept identified by research group Gartner as being one of the top technology trends of 2020. == Technology == Intelligent automation applies the assembly line concept of breaking tasks into repetitive steps to improve business processes. Rather than having humans perform each step, intelligent automation can replace steps with an intelligent software robot, improving efficiency. Intelligent automation integrates robotic process automation (RPA) with artificial intelligence techniques (such as machine learning, natural-language processing, and computer vision) enabling systems to interpret data, make decisions, and adapt to changing inputs. Modern platforms use a layered architecture combining workflow orchestration, low-code tools, integration middleware, and AI services to coordinate bots and data pipelines across organisational systems. == Applications == Intelligent automation is used to process unstructured content. Common real-world applications include self-driving cars, self-checkouts at grocery stores, smart home assistants, and appliances. Businesses can apply data and machine learning to build predictive analytics that react to consumer behavior changes, or to implement RPA to improve manufacturing floor operations. For example, the technology has also been used to automate the workflow behind distributing COVID-19 vaccines. Data provided by hospital systems’ electronic health records can be processed to identify and educate patients, and schedule vaccinations. Intelligent automation can provide real-time insights on profitability and efficiency. However, in an April 2022 survey by Alchemmy, despite three quarters of businesses acknowledging the importance of Artificial Intelligence to their future development, just a quarter of business leaders (25%) considered Intelligent Automation a “game changer” in understanding current performance. 42% of CTOs see “shortage of talent” as the main obstacle to implementing Intelligent Automation in their business, while 36% of CEOs see ‘upskilling and professional development of existing workforce’ as the most significant adoption barrier. IA is becoming increasingly accessible for firms of all sizes. With this in mind, it is expected to continue to grow rapidly in all industries. This technology has the potential to change the workforce. As it advances, it will be able to perform increasingly complex and difficult tasks. In addition, this may expose certain workforce issues as well as change how tasks are allocated. Tools such as Semrush's AI Visibility Toolkit and Enterprise AIO reflect these developments by analysing how entities are referenced and represented within responses produced by large-language-model-based systems. == Benefits == Streamline processes: Repetitive manual tasks can put a strain on the workforce. However, with AI agents, these tasks can be automated to allow teams to focus on more important matters that require human cognition. Intelligent automation can also be used to mitigate tasks with human error which in turn increases proficiency. This allows the opportunity for firms to scale production without the traditional negative consequences such as reduced quality or increased risk. Customer service improvement: Customer service can be significantly improved, providing the firm with a competitive advantage. IA utilizing chat features allows for instant curated responses to customers. In addition, it can give updates to customers, make appointments, manage calls, and personalize campaigns. Flexibility: Due to the wide range of applications, IA is useful across a variety of fields, technologies, projects and industries. In addition, IA can be integrated with current automated systems in place. This allows for optimized systems unique to each firm to best fit their individual needs. == Capabilities == Cognitive automation: Employs AI techniques to assist humans in decision-making and task completion Natural language processing: Allows computers to automate knowledge work Business process management: Enhances the consistency and agility of corporate operations Process mining: Applies data mining methods to discover, analyze, and improve business processes Intelligent document processing: Utilizes OCR and other advanced technologies to extract data from documents and convert it into structured, usable data Computer vision: Allows computers to extract information from digital images, videos, and other visual inputs Integration automation: Establishes a unified platform with automated workflows that integrate data, applications, and devices.
AI safety
AI safety is an interdisciplinary field focused on preventing accidents, misuse, or other harmful consequences arising from artificial intelligence systems. It encompasses AI alignment (which aims to ensure AI systems behave as intended), monitoring AI systems for risks, and enhancing their robustness. The field is particularly concerned with existential risks posed by advanced AI models. Beyond technical research, AI safety involves developing norms and policies that promote safety, including advocacy for regulations at different levels of government. The field gained significant popularity in 2023, with rapid progress in generative AI and public concerns voiced by researchers and CEOs about potential dangers. During the 2023 AI Safety Summit, the United States and the United Kingdom both established their own AI Safety Institute. However, researchers have expressed concern that AI safety measures are not keeping pace with the rapid development of AI capabilities. == Motivations == Scholars discuss current risks from critical systems failures, bias, and AI-enabled surveillance, as well as emerging risks like technological unemployment, digital manipulation, weaponization, AI-enabled cyberattacks and bioterrorism. They also discuss speculative risks from losing control of future artificial general intelligence (AGI) agents, or from AI enabling perpetually stable dictatorships. === Existential safety === Some have criticized concerns about AGI, such as Andrew Ng who compared them in 2015 to "worrying about overpopulation on Mars when we have not even set foot on the planet yet". Stuart J. Russell on the other side urges caution, arguing that "it is better to anticipate human ingenuity than to underestimate it". AI researchers have widely differing opinions about the severity and primary sources of risk posed by AI technology – though surveys suggest that experts take high consequence risks seriously. In two surveys of AI researchers, the median respondent was optimistic about AI overall, but placed a 5% probability on an "extremely bad (e.g. human extinction)" outcome of advanced AI. In a 2022 survey of the natural language processing community, 37% agreed or weakly agreed that it is plausible that AI decisions could lead to a catastrophe that is "at least as bad as an all-out nuclear war". == History == Risks from AI began to be seriously discussed at the start of the computer age: Moreover, if we move in the direction of making machines which learn and whose behavior is modified by experience, we must face the fact that every degree of independence we give the machine is a degree of possible defiance of our wishes. In 1988 Blay Whitby published a book outlining the need for AI to be developed along ethical and socially responsible lines. From 2008 to 2009, the Association for the Advancement of Artificial Intelligence (AAAI) commissioned a study to explore and address potential long-term societal influences of AI research and development. The panel was generally skeptical of the radical views expressed by science-fiction authors but agreed that "additional research would be valuable on methods for understanding and verifying the range of behaviors of complex computational systems to minimize unexpected outcomes". In 2011, Roman Yampolskiy introduced the term "AI safety engineering" at the Philosophy and Theory of Artificial Intelligence conference, listing prior failures of AI systems and arguing that "the frequency and seriousness of such events will steadily increase as AIs become more capable". In 2014, philosopher Nick Bostrom published the book Superintelligence: Paths, Dangers, Strategies. He has the opinion that the rise of AGI has the potential to create various societal issues, ranging from the displacement of the workforce by AI, manipulation of political and military structures, to even the possibility of human extinction. His argument that future advanced systems may pose a threat to human existence prompted Elon Musk, Bill Gates, and Stephen Hawking to voice similar concerns. In 2015, dozens of artificial intelligence experts signed an open letter on artificial intelligence calling for research on the societal impacts of AI and outlining concrete directions. To date, the letter has been signed by over 8000 people including Yann LeCun, Shane Legg, Yoshua Bengio, and Stuart Russell. In the same year, a group of academics led by professor Stuart J. Russell founded the Center for Human-Compatible AI at the University of California Berkeley and the Future of Life Institute awarded $6.5 million in grants for research aimed at "ensuring artificial intelligence (AI) remains safe, ethical and beneficial". In 2016, the White House Office of Science and Technology Policy and Carnegie Mellon University announced The Public Workshop on Safety and Control for Artificial Intelligence, which was one of a sequence of four White House workshops aimed at investigating "the advantages and drawbacks" of AI. In the same year, Concrete Problems in AI Safety – one of the first and most influential technical AI Safety agendas – was published. In 2017, the Future of Life Institute sponsored the Asilomar Conference on Beneficial AI, where more than 100 thought leaders formulated principles for beneficial AI including "Race Avoidance: Teams developing AI systems should actively cooperate to avoid corner-cutting on safety standards". In 2018, the DeepMind Safety team outlined AI safety problems in specification, robustness, and assurance. The following year, researchers organized a workshop at ICLR that focused on these problem areas. In 2021, Unsolved Problems in ML Safety was published, outlining research directions in robustness, monitoring, alignment, and systemic safety. In 2023, Rishi Sunak said he wants the United Kingdom to be the "geographical home of global AI safety regulation" and to host the first global summit on AI safety. The AI safety summit took place in November 2023, and focused on the risks of misuse and loss of control associated with frontier AI models. During the summit the intention to create the International Scientific Report on the Safety of Advanced AI was announced. In 2024, The US and UK forged a new partnership on the science of AI safety. The MoU was signed on 1 April 2024 by US commerce secretary Gina Raimondo and UK technology secretary Michelle Donelan to jointly develop advanced AI model testing, following commitments announced at an AI Safety Summit in Bletchley Park in November. In 2025, an international team of 96 experts chaired by Yoshua Bengio published the first International AI Safety Report. The report, commissioned by 30 nations and the United Nations, represents the first global scientific review of potential risks associated with advanced artificial intelligence. It details potential threats stemming from misuse, malfunction, and societal disruption, with the objective of informing policy through evidence-based findings, without providing specific recommendations. == Research focus == AI safety research areas include robustness, monitoring, and alignment. === Robustness === ==== Adversarial robustness ==== AI systems are often vulnerable to adversarial examples or "inputs to machine learning (ML) models that an attacker has intentionally designed to cause the model to make a mistake". For example, in 2013, Szegedy et al. discovered that adding specific imperceptible perturbations to an image could cause it to be misclassified with high confidence. This continues to be an issue with neural networks, though in recent work the perturbations are generally large enough to be perceptible. The image on the right is predicted to be an ostrich after the perturbation is applied. (Left) is a correctly predicted sample, (center) perturbation applied magnified by 10x, (right) adversarial example. Adversarial robustness is often associated with security. Researchers demonstrated that an audio signal could be imperceptibly modified so that speech-to-text systems transcribe it to any message the attacker chooses. Network intrusion and malware detection systems also must be adversarially robust since attackers may design their attacks to fool detectors. Models that represent objectives (reward models) must also be adversarially robust. For example, a reward model might estimate how helpful a text response is and a language model might be trained to maximize this score. Researchers have shown that if a language model is trained for long enough, it will leverage the vulnerabilities of the reward model to achieve a better score and perform worse on the intended task. This issue can be addressed by improving the adversarial robustness of the reward model. More generally, any AI system used to evaluate another AI system must be adversarially robust. This could include monitoring tools, since they could also potentially be tampered with to produce a higher reward. Large language models (LLMs) can be vulnerable to prom
MLOps
MLOps or ML Ops is a paradigm that aims to deploy and maintain machine learning models in production reliably and efficiently. It bridges the gap between machine learning development and production operations, ensuring that models are robust, scalable, and aligned with business goals. The word is a compound of "machine learning" and the continuous delivery practice (CI/CD) of DevOps in the software field. Machine learning models are tested and developed in isolated experimental systems. When an algorithm is ready to be launched, MLOps is practiced between data scientists, DevOps, and machine learning engineers to transition the algorithm to production systems. Similar to DevOps or DataOps approaches, MLOps seeks to increase automation and improve the quality of production models, while also focusing on business and regulatory requirements. While MLOps started as a set of best practices, it is slowly evolving into an independent approach to ML lifecycle management. MLOps applies to the entire lifecycle - from integrating with model generation (software development lifecycle, continuous integration/continuous delivery), orchestration, and deployment, to health, diagnostics, governance, and business metrics. == Definition == MLOps is a paradigm, including aspects like best practices, sets of concepts, as well as a development culture when it comes to the end-to-end conceptualization, implementation, monitoring, deployment, and scalability of machine learning products. Most of all, it is an engineering practice that leverages three contributing disciplines: machine learning, software engineering (especially DevOps), and data engineering. MLOps is aimed at productionizing machine learning systems by bridging the gap between development (Dev) and operations (Ops). Essentially, MLOps aims to facilitate the creation of machine learning products by leveraging these principles: CI/CD automation, workflow orchestration, reproducibility; versioning of data, model, and code; collaboration; continuous ML training and evaluation; ML metadata tracking and logging; continuous monitoring; and feedback loops. == History == Interest in operationalizing machine learning systems began to grow in the mid-2010s as ML projects started moving from experimentation to production use. The challenges associated with sustaining such systems were highlighted in a 2015 paper. The predicted growth in machine learning included an estimated doubling of ML pilots and implementations from 2017 to 2018, and again from 2018 to 2020. Reports show a majority (up to 88%) of corporate machine learning initiatives are struggling to move beyond test stages. However, those organizations that actually put machine learning into production saw a 3–15% profit margin increases. The MLOps market size was USD 2,191.8 Million in 2024, and is projected to be USD 16,613.4 Million in 2030. == Architecture == Machine Learning systems can be categorized in eight different categories: data collection, data processing, feature engineering, data labeling, model design, model training and optimization, endpoint deployment, and endpoint monitoring. Each step in the machine learning lifecycle is built in its own system, but requires interconnection. These are the minimum systems that enterprises need to scale machine learning within their organization. == Goals == There are a number of goals enterprises want to achieve through MLOps systems successfully implementing ML across the enterprise, including: Deployment and automation Reproducibility of models and predictions Diagnostics Governance and regulatory compliance Scalability Collaboration Business uses Monitoring and management A standard practice, such as MLOps, takes into account each of the aforementioned areas, which can help enterprises optimize workflows and avoid issues during implementation. Vendors such as Adaptive ML deliver commercial reinforcement learning operations (RLOps) and MLOps-infrastructure, targeting organizations deploying large language models in production. A common architecture of an MLOps system would include data science platforms where models are constructed and the analytical engines where computations are performed, with the MLOps tool orchestrating the movement of machine learning models, data and outcomes between the systems.