AI Art Backlash

AI Art Backlash — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

• 

  Machine unlearning

  Machine unlearning is a branch of machine learning focused on removing specific undesired element, such as private data, wrong or manipulated training data, outdated information, copyrighted material, harmful content, dangerous abilities, or misinformation, without needing to rebuild models from the ground up. Large language models, like the ones powering ChatGPT, may be asked not just to remove specific elements but also to unlearn a "concept," "fact," or "knowledge," which aren't easily linked to specific examples. New terms such as "model editing," "concept editing," and "knowledge unlearning" have emerged to describe this process. == History == Early research efforts were largely motivated by Article 17 of the GDPR, the European Union's privacy regulation commonly known as the "right to be forgotten" (RTBF), introduced in 2014. The GDPR did not anticipate that the development of large language models would make data erasure a complex task. This issue has since led to research on "machine unlearning," with a growing focus on removing copyrighted material, harmful content, dangerous capabilities, and misinformation. Just as early experiences in humans shape later ones, some concepts are more fundamental and harder to unlearn. A piece of knowledge may be so deeply embedded in the model's knowledge graph that unlearning it could cause internal contradictions, requiring adjustments to other parts of the graph to resolve them. Researchers have now also started studying unlearning in the context of removing incorrect or adversarially manipulated training data such as systematically biased labels or poisoning attacks. == Motivations == At present, machine unlearning is motivated by a growing range of concerns that extend well beyond the field's original focus on data privacy. A widely used taxonomy in the literature distinguishes two high-level categories of motivation. Access revocation covers cases where a data subject or rights holder requests the removal of data they own or control. This is most commonly associated with RTBF established by the European Union's General Data Protection Regulation (GDPR) and analogous legislation such as the California Consumer Privacy Act (CCPA). These regulations grant individuals the legal right to request erasure of their personal data from any system that has processed it, including models that were trained on it. Access revocation also encompasses the removal of copyrighted or pay-walled content that was incorporated into training corpora without the necessary licenses, a concern that has become prominent with the widespread use of largely web-scraped pre-training datasets. Model correction covers cases where the model exhibits undesirable behavior arising from the training data, regardless of any individual's request. This includes: Removal of toxic, biased, or unsafe outputs introduced by harmful content in the training set Correction of stale or factually incorrect associations, such as outdated knowledge encoded in a deployed model Removal of dangerous capabilities, such as detailed knowledge of the synthesis of chemical or biological agents Correction of the influence of data poisoning or adversarial attacks that have corrupted model behavior This second category has been formalized as corrective machine unlearning, which frames unlearning as a post-training mechanism for repairing the effects of bad or harmful training data. It is closely related to the AI safety literature, where data filtering alone has been found insufficient to prevent hazardous knowledge from being encoded in model weights, motivating unlearning as a complementary risk mitigation strategy. A further distinction has been drawn in the literature between removal {eliminating the influence of specific training data on model parameters) and suppression (preventing the model from generating specific outputs regardless of how that knowledge is encoded). These two goals are not equivalent: removing training data does not guarantee meaningful output suppression, and suppressing outputs does not constitute removal of the underlying training data's influence. == SISA Training == SISA is a training strategy consisting of four mechanisms designed to make machine unlearning more efficient by structuring how models are trained and updated. Its goal is to allow a system to remove the influence of specific data points without retraining an entire model from scratch. By reorganizing training data and workflows, SISA reduces the computational burden of unlearning requests. Sharding divides the training dataset into multiple disjoint subsets, or shards. Each shard is used to train a separate model instance. This ensures that a single data point affects only one shard, so unlearning it requires updating only the corresponding shard rather than the full model. Isolation refers to training each shard independently, with nothing shared across shards during the training process. This separation prevents cross-contamination between shards, ensuring that forgetting data in one shard does not require adjustments to any others. Slicing breaks the data within each shard into sequential slices and stores model states after each slice is trained on. When an unlearning request targets a piece of data, the system can roll back to the checkpoint before the point was seen and retrain only from that slice forward. This reduces retraining time even within a shard. Aggregation occurs at inference, when the model is queried. It combines the outputs of each shard to determine the output of the overall model. This is often through majority voting or averaging. This allows SISA-trained systems to behave like a single model despite being composed of multiple shard-level models. Together, these mechanisms enable machine learning systems to forget specific data points with far lower computational cost than full retraining. The trade-off is that sharding and slicing can lead to reduced model accuracy, worse generalization, and increased storage requirements for the intermediate checkpoints. This can be tolerable based on the needs of the individual or organization to comply with "right to be forgotten" or efficiently recover from backdoor attacks. == Algorithms == Machine unlearning algorithms are broadly categorized into exact and approximate methods, reflecting a fundamental trade-off between formal guarantees and computational tractability. === Exact Unlearning === Exact unlearning methods produce a model that is statistically indistinguishable from one retrained from scratch on the dataset with the forget data removed. The canonical framework for exact unlearning is SISA Training (Sharded, Isolated, Sliced, and Aggregated), introduced by Bourtoule et al. (2021). SISA partitions the training dataset into disjoint shards and trains a separate sub-model on each. At inference time, predictions are aggregated across sub-models. When an unlearning request is received, only the sub-model corresponding to the shard containing the target data requires retraining, reducing computational overhead proportionally to the number of shards. Exact methods provide the strongest guarantees but become prohibitively expensive for large pre-trained neural networks and are generally limited to settings where training can be structured in advance. === Approximate Unlearning === Approximate unlearning methods seek to produce a model whose behavior is sufficiently close to an exactly unlearned model without the cost of full retraining. These methods dominate practical applications. Common approaches include: Gradient Ascent: The model is fine-tuned by maximizing the loss on the forget set, directly degrading its performance on targeted data. This is the most direct approach but risks destabilizing performance on retained data. Random Labelling: The model is fine-tuned on the forget set using randomly shuffled labels, confusing its associations with the targeted data while producing a less aggressive weight shift than pure gradient ascent. Gradient Difference: Combines gradient ascent on the forget set with simultaneous gradient descent on the retain set, using the retain objective as a regularizer to preserve general model utility. KL Divergence Regularization: Minimizes the KL divergence between the outputs of the unlearned model and the original model on the retain set, anchoring behavior on data the model should remember. Weight Pruning and Fine-tuning: Parameters with the smallest L1-norm are pruned — targeting weights most weakly associated with general knowledge and potentially most associated with the forget set — followed by fine-tuning on the retain set to restore utility. Layer Reset and Fine-tuning: The first or last k layers are re-initialized to random weights and the model is subsequently fine-tuned on the retain set. This is a coarse but computationally simple approach. Selective Synaptic Dampening: Uses influence functions to estimate the effect of individual trainin

  Read more →
• 

  Lai–Robbins lower bound

  The Lai–Robbins lower bound gives an asymptotic lower bound on the regret that any uniformly good algorithm must incur in the stochastic multi-armed bandit problem. The original result was proved by Tze Leung Lai and Herbert Robbins in 1985 for parametric exponential families. Later work extended the statement to more general classes of distributions. == Multi-armed bandit problem == The multi-armed bandit problem (MAB) is a sequential game in which the player must trade off exploration (to learn) and exploitation (to earn). The player chooses among K {\displaystyle K} actions (arms) with unknown distributions ν = ( ν 1 , … , ν K ) {\displaystyle \nu =(\nu _{1},\dots ,\nu _{K})} . The player is assumed to know a class of distributions D {\displaystyle {\mathcal {D}}} such that for every k {\displaystyle k} one has ν k ∈ D {\displaystyle \nu _{k}\in {\mathcal {D}}} (for example, D {\displaystyle {\mathcal {D}}} may be the family of Gaussian or Bernoulli distributions). At each round t = 1 , … , T {\displaystyle t=1,\dots ,T} the player selects (pulls) an arm a t {\displaystyle a_{t}} and observes a reward X t ∼ ν a t {\displaystyle X_{t}\sim \nu _{a_{t}}} . We denote N a ( t ) := ∑ s = 1 t 1 { a s = a } {\displaystyle N_{a}(t):=\sum _{s=1}^{t}\mathbf {1} _{\{a_{s}=a\}}} the number of times arm a {\displaystyle a} has been pulled in the first t {\displaystyle t} rounds, μ ( ν ) := ( μ 1 , … , μ K ) {\displaystyle \mu (\nu ):=(\mu _{1},\dots ,\mu _{K})} the vector of arm means, where μ k = E X ∼ ν k [ X ] {\displaystyle \mu _{k}=\mathbb {E} _{X\sim \nu _{k}}[X]} , μ ∗ := max a μ a {\displaystyle \mu ^{}:=\max _{a}\mu _{a}} the highest mean Δ a := μ ∗ − μ a ≥ 0 {\displaystyle \Delta _{a}:=\mu ^{}-\mu _{a}\geq 0} the gap of arm a {\displaystyle a} . An arm a {\displaystyle a} with μ a = μ ∗ {\displaystyle \mu _{a}=\mu ^{}} is called an optimal arm; otherwise it is a suboptimal arm. The goal is to minimize the regret at horizon T {\displaystyle T} , defined by R T := ∑ a = 1 K Δ a E [ N a ( T ) ] . {\displaystyle R_{T}:=\sum _{a=1}^{K}\Delta _{a}\,\mathbb {E} [N_{a}(T)].} Intuitively, the regret is the (expected) total loss compared to always playing an optimal arm: regret = ∑ a ( cost of playing a ) × ( times a is played ) . {\displaystyle {\text{regret}}=\sum _{a}\ ({\text{cost of playing }}a)\times ({\text{times }}a{\text{ is played}}).} An MAB algorithm is a (possibly randomized) policy that, at each round t {\displaystyle t} , choose an arm a_t by using the observations received from previous turns. === Intuitive example === Suppose a farmer must choose, each year, one of K {\displaystyle K} seed varieties to plant. Each variety k {\displaystyle k} has an unknown average yield μ k {\displaystyle \mu _{k}} . If the farmer knew the best variety (with mean μ ∗ {\displaystyle \mu ^{}} ) he would plant it every year; in reality he must try varieties to learn which is best. The cumulative regret after T {\displaystyle T} years measures the total expected loss in yield due to imperfect knowledge. Remarks The model above is the stochastic MAB; there also exist adversarial variants. One may consider a fixed-horizon setting (known T {\displaystyle T} ) or an anytime setting (unknown T {\displaystyle T} ). == Lai–Robbins lower bound == The theorem gives the right amount of time we should pull a suboptimal arm k {\displaystyle k} to distinguish whether we are in the instance with ν k {\displaystyle \nu _{k}} or with ν ~ k {\displaystyle {\tilde {\nu }}_{k}} where ν ~ k {\displaystyle {\tilde {\nu }}_{k}} is such that μ ~ k > μ ∗ {\displaystyle {\tilde {\mu }}_{k}>\mu ^{}} . Knowning a lower bound on the number of pull of every suboptimal arm gives a lower bound on the regret as only suboptimal arms contribute to the regret. Before stating the formal theorem we need to define what is a consistent algorithm. === Consistency (uniformly good algorithms) === Let D {\displaystyle {\mathcal {D}}} be a class of probability distributions and consider K {\displaystyle K} arms with reward distributions ν = ( ν 1 , … , ν K ) ∈ D K {\displaystyle \nu =(\nu _{1},\dots ,\nu _{K})\in {\mathcal {D}}^{K}} . An algorithm is said to be consistent (also called uniformly good) on D K {\displaystyle {\mathcal {D}}^{K}} if, for every instance ν ∈ D K {\displaystyle \nu \in {\mathcal {D}}^{K}} , the expected regret R T ( ν ) {\displaystyle R_{T}(\nu )} grows subpolynomially: ∀ α > 0 , R T ( ν ) = o ( T α ) as T → ∞ {\displaystyle \forall \alpha >0,\qquad R_{T}(\nu )=o(T^{\alpha })\quad {\text{as }}T\to \infty } This assumption excludes algorithms that perform well on some instances but incur linear regret on others. === Formal lower bound === For any suboptimal arm a {\displaystyle a} . For a distribution ν a ∈ D {\displaystyle \nu _{a}\in {\mathcal {D}}} and a threshold x {\displaystyle x} , define K inf ( ν a , x , D ) := inf { KL ⁡ ( ν a , ν ′ ) : ν ′ ∈ D , μ ′ > x } {\displaystyle {\mathcal {K}}_{\inf }(\nu _{a},x,{\mathcal {D}}):=\inf {\Bigl \{}\operatorname {KL} (\nu _{a},\nu '):\nu '\in {\mathcal {D}},\ \mu '>x{\Bigr \}}} where KL ⁡ ( ⋅ , ⋅ ) {\displaystyle \operatorname {KL} (\cdot ,\cdot )} denotes the Kullback-Leibler divergence. Then, for any algorithm consistent on D K {\displaystyle {\mathcal {D}}^{K}} and for every instance ν ∈ D K {\displaystyle \nu \in {\mathcal {D}}^{K}} , every suboptimal arm a {\displaystyle a} satisfies E ν [ N a ( T ) ] ≥ ln ⁡ T K inf ( ν a , μ ∗ , D ) + o ( ln ⁡ T ) {\displaystyle \mathbb {E} _{\nu }[N_{a}(T)]\geq {\frac {\ln T}{{\mathcal {K}}_{\inf }(\nu _{a},\mu ^{},{\mathcal {D}})}}+o(\ln T)} Consequently, the regret satisfies R T ( ν ) ≥ ( ∑ a : μ a < μ ∗ Δ a K inf ( ν a , μ ∗ , D ) ) ln ⁡ T + o ( ln ⁡ T ) {\displaystyle R_{T}(\nu )\geq \left(\sum _{a:\,\mu _{a}<\mu ^{}}{\frac {\Delta _{a}}{{\mathcal {K}}_{\inf }(\nu _{a},\mu ^{},{\mathcal {D}})}}\right)\ln T+o(\ln T)} The original 1985 paper established this result for exponential families; later work showed that the bound holds under much weaker assumptions on D {\displaystyle {\mathcal {D}}} . === Intuition === Consistency imposes that, for every ν {\displaystyle \nu } , the number of pulls of an optimal arm must be large. This means that μ ∗ {\displaystyle \mu ^{}} is estimated very accurately. The goal is to determine, for a suboptimal arm k {\displaystyle k} , how many samples are needed to be confident, with the appropriate level of confidence, that μ k < μ ∗ {\displaystyle \mu _{k}<\mu ^{}} . To do so, we use what is called the most confusing instance: an instance close to ν {\displaystyle \nu } such that arm k {\displaystyle k} is optimal. We define it as ν ~ {\displaystyle {\tilde {\nu }}} such that, for all a ≠ k {\displaystyle a\neq k} , ν ~ a = ν a {\displaystyle {\tilde {\nu }}_{a}=\nu _{a}} , and ν ~ k {\displaystyle {\tilde {\nu }}_{k}} is chosen so that μ ~ k > μ ∗ {\displaystyle {\tilde {\mu }}_{k}>\mu ^{}} . The objective is to determine how many samples of arm k {\displaystyle k} are required to distinguish whether we are in the instance with ν k {\displaystyle \nu _{k}} or with ν ~ k {\displaystyle {\tilde {\nu }}_{k}} in terms of KL {\displaystyle \operatorname {KL} } distance. == Algorithms achieving the Lai–Robbins lower bound == Several algorithms are known to achieve the Lai–Robbins asymptotic lower bound under specific assumptions on the reward distribution class D {\displaystyle {\mathcal {D}}} . The following list summarizes a non-exhaustive list of algorithms matching the lower bound. == Extension to other problems == === Structured bandit === A more complexe is structured bandit where we know that the mean of each arm is in a set with some restriction. In this case we can prove a smaller lower bound that use the knowledge of this set. === Best arm identification (BAI) === A similar result has been proved for best arm identification, which is the same game except that, instead of minimizing the regret, the goal is to identify the best arm with probability 1 − δ {\displaystyle 1-\delta } using as few rounds as possible. === Reinforcement Learning (RL) === Similar results have been proved for regret minimization in average-reward reinforcement learning. The order is also ln ⁡ T {\displaystyle \ln T} , with a constant that depends on the problem.

  Read more →
• 

  Schema crosswalk

  A schema crosswalk is a table that shows equivalent elements (or "fields") in more than one database schema. It maps the elements in one schema to the equivalent elements in another. Crosswalk tables are often employed within or in parallel to enterprise systems, especially when multiple systems are interfaced or when the system includes legacy system data. In the context of Interfaces, they function as an internal extract, transform, load (ETL) mechanism. For example, this is a metadata crosswalk from MARC standards to Dublin Core: Crosswalks show people where to put the data from one scheme into a different scheme. They are often used by libraries, archives, museums, and other cultural institutions to translate data to or from MARC standards, Dublin Core, Text Encoding Initiative (TEI), and other metadata schemes. For example, an archive has a MARC record in its catalog describing a manuscript. Suppose the archive makes a digital copy of that manuscript and wants to display it on the web along with the information from the catalog. In that case, it will have to translate the data from the MARC catalog record into a different format, such as Metadata Object Description Schema, that is viewable on a webpage. Because MARC has various fields than MODS, decisions must be made about where to put the data into MODS. This type of "translating" from one format to another is often called "metadata mapping" or "field mapping," and is related to "data mapping", and "semantic mapping". Crosswalks also have several technical capabilities. They help databases using different metadata schemes to share information. They help metadata harvesters create union catalogs. They enable search engines to search multiple databases simultaneously with a single query. == Challenges for crosswalks == One of the biggest challenges for crosswalks is that no two metadata schemes are 100% equivalent. One scheme may have a field that doesn't exist in another scheme or a field that is split into two different fields in another scheme; this is why data is often lost when mapping from a complex scheme to a simpler one. For example, when mapping from MARC to Simple Dublin Core, the distinction between types of titles is lost: Simple Dublin Core only has one "Title" element, so all of the different types of MARC titles get lumped together without further distinctions. A future attempt to convert the metadata back into MARC would enter the information in the basic MARC 245 Title Statement field, with none of the original distinctions. This is why crosswalks are said to be "lateral" (one-way) mappings from one scheme to another. Separate crosswalks would be required to map from scheme A to scheme B and from scheme B to scheme A. === Difficulties in mapping === Other mapping problems arise when: One scheme has one element that needs to be split up with different parts of it placed in multiple other elements in the second scheme ("one-to-many" mapping) One scheme allows an element to be repeated more than once while another only allows that element to appear once with multiple terms in it Schemes have different data formats (e.g. John Doe or Doe, John) An element in one scheme is indexed, but the equivalent element in the other scheme is not Schemes may use different controlled vocabularies Schemes change their standards over time Some of these problems are not fixable. As Karen Coyle says in "Crosswalking Citation Metadata: The University of California's Experience," "The more metadata experience we have, the more it becomes clear that metadata perfection is not attainable, and anyone who attempts it will be sorely disappointed. When metadata is crosswalked between two or more unrelated sources, there will be data elements that cannot be reconciled in an ideal manner. The key to a successful metadata crosswalk is intelligent flexibility. It is essential to focus on the important goals and be willing to compromise to reach a practical conclusion to projects."

  Read more →
• 

  Enterprise data planning

  Enterprise data planning is the starting point for enterprise wide change. It states the destination and describes how you will get there. It defines benefits, costs and potential risks. It provides measures to be used along the way to judge progress and adjust the journey according to changing circumstances. Data is fundamental to investment enterprises. Effective, economic management of data underpins operations and enables transformations needed to satisfy customer demands, competition and regulation. Data warehouse(s) and other aspects of the overall data architecture are critical to the enterprise. EDMworks has created a strategic data planning approach for the Investment Sector. It consists of a planning process, planning intranets, templates and training materials. EDMworks planning process is based on the belief that extensive domain knowledge significantly shortens planning iterations and enables progressively higher quality plans to be produced and implemented. This approach drives the development of an effective and economic enterprise data architecture. Enterprise data planning is based on proven business disciplines. Key architectural layers for data and applications are then added in order to provide an enterprise wide understanding of the uses and interdependencies of data. This enables the definition of the core components of the EDM plan: Industry structure and business objectives Assessment of systems and services Target architecture for applications, data and infrastructure Target organization structures Systems, database, infrastructure and organizational plans Business case, costs, benefits, results and risks. EDMworks uses several components from the Open Systems Group TOGAF enterprise systems planning process. TOGAF acts as an extension to good business planning methods to provide a framework for the development of the systems and data architectural components. == History == James Martin was one of the pathfinders in data planning methodologies. He was one of the first to identify data as being an enterprise wide asset that required management. He developed a series of tools and methods to support that process. Most of the large consulting firms developed their own methods to address the same basic issue. Frequently, their approaches were incorporated into their own branded system development methodologies that encompassed the complete systems development life-cycle. Others, such as Ed Tozer, developed more focused offerings that dealt with the complexities of extracting key business needs from senior management and then defining relevant architectural visions for the specific enterprise. From these various sources, the concepts of Business, Data, Applications and Technology Architectures emerged. The Open Group Architectural Framework (TOGAF) has taken this work forward and has established a sound method in TOGAF version 9. EDMworks approach is to adopt these planning and architectural practices as a basis and then add two additional dimensions to the planning and implementation focus: Domain knowledge of the Investments sector. Investments is a complex global industry with a common set of characteristics about clients, information vendors, competition and regulation. Domain knowledge significantly improves the quality of the planning and implementation processes Development of people and teams. Change is a major feature of in any Enterprise Data Management program and people and teams both need development in order to make EDM effective throughout an organization.

  Read more →
• 

  Multi-agent reinforcement learning

  Multi-agent reinforcement learning (MARL) is a sub-field of reinforcement learning. It focuses on studying the behavior of multiple learning agents that coexist in a shared environment. Each agent is motivated by its own rewards, and does actions to advance its own interests; in some environments these interests are opposed to the interests of other agents, resulting in complex group dynamics. Multi-agent reinforcement learning is closely related to game theory and especially repeated games, as well as multi-agent systems. Its study combines the pursuit of finding ideal algorithms that maximize rewards with a more sociological set of concepts. While research in single-agent reinforcement learning is concerned with finding the algorithm that gets the biggest number of points for one agent, research in multi-agent reinforcement learning evaluates and quantifies social metrics, such as cooperation, reciprocity, equity, social influence, language and discrimination. == Definition == Similarly to single-agent reinforcement learning, multi-agent reinforcement learning is modeled as some form of a Markov decision process (MDP). Fix a set of agents I = { 1 , . . . , N } {\displaystyle I=\{1,...,N\}} . We then define: A set S {\displaystyle S} of environment states. One set A i {\displaystyle {\mathcal {A}}_{i}} of actions for each of the agents i ∈ I = { 1 , … , N } {\displaystyle i\in I=\{1,\dots ,N\}} . P a → ( s , s ′ ) = Pr ( s t + 1 = s ′ ∣ s t = s , a → t = a → ) {\displaystyle P_{\vec {a}}(s,s')=\Pr(s_{t+1}=s'\mid s_{t}=s,{\vec {a}}_{t}={\vec {a}})} is the probability of transition (at time t {\displaystyle t} ) from state s {\displaystyle s} to state s ′ {\displaystyle s'} under joint action a → {\displaystyle {\vec {a}}} . R → a → ( s , s ′ ) {\displaystyle {\vec {R}}_{\vec {a}}(s,s')} is the immediate joint reward after the transition from s {\displaystyle s} to s ′ {\displaystyle s'} with joint action a → {\displaystyle {\vec {a}}} . In settings with perfect information, such as the games of chess and Go, the MDP would be fully observable. In settings with imperfect information, especially in real-world applications like self-driving cars, each agent would access an observation that only has part of the information about the current state. In the partially observable setting, the core model is the partially observable stochastic game in the general case, and the decentralized POMDP in the cooperative case. == Cooperation vs. competition == When multiple agents are acting in a shared environment their interests might be aligned or misaligned. MARL allows exploring all the different alignments and how they affect the agents' behavior: In pure competition settings, the agents' rewards are exactly opposite to each other, and therefore they are playing against each other. Pure cooperation settings are the other extreme, in which agents get the exact same rewards, and therefore they are playing with each other. Mixed-sum settings cover all the games that combine elements of both cooperation and competition. === Pure competition settings === When two agents are playing a zero-sum game, they are in pure competition with each other. Many traditional games such as chess and Go fall under this category, as do two-player variants of video games like StarCraft. Because each agent can only win at the expense of the other agent, many complexities are stripped away. There is no prospect of communication or social dilemmas, as neither agent is incentivized to take actions that benefit its opponent. The Deep Blue and AlphaGo projects demonstrate how to optimize the performance of agents in pure competition settings. One complexity that is not stripped away in pure competition settings is autocurricula. As the agents' policy is improved using self-play, multiple layers of learning may occur. === Pure cooperation settings === MARL is used to explore how separate agents with identical interests can communicate and work together. Pure cooperation settings are explored in recreational cooperative games such as Overcooked, as well as real-world scenarios in robotics. In pure cooperation settings all the agents get identical rewards, which means that social dilemmas do not occur. In pure cooperation settings, oftentimes there are an arbitrary number of coordination strategies, and agents converge to specific "conventions" when coordinating with each other. The notion of conventions has been studied in language and also alluded to in more general multi-agent collaborative tasks. === Mixed-sum settings === Most real-world scenarios involving multiple agents have elements of both cooperation and competition. For example, when multiple self-driving cars are planning their respective paths, each of them has interests that are diverging but not exclusive: Each car is minimizing the amount of time it's taking to reach its destination, but all cars have the shared interest of avoiding a traffic collision. Zero-sum settings with three or more agents often exhibit similar properties to mixed-sum settings, since each pair of agents might have a non-zero utility sum between them. Mixed-sum settings can be explored using classic matrix games such as prisoner's dilemma, more complex sequential social dilemmas, and recreational games such as Among Us, Diplomacy and StarCraft II. Mixed-sum settings can give rise to communication and social dilemmas. == Social dilemmas == As in game theory, much of the research in MARL revolves around social dilemmas, such as prisoner's dilemma, chicken and stag hunt. While game theory research might focus on Nash equilibria and what an ideal policy for an agent would be, MARL research focuses on how the agents would learn these ideal policies using a trial-and-error process. The reinforcement learning algorithms that are used to train the agents are maximizing the agent's own reward; the conflict between the needs of the agents and the needs of the group is a subject of active research. Various techniques have been explored in order to induce cooperation in agents: Modifying the environment rules, adding intrinsic rewards, and more. === Sequential social dilemmas === Social dilemmas like prisoner's dilemma, chicken and stag hunt are "matrix games". Each agent takes only one action from a choice of two possible actions, and a simple 2x2 matrix is used to describe the reward that each agent will get, given the actions that each agent took. In humans and other living creatures, social dilemmas tend to be more complex. Agents take multiple actions over time, and the distinction between cooperating and defecting is not as clear cut as in matrix games. The concept of a sequential social dilemma (SSD) was introduced in 2017 as an attempt to model that complexity. There is ongoing research into defining different kinds of SSDs and showing cooperative behavior in the agents that act in them. == Autocurricula == An autocurriculum (plural: autocurricula) is a reinforcement learning concept that's salient in multi-agent experiments. As agents improve their performance, they change their environment; this change in the environment affects themselves and the other agents. The feedback loop results in several distinct phases of learning, each depending on the previous one. The stacked layers of learning are called an autocurriculum. Autocurricula are especially apparent in adversarial settings, where each group of agents is racing to counter the current strategy of the opposing group. The Hide and Seek game is an accessible example of an autocurriculum occurring in an adversarial setting. In this experiment, a team of seekers is competing against a team of hiders. Whenever one of the teams learns a new strategy, the opposing team adapts its strategy to give the best possible counter. When the hiders learn to use boxes to build a shelter, the seekers respond by learning to use a ramp to break into that shelter. The hiders respond by locking the ramps, making them unavailable for the seekers to use. The seekers then respond by "box surfing", exploiting a glitch in the game to penetrate the shelter. Each "level" of learning is an emergent phenomenon, with the previous level as its premise. This results in a stack of behaviors, each dependent on its predecessor. Autocurricula in reinforcement learning experiments are compared to the stages of the evolution of life on Earth and the development of human culture. A major stage in evolution happened 2-3 billion years ago, when photosynthesizing life forms started to produce massive amounts of oxygen, changing the balance of gases in the atmosphere. In the next stages of evolution, oxygen-breathing life forms evolved, eventually leading up to land mammals and human beings. These later stages could only happen after the photosynthesis stage made oxygen widely available. Similarly, human culture could not have gone through the Industrial Revolution in the 18th century without the resources and insights gaine

  Read more →
• 

  Super column

  A super column is a tuple (a pair) with a binary super column name and a value that maps it to many columns. They consist of a key–value pairs, where the values are columns. Theoretically speaking, super columns are (sorted) associative array of columns. Similar to a regular column family where a row is a sorted map of column names and column values, a row in a super column family is a sorted map of super column names that maps to column names and column values. A super column is part of a keyspace together with other super columns and column families, and columns. == Code example == Written in the JSON-like syntax, a super column definition can be like this: Where: "databases" are keyspace; "Cassandra" and "HBase" are rowKeys; "name" and "address" are super column names; "firstName", "city", "age", etc. are column names.

  Read more →
• 

  Ubiquitous robot

  Ubiquitous robot is a term used in an analogous way to ubiquitous computing. Software useful for "integrating robotic technologies with technologies from the fields of ubiquitous and pervasive computing, sensor networks, and ambient intelligence". The emergence of mobile phone, wearable computers and ubiquitous computing makes it likely that human beings will live in a ubiquitous world in which all devices are fully networked. The existence of ubiquitous space resulting from developments in computer and network technology will provide motivations to offer desired services by any IT device at any place and time through user interactions and seamless applications. This shift has hastened the ubiquitous revolution, which has further manifested itself in the new multidisciplinary research area, ubiquitous robotics. It initiates the third generation of robotics following the first generation of the industrial robot and the second generation of the personal robot. Ubiquitous robot (Ubibot) is a robot incorporating three components including virtual software robot or avatar, real-world mobile robot and embedded sensor system in surroundings. Software robot within a virtual world can control a real-world robot as a brain and interact with human beings. Researchers of KAIST, Korea describe these three components as a Sobot (Software robot), Mobot (Mobile robot), and Embot (Embedded robot).

  Read more →
• 

  TurboQuant

  TurboQuant is an online vector quantization algorithm for compressing high-dimensional Euclidean vectors while preserving their geometric structure. It was proposed in 2025 by Amir Zandieh, Majid Daliri, Majid Hadian, and Vahab Mirrokni in the paper TurboQuant: Online Vector Quantization with Near-optimal Distortion Rate. The paper lists Zandieh and Mirrokni as affiliated with Google Research, Daliri with New York University, and Hadian with Google DeepMind. The method was developed for applications including large language model (LLM) inference, key–value (KV) cache compression, vector databases, and nearest neighbor search. TurboQuant consists of two related algorithms: TurboQuantmse, which is optimized for mean squared error (MSE), and TurboQuantprod, which is optimized for unbiased inner product estimation. The algorithm uses a random rotation of input vectors, applies scalar quantizers to the rotated coordinates, and, for inner-product estimation, applies a one-bit Quantized Johnson–Lindenstrauss (QJL) transform to the residual error. == Background == Vector quantization is a compression method that maps high-dimensional vectors to a finite set of codewords. The problem has roots in Shannon's source coding theory and rate–distortion theory. In machine learning and information retrieval, vector quantization is used to reduce the memory required to store embeddings, activation vectors, and other numerical representations. In Transformer-based large language models, the KV cache stores key and value vectors from previous tokens during autoregressive decoding. The size of this cache grows with context length, the number of attention heads, and the number of concurrent requests, making it a major memory bottleneck in LLM serving. Similar compression problems appear in vector search, where large collections of embedding vectors must be stored and searched efficiently. Earlier approaches to vector quantization include product quantization, scalar quantization, and data-dependent k-means codebook construction. The TurboQuant paper argues that many existing methods either require offline preprocessing and calibration or suffer from suboptimal distortion guarantees in online settings. == Algorithm == === TurboQuantmse === TurboQuantmse is the version of the algorithm optimized for mean-squared error. For a unit vector x ∈ S d − 1 {\displaystyle x\in S^{d-1}} , the algorithm first applies a random rotation matrix Π ∈ R d × d {\displaystyle \Pi \in \mathbb {R} ^{d\times d}} and sets z = Π x {\displaystyle z=\Pi x} . Each coordinate of the rotated vector follows a shifted and scaled beta distribution, which converges to a normal distribution in high dimensions. In high dimensions, distinct coordinates also become nearly independent, allowing the algorithm to apply scalar quantizers independently to each coordinate. The scalar quantizer is constructed by solving a one-dimensional continuous k-means or Lloyd–Max quantization problem. If the centroids are c 1 , c 2 , … , c 2 b {\displaystyle c_{1},c_{2},\ldots ,c_{2^{b}}} , the quantization step stores, for each coordinate, i d x j = ⁡ a r g m i n k ∈ [ 2 b ] | z j − c k | . {\displaystyle \mathrm {idx} _{j}=\operatorname {} {arg\,min}_{k\in [2^{b}]}|z_{j}-c_{k}|.} During dequantization, the stored index for each coordinate is replaced by the corresponding centroid, giving a reconstructed rotated vector z ~ {\displaystyle {\tilde {z}}} . The algorithm then rotates back: x ~ = Π ⊤ z ~ . {\displaystyle {\tilde {x}}=\Pi ^{\top }{\tilde {z}}.} The paper gives the following bound for TurboQuantmse: D m s e ≤ 3 π 2 ⋅ 1 4 b . {\displaystyle D_{\mathrm {mse} }\leq {\frac {\sqrt {3\pi }}{2}}\cdot {\frac {1}{4^{b}}}.} It also reports finer-grained MSE values of approximately 0.36, 0.117, 0.03, and 0.009 for bit-widths b = 1 , 2 , 3 , 4 {\displaystyle b=1,2,3,4} , respectively. === TurboQuantprod === TurboQuantprod is optimized for unbiased inner-product estimation. The authors note that an MSE-optimized quantizer may introduce bias when used to estimate inner products. To address this, TurboQuantprod first applies TurboQuantmse with bit-width b − 1 {\displaystyle b-1} , then applies a one-bit Quantized Johnson–Lindenstrauss transform to the remaining residual vector. Let r = x − Q m s e − 1 ( Q m s e ( x ) ) {\displaystyle r=x-Q_{\mathrm {mse} }^{-1}(Q_{\mathrm {mse} }(x))} be the residual after MSE quantization, and let γ = ‖ r ‖ 2 {\displaystyle \gamma =\|r\|_{2}} . The QJL step stores a sign vector for the residual. For γ ≠ 0 {\displaystyle \gamma \neq 0} , this can be written using the normalized residual u = r / γ {\displaystyle u=r/\gamma } : q j l = sign ⁡ ( S u ) , {\displaystyle qjl=\operatorname {sign} (Su),} where S ∈ R d × d {\displaystyle S\in \mathbb {R} ^{d\times d}} is a random projection matrix. Since the sign function is invariant under positive rescaling, this is equivalent to sign ⁡ ( S r ) {\displaystyle \operatorname {sign} (Sr)} when r ≠ 0 {\displaystyle r\neq 0} . If γ = 0 {\displaystyle \gamma =0} , the residual correction is zero. TurboQuantprod stores the MSE quantization, the QJL sign vector, and the residual norm: Q p r o d ( x ) = [ Q m s e ( x ) , q j l , γ ] . {\displaystyle Q_{\mathrm {prod} }(x)=\left[Q_{\mathrm {mse} }(x),qjl,\gamma \right].} The dequantized vector is reconstructed as x ~ = x ~ m s e + π / 2 d γ S ⊤ q j l . {\displaystyle {\tilde {x}}={\tilde {x}}_{\mathrm {mse} }+{\frac {\sqrt {\pi /2}}{d}}\,\gamma S^{\top }qjl.} The paper proves that TurboQuantprod is unbiased for inner-product estimation: E x ~ [ ⟨ y , x ~ ⟩ ] = ⟨ y , x ⟩ . {\displaystyle \mathbb {E} _{\tilde {x}}\left[\langle y,{\tilde {x}}\rangle \right]=\langle y,x\rangle .} It also gives the distortion bound D p r o d ≤ 3 π 2 ⋅ ‖ y ‖ 2 2 d ⋅ 1 4 b . {\displaystyle D_{\mathrm {prod} }\leq {\frac {\sqrt {3\pi }}{2}}\cdot {\frac {\|y\|_{2}^{2}}{d}}\cdot {\frac {1}{4^{b}}}.} == Performance and applications == The TurboQuant paper reports that the algorithm achieves near-optimal distortion rates within a small constant factor of information-theoretic lower bounds. The authors report that, for KV cache quantization, TurboQuant achieved quality neutrality at 3.5 bits per channel and marginal degradation at 2.5 bits per channel. In long-context LLM experiments using Llama 3.1 8B Instruct, the paper evaluated the method on a "needle-in-a-haystack" retrieval task with document lengths from 4,000 to 104,000 tokens. It reported that TurboQuant matched the uncompressed full-precision baseline while using more than 4× compression, and compared the method against PolarQuant, SnapKV, PyramidKV, and KIVI. Google Research stated that TurboQuant was evaluated on long-context benchmarks including LongBench, Needle in a Haystack, ZeroSCROLLS, RULER, and L-Eval using open-source models including Gemma and Mistral. According to a report in Tom's Hardware, Google described the method as reducing KV-cache memory by at least six times and achieving up to an eightfold improvement in attention-logit computation on Nvidia H100 GPUs compared with unquantized 32-bit keys. TurboQuant has also been applied to nearest-neighbor vector search. The original paper reports experiments on DBpedia entity embeddings and GloVe embeddings, comparing TurboQuant with product quantization and other vector-search quantization baselines. == Relationship to other methods == TurboQuant is related to several methods for efficient large language model inference and high-dimensional search: Product quantization – a vector quantization technique widely used for approximate nearest-neighbor search Quantization (machine learning) – reducing the numerical precision of weights, activations, or cached tensors in machine learning models PagedAttention – a memory-management algorithm for LLM serving that reduces fragmentation in the KV cache Johnson–Lindenstrauss lemma – a result in high-dimensional geometry used in random projection methods Lloyd's algorithm – an algorithm for scalar and vector quantization, including k-means-style codebook construction Unlike PagedAttention, which focuses on memory allocation and cache layout, TurboQuant reduces the numerical storage cost of the vectors themselves. Unlike many product-quantization methods, TurboQuant is designed to be data-oblivious and online, avoiding dataset-specific codebook training. == Limitations == The strongest performance claims for TurboQuant come from the original paper and Google Research's own publication. Coverage in technology media has noted that the broader impact of the method will depend on real-world implementation details, workloads, and hardware architectures.

  Read more →
• 

  Summify

  Summify was a social news aggregator founded by Mircea Paşoi and Cristian Strat, two former Google and Microsoft interns from Romania. The service emailed its users a periodic summary of news articles shared from their social networks based on their relevance and importance. The platform supported Twitter, Facebook, and Google Reader accounts. == History == In 2009, Paşoi and Strat created ReadFu, a plugin that provided a contextual summary and statistics of the target page of a hyperlink. In January 2010, ReadFu was accepted into the Vancouver-based start-up incubator Bootup Labs. On March 20, 2010 the service was renamed to Summify and a private beta began. On August 11, 2010 Paşoi and Strat announced a new direction for the service. It would become a real-time social news reader that aggregates incoming news from social networks and displays articles by importance using social reactions. After some feedback that the users preferred article digests by email more than the real-time news reader version, Summify discontinued the news reader version. In March 2011, Summify completed a Seed round, with investors including Rob Glaser, Accel Partners, and Stewart Butterfield. Summify received coverage from various news and media outlets such as TechCrunch. It was also featured in various news platforms, such as Time, The Globe and Mail, Mashable, VentureBeat, Gizmodo, Lifehacker, and The Next Web. Summify released a free app on the Apple App Store on July 8, 2011. The app allowed users to read their web summaries from iOS mobile devices. Summify was acquired by Twitter on January 19, 2012. The service shut down soon after, on June 22, 2012.

  Read more →
• 

  Artificial imagination

  Artificial imagination is a narrow subcomponent of artificial general intelligence which generates, simulates, and facilitates real or possible fiction models to create predictions, inventions, or conscious experiences. The term artificial imagination is also used to describe a property of machines or programs. Some of the traits that researchers hope to simulate include creativity, vision, digital art, humor, and satire. Practitioners in the field are researching various aspects of Artificial imagination, such as Artificial (visual) imagination, Artificial (aural) Imagination, modeling/filtering content based on human emotions and Interactive Search. Some articles on the topic speculate on how artificial imagination may evolve to create an artificial world "people may be comfortable enough to escape from the real world". Some researchers such as G. Schleis and M. Rizki have focused on using artificial neural networks to simulate artificial imagination. Another important project is being led by Hiroharu Kato and Tatsuya Harada at the University of Tokyo in Japan. They have developed a computer capable of translating a description of an object into an image, which could be the easiest way to define what imagination is. Their idea is based on the concept of an image as a series of pixels divided into short sequences that correspond to a specific part of an image. The scientists call this sequences "visual words" and those can be interpreted by the machine using statistical distribution to read an create an image of an object the machine has not encountered. The topic of artificial imagination has garnered interest from scholars outside the computer science domain, such as noted communications scholar Ernest Bormann, who came up with the Symbolic Convergence Theory and worked on a project to develop artificial imagination in computer systems. An interdisciplinary research seminar organized by the artist Grégory Chatonsky on artificial imagination and postdigital art has taken place since 2017 at the Ecole Normale Supérieure in Paris. == Use in interactive search == The typical application of artificial imagination is for an interactive search. Interactive searching has been developed since the mid-1990s, accompanied by the World Wide Web's development and the optimization of search engines. Based on the first query and feedback from a user, the databases to be searched are reorganized to improve the searching results. Artificial imagination allows us to synthesize images and to develop a new image, whether it is in the database, regardless its existence in the real world. For example, the computer shows results that are based on the answer from the initial query. The user selects several relevant images, and then the technology analyzes these selections and reorganizes the images' ranks to fit the query. In this process, artificial imagination is used to synthesize the selected images and to improve the searching result with additional relevant synthesized images. This technique is based on several algorithms, including the Rocchio algorithm and the evolutionary algorithm. The Rocchio algorithm, locating a query point near relevant examples and far away from irrelevant examples, is simple and works well in a small system where the databases are arranged in certain ranks. The evolutionary synthesis is composed of two steps: a standard algorithm and an enhancement of the standard algorithm. Through feedback from the user, there would be additional images synthesized so as to be suited to what the user is looking for. == General artificial imagination == Artificial imagination has a more general definition and wide applications. The traditional fields of artificial imagination include visual imagination and aural imagination. More generally, all the actions to form ideas, images and concepts can be linked to imagination. Thus, artificial imagination means more than only generating graphs. For example, moral imagination is an important research subfield of artificial imagination, although classification of artificial imagination is difficult. Morals are an important part to human beings' logic, while artificial morals are important in artificial imagination and artificial intelligence. A common criticism of artificial intelligence is whether human beings should take responsibility for machines' mistakes or decisions and how to develop well-behaved machines. As nobody can give a clear description of the best moral rules, it is impossible to create machines with commonly accepted moral rules. However, recent research about artificial morals circumvent the definition of moral. Instead, machine learning methods are applied to train machines to imitate human morals. As the data about moral decisions from thousands of different people are considered, the trained moral model can reflect widely accepted rules. Memory is another major field of artificial imagination. Researchers such as Aude Oliva have performed extensive work on artificial memory, especially visual memory. Compared to visual imagination, the visual memory focuses more on how machine understand, analyse and store pictures in a human way. In addition, characters like spatial features are also considered. As this field is based on the brains' biological structures, extensive research on neuroscience has also been performed, which makes it a large intersection between biology and computer science.

  Read more →
• 

  Spatial computing

  Spatial computing refers to 3D human–computer interaction techniques that are perceived by users as taking place in the real world, in and around their bodies and physical environments, instead of constrained to and perceptually behind computer screens or in purely virtual worlds. This concept inverts the long-standing practice of teaching people to interact with computers in digital environments, and instead teaches computers to better understand and interact with people more naturally in the human world. This concept overlaps with and encompasses others including extended reality, augmented reality, mixed reality, natural user interface, contextual computing, affective computing, and ubiquitous computing. The usage for labeling and discussing these adjacent technologies is imprecise. Spatial computing devices include sensors—such as RGB cameras, depth cameras, 3D trackers, inertial measurement units, or other tools—to sense and track nearby human bodies (including hands, arms, eyes, legs, mouths) during ordinary interactions with people and computers in a 3D space. They further use computer vision to attempt to understand real world scenes, such as rooms, streets or stores, to read labels, to recognize objects, create 3D maps, and more. Quite often they also use extended reality and mixed reality to superimpose virtual 3D graphics and virtual 3D audio onto the human visual and auditory system as a way of providing information more naturally and contextually than traditional 2D screens. Spatial computing often refers to personal computing devices like headsets and headphones, but other human-computer interactions that leverage real-time spatial positioning for displays, like projection mapping or cave automatic virtual environment displays, can also be considered spatial computing if they leverage human-computer input for the participants. == History == The term "spatial computing" apparently originated in the field of GIS around 1985 or earlier to describe computations on large-scale geospatial information. Early examples of spatial computing in GIS include ArcInfo and its iterations, initially released in 1981, a part of ArcGIS along with ArcEditor, which together provide mapping, analysis, editing, and geoprocessing for geodatabases. This is somewhat related to the modern use, but on the scale of continents, cities, and neighborhoods. Modern spatial computing is more centered on the human scale of interaction, around the size of a living room or smaller. But it is not limited to that scale in the aggregate. In the early 1990s, as field of virtual reality was beginning to be commercialized beyond academic and military labs, a startup called Worldesign in Seattle used the term Spatial Computing to describe the interaction between individual people and 3D spaces, operating more at the human end of the scale than previous GIS examples may have contemplated. The company built a CAVE-like environment it called the Virtual Environment Theater, whose 3D experience was of a virtual flyover of the Giza Plateau, circa 3000 BC. Robert Jacobson, CEO of Worldesign, attributes the origins of the term to experiments at the Human Interface Technology Lab, at the University of Washington, under the direction of Thomas A. Furness III. Jacobson was a co-founder of that lab before spinning off this early VR startup. In 1997, an academic publication by T. Caelli, Peng Lam, and H. Bunke called "Spatial Computing: Issues in Vision, Multimedia and Visualization Technologies" introduced the term more broadly for academic audiences, focusing on a variety of topics such as image processing, dead reckoning navigation, object recognition, and visualizing spatial data. The specific term "spatial computing" was later referenced again in 2003 by Simon Greenwold, as "human interaction with a machine in which the machine retains and manipulates referents to real objects and spaces". MIT Media Lab alumnus John Underkoffler gave a TED talk in 2010 giving a live demo of the multi-screen, multi-user spatial computing systems being developed by Oblong Industries, which sought to bring to life the futuristic interfaces conceptualized by Underkoffler in the films Minority Report and Iron Man. Google Earth, initially released by Keyhole Inc. in 2001 and re-released by Google in 2005 can be considered a capable GIS and includes advanced geospatial tools and capabilities. == Notable instances of the use of spatial computing == In 2019, Microsoft HoloLens released a video outlining Airbus' partnership with Microsoft Azure to utilize the latter's mixed reality services for streamlining and improving the aircraft design process, as well as reducing the error in development. Airbus utilized the HoloLens 2 to this end, and the executive vice president of engineering claimed that their design process' validation phases were "hugely accelerated by 80 percent", as well as "strongly believe[d]" that up to 30% improvements in their industrial tasks could be attained with the HoloLens 2. During the presentational video, Airbus cited the maturity of Microsoft Azure services as "key" for their usage of the HoloLens 2. Also in 2019, the U.S. army partnered with Microsoft to produce a HoloLens based Integrated Visual Augmentation System (IVAS) to enhance infantry members by giving troops various abilities, including but not limited to using holographs to train, projecting 3D maps into their vision, and seeing through smoke and corners. Microsoft received tens of thousands of hours of feedback for their systems by 2021. Sergeant Marc Krugh at the time claimed that Microsoft's partnership has already caused the army to rethink some of its troops' operation strategy. == Products == === Apple Vision Pro === Apple announced Apple Vision Pro, a device it markets as a "spatial computer", on June 5, 2023. It includes several features such as Spatial Audio, two 4K micro-OLED displays, the Apple R1 chip and eye tracking, and released in the United States on February 2, 2024. In announcing the platform, Apple invoked its history of popularizing 2D graphical user interfaces that supplanted prior human-computer interface mechanisms such as the command line. Apple suggests the introduction of spatial computing as a new category of interactive device, on the same level of importance as the introduction of the 2D GUI. Apple Vision Pro runs on a new operating system called visionOS, which combines eye tracking, gesture recognition, and voice input to enable immersive interaction without physical controllers. The platform is aimed at productivity, entertainment, collaboration, and enterprise use cases. === Magic Leap === Magic Leap had also previously used the term “spatial computing” to describe its own devices. Its first headset, the Magic Leap 1, was released on August 8, 2018. Magic Leap’s technology enables the display of content into the real world using an optical see-through head-mounted display, which projects an overlay of a virtual world into the user’s field of view. This allows for an experience where the physical and digital worlds are perceived simultaneously. === Microsoft Hololens === On February 24, 2019, Microsoft released the HoloLens 2, which includes mixed reality tools and can generate interactable, manipulatable holograms in 3D space. The holograms in question can be related to a physical object or completely independent and free-floating. The Azure Spatial Anchors cloud service was released simultaneously, which gives the holograms capability to persist across time and many individuals' devices. === Meta Quest === The Meta Quest 3, a mixed reality gaming headset that includes spatial audio, two color cameras, and grants the ability to interact with virtual characters released on October 9, 2023, at a notably cheaper price than the Apple Vision Pro, but with reduced capabilities. === Snap Spectacles === Spectacles (product) are augmented reality glasses developed by Snap Inc.. The latest generation includes a 46-degree stereoscopic display, adjustable tint, and Snapdragon processors. Spectacles allow users to interact with a collection of augmented reality experiences designed for education, entertainment, and utility. Currently, the device is in the hands of selected developers and creators, as part of an experimental AR ecosystem focused on creativity, use case exploration and expression.

  Read more →
• 

  Wearable computer

  A wearable computer, also known as a body-borne computer or wearable, is a computing device worn on the body. The definition of 'wearable computer' may be narrow or broad, extending to smartphones or even ordinary wristwatches. Wearables may be for general use, in which case they are just a particularly small example of mobile computing. Alternatively, they may be for specialized purposes such as fitness trackers. They may incorporate special sensors such as accelerometers, heart rate monitors, or on the more advanced side, electrocardiogram (ECG) and blood oxygen saturation (SpO2) monitors. Under the definition of wearable computers, we also include novel user interfaces such as Google Glass, an optical head-mounted display controlled by gestures. It may be that specialized wearables will evolve into general all-in-one devices, as happened with the convergence of PDAs and mobile phones into smartphones. Wearables are typically worn on the wrist (e.g. fitness trackers), hung from the neck (like a necklace), strapped to the arm or leg (electronic tagging), or on the head (as glasses or a helmet), though some have been located elsewhere (e.g. on a finger or in a shoe). Devices carried in a pocket or bag – such as smartphones and before them, pocket calculators and PDAs, may or may not be regarded as 'worn'. Wearable computers have various technical issues common to other mobile computing, such as batteries, heat dissipation, software architectures, wireless and personal area networks, and data management. Many wearable computers are active all the time, e.g. processing or recording data continuously. == Applications == Wearable computers are not only limited to computers such as fitness trackers that are worn on wrists; they also include wearables such as heart pacemakers and other prosthetics. They are used most often in research that focuses on behavioral modeling, health monitoring systems, IT and media development, where the person wearing the computer actually moves or is otherwise engaged with his or her surroundings. Wearable computers have been used for the following: general-purpose computing (e.g. smartphones and smartwatches) sensory integration, e.g. to help people see better or understand the world better (whether in task-specific applications like camera-based welding helmets or for everyday use like Google Glass) behavioral modeling health care monitoring systems service management electronic textiles and fashion design, e.g. Microsoft's 2011 prototype "The Printing Dress". Wearable computing is the subject of active research, especially the form-factor and location on the body, with areas of study including user interface design, augmented reality, and pattern recognition. The use of wearables for specific applications, for compensating disabilities or supporting elderly people steadily increases. == Operating systems == The dominant operating systems for wearable computing are: FreeRTOS is a real-time operating system kernel for embedded devices; most of the Smartbands that are currently available in the market are based on FreeRTOS, which include Huawei, Honor, Lenovo, realme, TCL and Xiaomi smartbands. LiteOS is a lightweight open source real-time operating system that is part of Huawei's "1+8+N" Internet of Things solution. Tizen OS from Samsung (there was an announcement in May 2021 that Wear OS and Tizen OS will merge and will be called simply Wear.) watchOS watchOS is a proprietary mobile operating system developed by Apple Inc. to run on the Apple Watch. Wear OS Wear OS (previously known as Android Wear) is a smartwatch operating system developed by Google Inc. == History == Due to the varied definitions of wearable and computer, the first wearable computer could be as early as the first abacus on a necklace, a 16th-century abacus ring, a wristwatch and 'finger-watch' owned by Queen Elizabeth I of England, or the covert timing devices hidden in shoes to cheat at roulette by Thorp and Shannon in the 1960s and 1970s. However, a general-purpose computer is not merely a time-keeping or calculating device, but rather a user-programmable item for arbitrary complex algorithms, interfacing, and data management. By this definition, the wearable computer was invented by Steve Mann, in the late 1970s: Steve Mann, a professor at the University of Toronto, was hailed as the father of the wearable computer and the ISSCC's first virtual panelist, by moderator Woodward Yang of Harvard University (Cambridge Mass.). The development of wearable items has taken several steps of miniaturization from discrete electronics over hybrid designs to fully integrated designs, where just one processor chip, a battery, and some interface conditioning items make the whole unit. === 1500s === Queen Elizabeth I of England received a watch from Robert Dudley in 1571, as a New Year's present; it may have been worn on the forearm rather than the wrist. She also possessed a 'finger-watch' set in a ring, with an alarm that prodded her finger. === 1600s === The Qing dynasty saw the introduction of a fully functional abacus on a ring, which could be used while it was being worn. === 1960s === In 1961, mathematicians Edward O. Thorp and Claude Shannon built some computerized timing devices to help them win a game of roulette. One such timer was concealed in a shoe and another in a pack of cigarettes. Various versions of this apparatus were built in the 1960s and 1970s. Thorp refers to himself as the inventor of the first "wearable computer". In other variations, the system was a concealed cigarette-pack-sized analog computer designed to predict the motion of roulette wheels. A data-taker would use microswitches hidden in his shoes to indicate the speed of the roulette wheel, and the computer would indicate an octant of the roulette wheel to bet on by sending musical tones via radio to a miniature speaker hidden in a collaborator's ear canal. The system was successfully tested in Las Vegas in June 1961, but hardware issues with the speaker wires prevented it from being used beyond test runs. This was not a wearable computer because it could not be re-purposed during use; rather it was an example of task-specific hardware. This work was kept secret until it was first mentioned in Thorp's book Beat the Dealer (revised ed.) in 1966 and later published in detail in 1969. === 1970s === Pocket calculators became mass-market devices in 1970, starting in Japan. Programmable calculators followed in the late 1970s, being somewhat more general-purpose computers. The HP-01 algebraic calculator watch by Hewlett-Packard was released in 1977. A camera-to-tactile vest for the blind, launched by C.C. Collins in 1977, converted images into a 1024-point, ten-inch square tactile grid on a vest. === 1980s === The 1980s saw the rise of more general-purpose wearable computers. In 1981, Steve Mann designed and built a backpack-mounted 6502-based wearable multimedia computer with text, graphics, and multimedia capability, as well as video capability (cameras and other photographic systems). Mann went on to be an early and active researcher in the wearables field, especially known for his 1994 creation of the Wearable Wireless Webcam, the first example of lifelogging. Seiko Epson released the RC-20 Wrist Computer in 1984. It was an early smartwatch, powered by a computer on a chip. In 1989, Reflection Technology marketed the Private Eye head-mounted display, which scans a vertical array of LEDs across the visual field using a vibrating mirror. This display gave rise to several hobbyist and research wearables, including Gerald "Chip" Maguire's IBM/Columbia University Student Electronic Notebook, Doug Platt's Hip-PC, and Carnegie Mellon University's VuMan 1 in 1991. The Student Electronic Notebook consisted of the Private Eye, Toshiba diskless AIX notebook computers (prototypes), a stylus based input system and a virtual keyboard. It used direct-sequence spread spectrum radio links to provide all the usual TCP/IP based services, including NFS mounted file systems and X11, which all ran in the Andrew Project environment. The Hip-PC included an Agenda palmtop used as a chording keyboard attached to the belt and a 1.44 megabyte floppy drive. Later versions incorporated additional equipment from Park Engineering. The system debuted at "The Lap and Palmtop Expo" on 16 April 1991. VuMan 1 was developed as part of a Summer-term course at Carnegie Mellon's Engineering Design Research Center, and was intended for viewing house blueprints. Input was through a three-button unit worn on the belt, and output was through Reflection Tech's Private Eye. The CPU was an 8 MHz 80188 processor with 0.5 MB ROM. === 1990s === In the 1990s PDAs became widely used, and in 1999 were combined with mobile phones in Japan to produce the first mass-market smartphone. In 1993, the Private Eye was used in Thad Starner's wearable, based on Doug Platt's system and built from a kit from Park Enterprises, a Pri

  Read more →
• 

  Elasticity (data store)

  The elasticity of a data store relates to the flexibility of its data model and clustering capabilities. The greater the number of data model changes that can be tolerated, and the more easily the clustering can be managed, the more elastic the data store is considered to be. == Types == === Clustering elasticity === Clustering elasticity is the ease of adding or removing nodes from the distributed data store. Usually, this is a difficult and delicate task to be done by an expert in a relational database system. Some NoSQL data stores, like Apache Cassandra have an easy solution, and a node can be added/removed with a few changes in the properties and by adding specifying at least one seed. === Data-modelling elasticity === Relational databases are most often very inelastic, as they have a predefined data model that can only be adapted through redesign. Most NoSQL data stores, however, do not have a fixed schema. Each row can have a different number and even different type of columns. Concerning the data store, modifications in the schema are no problem. This makes this kind of data stores more elastic concerning the data model. The drawback is that the programmer has to take into account that the data model may change over time.

  Read more →
• 

  FAIR data

  FAIR data is data which meets the 2016 FAIR principles of findability, accessibility, interoperability, and reusability (FAIR). The FAIR principles emphasize machine-actionability (i.e., the capacity of computational systems to find, access, interoperate, and reuse data with none or minimal human intervention) because humans increasingly rely on computational support to deal with data as a result of the increase in the volume, complexity, and rate of production of data. The abbreviation FAIR/O data is sometimes used to indicate that the dataset or database in question complies with the FAIR principles and also carries an explicit data‑capable open license. == FAIR principles published by GO FAIR == Findable The first step in (re)using data is to find them. Metadata and data should be easy to find for both humans and computers. Machine-readable metadata are essential for automatic discovery of datasets and services, so this is an essential component of the FAIRification process. F1. (Meta)data are assigned a globally unique and persistent identifier F2. Data are described with rich metadata (defined by R1 below) F3. Metadata clearly and explicitly include the identifier of the data they describe F4. (Meta)data are registered or indexed in a searchable resource Accessible Once the user finds the required data, they need to know how they can be accessed, possibly including authentication and authorisation. A1. (Meta)data are retrievable by their identifier using a standardised communications protocol A1.1 The protocol is open, free, and universally implementable A1.2 The protocol allows for an authentication and authorisation procedure, where necessary A2. Metadata are accessible, even when the data are no longer available Interoperable The data usually need to be integrated with other data. In addition, the data need to interoperate with applications or workflows for analysis, storage, and processing. I1. (Meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation I2. (Meta)data use vocabularies that follow FAIR principles I3. (Meta)data include qualified references to other (meta)data Reusable The ultimate goal of FAIR is to optimise the reuse of data. To achieve this, metadata and data should be well-described so that they can be replicated and/or combined in different settings. R1. (Meta)data are richly described with a plurality of accurate and relevant attributes R1.1. (Meta)data are released with a clear and accessible data usage license R1.2. (Meta)data are associated with detailed provenance R1.3. (Meta)data meet domain-relevant community standards The principles refer to three types of entities: data (or any digital object), metadata (information about that digital object), and infrastructure. For instance, principle F4 defines that both metadata and data are registered or indexed in a searchable resource (the infrastructure component). === Acceptance and implementation === Before FAIR, a 2007 OECD report was the most influential paper discussing similar ideas related to data accessibility. In January 2014, the Lorentz Centre at Leiden University hosted a workshop entitled "Jointly designing a data FAIRPORT" where the participants first formulated the FAIR principles. After further discussions, they were published in the March 2016 issue of Scientific Data. At the 2016 G20 Hangzhou summit, the G20 leaders issued a statement endorsing the application of FAIR principles to research. Also in 2016, a group of Australian organisations developed a Statement on FAIR Access to Australia's Research Outputs, which aimed to extend the principles to research outputs more generally. In 2017, Germany, Netherlands and France agreed to establish an international office to support the FAIR initiative, the GO FAIR International Support and Coordination Office. Other international organisations active in the research data ecosystem, such as CODATA or Research Data Alliance (RDA) also support FAIR implementations by their communities. FAIR principles implementation assessment is being explored by FAIR Data Maturity Model Working Group of RDA, CODATA's strategic Decadal Programme "Data for Planet: Making data work for cross-domain challenges" mentions FAIR data principles as a fundamental enabler of data driven science. The Association of European Research Libraries recommends the use of FAIR principles. A 2017 paper by advocates of FAIR data reported that awareness of the FAIR concept was increasing among various researchers and institutes, but also, understanding of the concept was becoming confused as different people apply their own differing perspectives to it. Guides on implementing FAIR data practices state that the cost of a data management plan in compliance with FAIR data practices should be 5% of the total research budget. In 2019 the Global Indigenous Data Alliance (GIDA) released the CARE Principles for Indigenous Data Governance as a complementary guide. The CARE principles extend principles outlined in FAIR data to include Collective benefit, Authority to control, Responsibility, and Ethics to ensure data guidelines address historical contexts and power differentials. The CARE Principles for Indigenous Data Governance were drafted at the International Data Week and Research Data Alliance Plenary co-hosted event, "Indigenous Data Sovereignty Principles for the Governance of Indigenous Data Workshop", held 8 November 2018, in Gaborone, Botswana. The lack of information on how to implement the guidelines have led to inconsistent interpretations of them. In January 2020, representatives of nine groups of universities around the world produced the Sorbonne declaration on research data rights, which included a commitment to FAIR data, and called on governments to provide support to enable it. In 2021, researchers identified the FAIR principles as a conceptual component of data catalog software tools, with the other components being metadata management, business context and data responsibility roles. In April 2022, Matthias Scheffler and colleagues argued in Nature that FAIR principles are "a must" so that data mining and artificial intelligence can extract useful scientific information from the data. There have been moves in the geosciences to establish FAIR data by use of decimal georeferencing However, making data (and research outcomes) FAIR is a challenging task, and it is challenging to assess the FAIRness. In 2020, the FAIR Data Maturity Model Working Group published a set of guidelines for assessing "FAIRness".

  Read more →
• 

  Xulvi-Brunet–Sokolov algorithm

  Xulvi-Brunet and Sokolov's algorithm generates networks with chosen degree correlations. This method is based on link rewiring, in which the desired degree is governed by parameter ρ. By varying this single parameter it is possible to generate networks from random (when ρ = 0) to perfectly assortative or disassortative (when ρ = 1). This algorithm allows to keep network's degree distribution unchanged when changing the value of ρ. == Assortative model == In assortative networks, well-connected nodes are likely to be connected to other highly connected nodes. Social networks are examples of assortative networks. This means that an assortative network has the property that almost all nodes with the same degree are linked only between themselves. The Xulvi-Brunet–Sokolov algorithm for this type of networks is the following. In a given network, two links connecting four different nodes are chosen randomly. These nodes are ordered by their degrees. Then, with probability ρ, the links are randomly rewired in such a way that one link connects the two nodes with the smaller degrees and the other connects the two nodes with the larger degrees. If one or both of these links already existed in the network, the step is discarded and is repeated again. Thus, there will be no self-connected nodes or multiple links connecting the same two nodes. Different degrees of assortativity of a network can be achieved by changing the parameter ρ. Assortative networks are characterized by highly connected groups of nodes with similar degree. As assortativity grows, the average path length and clustering coefficient increase. == Disassortative model == In disassortative networks, highly connected nodes tend to connect to less-well-connected nodes with larger probability than in uncorrelated networks. Examples of such networks include biological networks. The Xulvi-Brunet and Sokolov's algorithm for this type of networks is similar to the one for assortative networks with one minor change. As before, two links of four nodes are randomly chosen and the nodes are ordered with respect to their degrees. However, in this case, the links are rewired (with probability p) such that one link connects the highest connected node with the node with the lowest degree and the other link connects the two remaining nodes randomly with probability 1 − ρ. Similarly, if the new links already existed, the previous step is repeated. This algorithm does not change the degree of nodes and thus the degree distribution of the network.

  Read more →