AI Code Tester

AI Code Tester — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

Mooky (app)

Mooky was a location-based social and dating application, designed to help its users to find the perfect match by providing a large scale of filters. Mooky was free of charge. The app made use of mobile devices' geolocation, a feature of smart phones and other devices which allows users to locate other users who are nearby. == History == Mooky was published on Google Play on April 17, 2016, by Mooky BV. The latest version of this application was version 1.0.6. == Overview == === How it works === Mooky used Facebook to build a user profile with photos and basic information, like the user's surname and age. From there on the user had to fill in their Mooky profile, which contains information about the user's height, posture, hair color, eye color, ethnicity and religion. After this the user could select its preferences to find matches nearby. === User verification === Mooky asked their users to take a selfie holding a piece of paper saying 'Mooky'. Mooky would then manually accept or decline the user verification.
Read more →
Metaclass (knowledge representation)

In knowledge representation, particularly in the Semantic Web, a metaclass is a class whose instances can themselves be classes. Similar to their role in programming languages, metaclasses in ontology languages can have properties otherwise applicable only to individuals, while retaining the same class's ability to be classified in a concept hierarchy. This enables knowledge about instances of those metaclasses to be inferred by semantic reasoners using statements made in the metaclass. Metaclasses thus enhance the expressivity of knowledge representations in a way that can be intuitive for users. While classes are suitable to represent a population of individuals, metaclasses can, as one of their feature, be used to represent the conceptual dimension of an ontology. Metaclasses are supported in the Web Ontology Language (OWL) and the data-modeling vocabulary RDFS. Metaclasses are often modeled by setting them as the object of claims involving rdf:type and rdfs:subClassOf—built-in properties commonly referred to as instance of and subclass of. Instance of entails that the subject of the claim is an instance, i.e. an individual that is a member of a class. Subclass of entails that the subject is a class. In the context of instance of and subclass of, the key difference between metaclasses and ordinary classes is that metaclasses are the object of instance of claims used on a class, while ordinary classes are not objects of such claims. (e.g. in a claim Bob instance of Human, Bob is the subject and an Instance, while the object, Human, is an ordinary class; but a further claim that Human instance of Animal species makes "Animal species" a metaclass because it has a member, "Human", that is also a Class). OWL 2 DL supports metaclasses by a feature called punning, in which one entity is interpreted as two different types of thing—a class and an individual—depending on its syntactic context. For example, through punning, an ontology could have a concept hierarchy such as Harry the eagle instance of golden eagle, golden eagle subclass of bird, and golden eagle instance of species. In this case, the punned entity would be golden eagle, because it is represented as a class (second claim) and an instance (third claim); whereas the metaclass would be species, as it has an instance that is a class. Punning also enables other properties that would otherwise be applicable only to ordinary instances to be used directly on classes, for example "golden eagle conservation status least concern." Having arisen from the fields of knowledge representation, description logic and formal ontology, Semantic Web languages have a closer relationship to philosophical ontology than do conventional programming languages such as Java or Python. Accordingly, the nature of metaclasses is informed by philosophical notions such as abstract objects, the abstract and concrete, and type-token distinction. Metaclasses permit concepts to be construed as tokens of other concepts while retaining their ontological status as types. This enables types to be enumerated over, while preserving the ability to inherit from types. For example, metaclasses could allow a machine reasoner to infer from a human-friendly ontology how many elements are in the periodic table, or, given that number of protons is a property of chemical element and isotopes are a subclass of elements, how many protons exist in the isotope hydrogen-2. Metaclasses are sometime organized by levels, in a similar way to the simple Theory of types where classes that are not metaclasses are assigned the first level, classes of classes in the first level are in the second level, classes of classes in the second level on the next and so on. == Examples == Following the type-token distinction, real world objects such as Abraham Lincoln or the planet Mars are regrouped into classes of similar objects. Abraham Lincoln is said to be an instance of human, and Mars is an instance of planet. This is a kind of is-a relationship. Metaclasses are class of classes, such as for example the nuclide concept. In chemistry, atoms are often classified as elements and, more specifically, isotopes. The glass of water one last drank has many hydrogen atoms, each of which is an instance of hydrogen. Hydrogen itself, a class of atoms, is an instance of nuclide. Nuclide is a class of classes, hence a metaclass. == Implementations == === RDF and RDFS === In RDF, the rdf:type property is used to state that a resource is an instance of a class. This enables metaclasses to be easily created by using rdf:type in a chain-like fashion. For example, in the two triples the resource species is a metaclass, because golden eagle is used as a class in the first statement and the class golden eagle is said to be an instance of the class species in the second statement. This way of doing allows :species to have non-class instances. RDF also provides rdf:Property as a way to create properties beyond those defined in the built-in vocabulary. Properties can be used directly on metaclasses, for example "species quantity 8.7 million", where quantity is a property defined via rdf:Property and species is a metaclass per the preceding example above. RDFS, an extension of RDF, introduced rdfs:Class and rdfs:subClassOf and enriched how vocabularies can classify concepts. Whereas rdf:type enables vocabularies to represent instantiation, the property rdfs:subClassOf enables vocabularies to represent subsumption. RDFS thus makes it possible for vocabularies to represent taxonomies, also known as subsumption hierarchies or concept hierarchies, which is an important addition to the type–token distinction made possible by RDF. Notably, the resource rdfs:Class is an instance of itself, demonstrating both the use of metaclasses in the language's internal implementation and a reflexive usage of rdf:type. RDFS is its own metamodel. This allows a second way to express that a resource is a metaclass. A triple to instantiate rdfs:Class, for example :golden_eagle rdf:type rdfs:Class will declare :golden_eagle as a class. It's also possible to subclass the rdfs:Class resource to declare a meta-class resource, for example :species rdfs:SubclassOf. By deduction, any instance of :species is then a class, so it is a class with class-instances, a meta-class.. This second way does not allows non-class instances of species and explicitly declares :tpecies as a meta-class. === OWL === In some OWL flavors like OWL1-DL, entities can be either classes or instances, but cannot be both. This limitations forbids metaclasses and metamodeling. This is not the case in the OWL1 full flavor, but this allows the model to be computationally undecidable. In OWL2, metaclasses can implemented with punning, that is a way to treat classes as if they were individuals. Other approaches have also been proposed and used to check the properties of ontologies at a meta level. ==== Punning ==== OWL 2 supports metaclasses through a feature called punning. In metaclasses implemented by punning, the same subject is interpreted as two fundamentally different types of thing—a class and an individual—depending on its syntactic context. This is similar to a pun in natural language, where different senses of the same word are emphasized to illustrate a point. Unlike in natural language, where puns are typically used for comedic or rhetorical effect, the main goal of punning in Semantic Web technologies is to make concepts easier to represent, closer to how they are discussed in everyday speech or academic literature. Although OWL 2 permits the same symbol to assume different roles, its standard semantics (known as Direct Semantics) still interprets the symbol differently depending on whether it is used as an individual, a class, or a property. === Protégé === In the ontology editor Protégé, metaclasses are templates for other classes who are their instances. == Classification == Some ontologies like the Cyc AI project's classifies classes and metaclasses. Classes are divided into fixed-order classes and variable-order classes. In the case of fixed-order classes, an order is attributed for metaclasses by measuring the distance to individuals with respect to the number of "instance of" triples that are necessary to find an individual. Classes that are not metaclasses are classes of individuals, so their order is "1" (first-order classes). Metaclasses that are classes of first-order classes' order is "2" (second-order classes), and so on. Variable-order metaclasses, on the other hand, can have instances; one example of variable-order metaclass is the class of all fixed-order classes.
Read more →
Issue tree

An issue tree, also called logic tree, is a graphical breakdown of a question that dissects it into its different components vertically and that progresses into details as it reads to the right. Issue trees are useful in problem solving to identify the root causes of a problem as well as to identify its potential solutions. They also provide a reference point to see how each piece fits into the whole picture of a problem. == Types == According to professor of strategy Arnaud Chevallier, elaborating an approach used at McKinsey & Company, there are two types of issue trees: diagnostic ones and solution ones. Diagnostic trees break down a "why" key question, identifying all the possible root causes for the problem. Solution trees break down a "how" key question, identifying all the possible alternatives to fix the problem. == Rules == Four basic rules can help ensure that issue trees are optimal, according to Chevallier: Consistently answer a "why" or a "how" question Progress from the key question to the analysis as it moves to the right Have branches that are mutually exclusive and collectively exhaustive (MECE) Use an insightful breakdown The requirement for issue trees to be collectively exhaustive implies that divergent thinking is a critical skill. == Applications == === In management interviews === Issue trees are used to answer questions in case interviews for management consulting positions. A quantitative type of question, the market sizing question, requires the interviewee to estimate the size of a data group such as a specific segment of a population, an amount of objects, a company's revenues, or similar. The candidates are expected to use a structured and logical method of arriving at their answer, and using an issue tree provides a diagram to aid the candidate's logical reasoning. Issue trees are used for other types of case interview questions as well.
Read more →
Z.ai

Knowledge Atlas Technology Joint Stock Co., Ltd., branded internationally as Z.ai, is a Chinese technology company specializing in artificial intelligence (AI). The company was formerly known as Zhipu AI outside China until its rebranding in 2025. Z.ai's flagship product is the GLM (General Language Model) family of large language models, which the company has released under the free and open-source MIT License since July 2025. As of 2024, it is one of China's "AI tiger" companies by investors and considered to be the third-largest LLM market player in China's AI industry according to the International Data Corporation. In January 2025, the United States Commerce Department blacklisted the company in its Entity List due to national security concerns. == History == Founded in 2019, the startup company began from Tsinghua University and was later spun out as an independent company. Researchers published an Association for Computational Linguistics conference paper in May 2022 introducing the GLM (General Language Model) training algorithm, which uses an "autoregressive blank infilling" strategy that creates cloze tests by randomly removing segments of input text and trains the model to autoregressively regenerate the removed text. In 2023, it raised 2.5 billion yuan (approx. 350 million in USD) from Alibaba Group and Tencent, along with Meituan, Ant Group, Xiaomi, and HongShan. In March 2024, Zhipu AI announced it was developing a Sora-like technology to achieve artificial general intelligence (AGI). In May 2024, the Saudi Arabian finance firm Prosperity7 Ventures, LLC participated in a USD $400 million financing round for Zhipu AI with a valuation of approximately 3 billion USD. In July 2024, they debuted the Ying text-to-video model. Zhipu released GLM-4-Plus in August 2024. In October 2024, Zhipu released GLM-4-Voice, an end-to-end speech large language model that can adjust its tone or dialect. Zhipu disclosed in April 2025 that it had started preparing for its initial public offering (IPO) and released two models under the free and open-source MIT License. In May 2025, the company sealed a 61.28 million yuan deal from the Chinese government for city projects in Hangzhou. In July 2025, Zhipu AI released GLM-4.5 and GLM-4.5 Air, their next generation language models, and the company rebranded itself as Z.ai internationally. In August 2025, Z.ai announced that their GLM models are compatible with Huawei's Ascend processors. On August 11, 2025, Z.ai released a new vision-language model (VLM) with a total of 106B parameters, GLM-4.5V. In late September 2025, the company released GLM-4.6 using China's domestic chips such as those from Cambricon Technologies. Z.ai released GLM-4.6V and GLM-4.7 in December 2025. That same year, the company changed its official name to Knowledge Atlas Technology JSC Ltd. On 8 January 2026, Z.ai held its IPO on the Hong Kong Stock Exchange to become a listed company. It is considered to be China's first major LLM company that went through an IPO. On February 11, 2026, Z.ai released GLM-5. In late February 2026, Z.ai's shares fell by 23%, and had a shortage of compute resources, leading to user complaints and Z.ai issuing a public call for support. Z.ai also restricted new user signups. In late March, 2026, Z.ai released the GLM-5.1 model to subscription users. On April 8th, 2026, Z.ai released GLM-5.1 as open-source. The same day, Z.ai increased its API prices by 10%, but maintained a lower price than its United States competitor Anthropic's Opus 4.6 model. On release, the company's share price increased 11.5%. == Description == Z.ai provides the following products and services: General Language Model (commonly abbreviated as GLM; formerly known as ChatGLM), a series of pre-trained dialogue models initially developed by Zhipu AI and Tsinghua KEG in 2023. GLM 4.5, released in July 2025 by Z.ai, can run on eight NVIDIA H20 chips. The release of GLM-4.6 in late September 2025 marked the first integration of FP8 and Int4 quantization on Cambricon chips. It also supports native FP8 on Moore Threads GPUs. Ying, a text-to-video model that generates image and text prompts into a six-second video clip for around 30 seconds. AutoGLM, an AI agent application that uses voice commands to complete tasks within a smartphone. The app can analyze complex tasks such as ordering an item from a nearby store and repeating an order based from the user's shopping history. AMiner, created by Jie Tang (co-founder of Z.ai) in March 2006, now owned by Z.ai. Z.ai has offices in the Middle East, United Kingdom, Singapore, and Malaysia, along with innovation center projects across Southeast Asia (2025). In January 2025, the United States Commerce Department added the company to its Entity List, citing national security concerns. == Models ==
Read more →
Internet Security Awareness Training

Internet Security Awareness Training (ISAT) is the training given to members of an organization regarding the protection of various information assets of that organization. ISAT is a subset of general security awareness training (SAT). Even small and medium enterprises are generally recommended to provide such training, but organizations that need to comply with government regulations (e.g., the Gramm–Leach–Bliley Act, the Payment Card Industry Data Security Standard, Health Insurance Portability and Accountability Act, Sarbanes–Oxley Act) normally require formal ISAT for annually for all employees. Often such training is provided in the form of online courses. ISAT, also referred to as Security Education, Training, and Awareness (SETA), organizations train and create awareness of information security management within their environment. It is beneficial to organizations when employees are well trained and feel empowered to take important actions to protect themselves and organizational data. The SETA program target must be based on user roles within organizations and for positions that expose the organizations to increased risk levels, specialized courses must be required. == Coverage == There are general topics to cover for the training, but it is necessary for each organization to have a coverage strategy based on its needs, as this will ensure the training is practical and captures critical topics relevant to the organization. As the threat landscape changes very frequently, organizations should continuously review their training programs to ensure relevance with current trends. Topics covered in ISAT include: Appropriate methods for protecting sensitive information on personal computer systems, including password policy Various computer security concerns, including spam, malware, phishing, social engineering, etc. Consequences of failure to properly protect information, including potential job loss, economic consequences to the firm, damage to individuals whose private records are divulged, and possible civil and criminal law penalties. Being Internet Security Aware means you understand that there are people actively trying to steal data that is stored within your organization's computers. (This often focuses on user names and passwords, so that criminal elements can ultimately get access to bank accounts and other high-value IT assets.) That is why it is important to protect the assets of the organization and stop that from happening. The general scope should include topics such as password security, Email phishing, Social engineering, Mobile device security, Sensitive data security, and Business communications. In contrast, those requiring specialized knowledge are usually required to take technical and in-depth training courses. Suppose an organization determines that it is best to use one of the available training tools on the market, it must ensure it sets objectives that the training can meet, including confirming the training will provide employees with the knowledge to understand risks and the behaviors needed in managing them, actions to take to prevent or detect security incidents, using language easily understandable by the trainees, and ensuring the pricing is reasonable. Organizations are recommended to base ISAT training content on employee roles and their culture; the policy should guide that training for all employees and gave the following as examples of sources of reference materials: National Institute of Standards and Technology (NIST) Special Publication 800-50, Building an Information Technology Security Awareness and Training Program International Standards Organization (ISO) 27002:2013, Information technology—Security techniques—Code of practice for information security controls International Standards Organization (ISO) 27001:2013, Information technology — Security techniques — Information security management systems COBIT 5 Appendix F.2, Detailed Guidance: Services, Infrastructure and Applications Enabler, Security Awareness The training must focus on current threats specific to an organization and the impacts if that materializes as a result of user actions. Including practical examples and ways of dealing with scenarios help users know the appropriate measures to take. It is a good practice to periodically train customers of specific organizations on threats they face from people with malicious intentions. Coverage strategy for SAT should be driven by an organization's policy. It can help truly determine the level of depth of the training and where it should be conducted at a global level or business unit level, or a combination of both. A policy also empowers a responsible party within the organization to run the training. == Importance == Studies show that well-structured security awareness training can significantly reduce the likelihood of cyber incidents caused by human error. According to the Ponemon Institute, organizations that implement regular security training experience up to 70% fewer successful phishing attacks. Additionally, a 2023 Verizon Data Breach Investigations Report found that 74% of breaches involve the human element, highlighting the need for continuous education. Employees are key in whether organizations are breached or not; there must be a policy on creating awareness and training them on emerging threats and actions to take in safeguarding sensitive information and reporting any observed unusual activity within the corporate environment. Research has shown that SAT has helped reduce cyber-attacks within organizations, especially when it comes to phishing, as trainees learned to identify these attack modes and give them the self-assurance to take action appropriately. There is an increase in phishing attacks, and it has become increasingly important for people to understand how to these attacks work, and the actions required to prevent these and SAT has shown a significant impact on the number of successful phishing attacks against organizations. == Compliance Requirements == Various regulations and laws mandate SAT for organizations in specific industries, including the Gramm–Leach–Bliley Act (GLBA) for the financial services, the Federal Information Security Modernization Act of 2014 for federal agencies, and the European Union's General Data Protection Regulation (GDPR). === Federal Information Security Modernization Act === Employees and contractors in federal agencies are required to receive Security Awareness Training annually, and the program needs to address job-related information security risks linked that provide them with the knowledge to lessen security risks. === Health Insurance Portability and Accountability Act === The Health Insurance Portability and Accountability Act has the Security Rule, and Privacy Rule requiring the creation of a security awareness training program and ensuring employees are trained accordingly. === Payment Card Industry Data Security Standard === The Payment Card Industry Security Standards Council, the governing council for stakeholders in the payment industry, formed by American Express, Discover, JCB International, MasterCard, and Visa that developed the DSS as a requirement for the payment industry. Requirement 12.6 requires member organizations to institute a formal security awareness program. There is a published guide for organizations to adhere to when setting up the program. === US States Training Regulations === Some States mandate Security Awareness Training whiles other do not but simply recommend voluntary training. Among states that require the training for its employees include: Colorado (The Colorado Information Security Act, Colorado Revised Statutes 24-37.5-401 et seq.) Connecticut (13 FAM 301.1-1 Cyber Security Awareness Training (PS800)) Florida (Florida Statutes Chapter 282) Georgia (Executive Order GA E.O.182 mandated training within 90 days of issue) Illinois (Cook County) Indiana (IN H 1240) Louisiana (Louisiana Division of Administration, Office of Technology Services p. 52: LA H 633) Maryland (20-07 IT Security Policy) Montana (Mandatory cyber training for executive branch state employees) Nebraska Nevada (agency-by-agency state employee requirement - State Security Standard 123 – IT Security) New Hampshire New Jersey ( NJ A 1654) North Carolina Ohio (IT-15 - Security Awareness and Training) Pennsylvania Texas Utah Vermont Virginia West Virginia (WV Code Section 5A-6-4a) == Training Techniques == Below are some common training techniques, even though some can be blended depending on the operating environment: Interactive video training – This technique allows users to be trained using two-way interactive audio and video instruction. Web-based training – This method allows employees or users to take the training independently and usually has a testing component to determine if learning has taken place. If not, users can be allowed to retake the course and test to ensure there is a complete understanding
Read more →
Semantic knowledge management

In computer science, semantic knowledge management is a set of practices that seeks to classify content so that the knowledge it contains may be immediately accessed and transformed for delivery to the desired audience, in the required format. This classification of content is semantic in its nature – identifying content by its type or meaning within the content itself and via external, descriptive metadata – and is achieved by employing XML technologies. The specific outcomes of these practices are: Maintain content for multiple audiences together in a single document Transform content into various delivery formats without re-authoring Search for content more effectively Involve more subject-matter experts in the creation of content without reducing quality Reduce production costs for delivery formats Reduce the manual administration of getting the right knowledge to the right people Reduce the cost and time to localize content == Notable semantic knowledge management systems == Learn eXact Thinking Cap LCMS Thinking Cap LMS Xyleme LCMS iMapping
Read more →
Catholic Church and artificial intelligence

The Catholic Church views artificial intelligence as a significant technological development that must be governed by strict ethical principles rooted in human dignity and the common good. In January 2025, the Church issued the doctrinal note Antiqua et nova co-issued by the Dicastery for the Doctrine of the Faith and the Dicastery for Culture and Education. It addresses the "relationship between artificial intelligence and human intelligence" and offers reflections on the "anthropological and ethical challenges raised by AI". In August 2025, Time magazine included Pope Leo XIV in its 2025 list of the World’s Most Influential People in Artificial Intelligence. In May 2026, Pope Leo XIV approved the creation of a new Vatican commission on artificial intelligence. He released his first papal encyclical, titled Magnifica humanitas, on the topic later in the month.
Read more →
Catastrophic interference

Catastrophic interference, also known as catastrophic forgetting, is the tendency of an artificial neural network to abruptly and drastically forget previously learned information upon learning new information. Neural networks are an important part of the connectionist approach to cognitive science. The issue of catastrophic interference when modeling human memory with connectionist models was originally brought to the attention of the scientific community by research from McCloskey and Cohen (1989), and Ratcliff (1990). It is a radical manifestation of the 'sensitivity-stability' dilemma or the 'stability-plasticity' dilemma. Specifically, these problems refer to the challenge of making an artificial neural network that is sensitive to, but not disrupted by, new information. Lookup tables and connectionist networks lie on the opposite sides of the stability plasticity spectrum. The former remains completely stable in the presence of new information but lacks the ability to generalize, i.e. infer general principles, from new inputs. On the other hand, connectionist networks like the standard backpropagation network can generalize to unseen inputs, but they are sensitive to new information. Backpropagation models can be analogized to human memory insofar as they have a similar ability to generalize, but these networks often exhibit less stability than human memory. Notably, these backpropagation networks are susceptible to catastrophic interference. This is an issue when modelling human memory, because unlike these networks, humans typically do not show catastrophic forgetting. == Discovery == The term catastrophic interference was originally coined by McCloskey and Cohen (1989) but was also brought to the attention of the scientific community by research from Ratcliff (1990). === The Sequential Learning Problem: McCloskey and Cohen (1989) === McCloskey and Cohen (1989) noted the problem of catastrophic interference during two different experiments with backpropagation neural network modelling. Experiment 1: Learning the ones and twos addition facts In their first experiment they trained a standard backpropagation neural network on a single training set consisting of 17 single-digit ones problems (i.e., 1 + 1 through 9 + 1, and 1 + 2 through 1 + 9) until the network could represent and respond properly to all of them. The error between the actual output and the desired output steadily declined across training sessions, which reflected that the network learned to represent the target outputs better across trials. Next, they trained the network on a single training set consisting of 17 single-digit twos problems (i.e., 2 + 1 through 2 + 9, and 1 + 2 through 9 + 2) until the network could represent, respond properly to all of them. They noted that their procedure was similar to how a child would learn their addition facts. Following each learning trial on the twos facts, the network was tested for its knowledge on both the ones and twos addition facts. Like the ones facts, the twos facts were readily learned by the network. However, McCloskey and Cohen noted the network was no longer able to properly answer the ones addition problems even after one learning trial of the twos addition problems. The output pattern produced in response to the ones facts often resembled an output pattern for an incorrect number more closely than the output pattern for a correct number. This is considered to be a drastic amount of error. Furthermore, the problems 2+1 and 1+2, which were included in both training sets, even showed dramatic disruption during the first learning trials of the twos facts. Experiment 2: Replication of Barnes and Underwood (1959) study In their second connectionist model, McCloskey and Cohen attempted to replicate the study on retroactive interference in humans by Barnes and Underwood (1959). They trained the model on A-B and A-C lists and used a context pattern in the input vector (input pattern), to differentiate between the lists. Specifically the network was trained to respond with the right B response when shown the A stimulus and A-B context pattern and to respond with the correct C response when shown the A stimulus and the A-C context pattern. When the model was trained concurrently on the A-B and A-C items then the network readily learned all of the associations correctly. In sequential training the A-B list was trained first, followed by the A-C list. After each presentation of the A-C list, performance was measured for both the A-B and A-C lists. They found that the amount of training on the A-C list in Barnes and Underwood study that lead to 50% correct responses, lead to nearly 0% correct responses by the backpropagation network. Furthermore, they found that the network tended to show responses that looked like the C response pattern when the network was prompted to give the B response pattern. This indicated that the A-C list apparently had overwritten the A-B list. This could be likened to learning the word dog, followed by learning the word stool and then finding that you think of the word stool when presented with the word dog. McCloskey and Cohen tried to reduce interference through a number of manipulations including changing the number of hidden units, changing the value of the learning rate parameter, overtraining on the A-B list, freezing certain connection weights, changing target values 0 and 1 instead 0.1 and 0.9. However, none of these manipulations satisfactorily reduced the catastrophic interference exhibited by the networks. Overall, McCloskey and Cohen (1989) concluded that: at least some interference will occur whenever new learning alters the weights involved in representing old learning the greater the amount of new learning, the greater the disruption in old knowledge interference was catastrophic in the backpropagation networks when learning was sequential but not concurrent === Constraints Imposed by Learning and Forgetting Functions: Ratcliff (1990) === Ratcliff (1990) used multiple sets of backpropagation models applied to standard recognition memory procedures, in which the items were sequentially learned. After inspecting the recognition performance models he found two major problems: Well-learned information was catastrophically forgotten as new information was learned in both small and large backpropagation networks. Even one learning trial with new information resulted in a significant loss of the old information, paralleling the findings of McCloskey and Cohen (1989). Ratcliff also found that the resulting outputs were often a blend of the previous input and the new input. In larger networks, items learned in groups (e.g. AB then CD) were more resistant to forgetting than were items learned singly (e.g. A then B then C...). However, the forgetting for items learned in groups was still large. Adding new hidden units to the network did not reduce interference. Discrimination between the studied items and previously unseen items decreased as the network learned more. This finding contradicts studies on human memory, which indicated that discrimination increases with learning. Ratcliff attempted to alleviate this problem by adding 'response nodes' that would selectively respond to old and new inputs. However, this method did not work as these response nodes would become active for all inputs. A model which used a context pattern also failed to increase discrimination between new and old items. == Proposed solutions == The main cause of catastrophic interference seems to be overlap in the representations at the hidden layer of distributed neural networks. In a distributed representation, each input tends to create changes in the weights of many of the nodes. Catastrophic forgetting occurs because when many of the weights where "knowledge is stored" are changed, it is unlikely for prior knowledge to be kept intact. During sequential learning, the inputs become mixed, with the new inputs being superimposed on top of the old ones. Another way to conceptualize this is by visualizing learning as a movement through a weight space. This weight space can be likened to a spatial representation of all of the possible combinations of weights that the network could possess. When a network first learns to represent a set of patterns, it finds a point in the weight space that allows it to recognize all of those patterns. However, when the network then learns a new set of patterns, it will move to a place in the weight space for which the only concern is the recognition of the new patterns. To recognize both sets of patterns, the network must find a place in the weight space suitable for recognizing both the new and the old patterns. Below are a number of techniques which have empirical support in successfully reducing catastrophic interference in backpropagation neural networks: === Orthogonality === Many of the early techniques in reducing representational overlap involved making either the input vecto
Read more →
Sanctuary (app)

Sanctuary is a mobile app focusing on astrology and mystical services. Users enter their birthday, time of birth, and place of birth information into the app and receive a birth chart as well as daily horoscope readings. Users can also sign up for a monthly membership and receive on-demand astrological readings via a text message format. The service has been described as being “Talkspace for astrology" and "Uber for astrological readings". The mobile app uses an A.I.-driven interface. On May 14, 2019, Apple featured Sanctuary as the App of the Day. == History == Sanctuary initially began as project within the incubator of Lorne Michaels’ Broadway Video Ventures. The app officially launched on March 21, 2019. Its backers include Broadway Video Ventures, Greycroft Partners, and Shari Redstone.
Read more →
Retrieval-based Voice Conversion

Retrieval-based Voice Conversion (RVC) is an open source voice conversion AI algorithm that enables realistic speech-to-speech transformations, accurately preserving the intonation and audio characteristics of the original speaker. == Overview == In contrast to text-to-speech systems such as ElevenLabs, RVC differs by providing speech-to-speech outputs instead. It maintains the modulation, timbre and vocal attributes of the original speaker, making it suitable for applications where emotional tone is crucial. The algorithm enables both pre-processed and real-time voice conversion with low latency. This real-time capability marks a significant advancement over previous AI voice conversion technologies, such as So-vits SVC. Its speed and accuracy have led many to note that its generated voices sound near-indistinguishable from "real life", provided that sufficient computational specifications and resources (e.g., a powerful GPU and ample RAM) are available when running it locally and that a high-quality voice model is used. == Technical foundation == Retrieval-based Voice Conversion (RVC) utilizes a hybrid approach that integrates feature extraction with retrieval-based synthesis. Instead of directly mapping source speaker features to the target speaker using statistical models, RVC retrieves relevant segments from a target speech database, aiming to enhance the naturalness and speaker fidelity of the converted speech. At a high level, the RVC system typically comprises three main components: (1) a content feature extractor, such as a phonetic posteriorgram (PPG) encoder or self-supervised models like HuBERT; (2) a vector retrieval module that searches a target voice database for the most similar speech units; and (3) a vocoder or neural decoder that synthesizes waveform output from the retrieved representations. The retrieval-based paradigm aims to mitigate the oversmoothing effect commonly observed in fully neural sequence-to-sequence models, potentially leading to more expressive and natural-sounding speech. Furthermore, with the incorporation of high-dimensional embeddings and k-nearest-neighbor search algorithms, the model can perform efficient matching across large-scale databases without significant computational overhead. Recent RVC frameworks have incorporated adversarial learning strategies and GAN-based vocoders, such as HiFi-GAN, to enhance synthesis quality. These integrations have been shown to produce clearer harmonics and reduce reconstruction errors. == Research developments == Research on RVC has recently explored the use of self-supervised learning (SSL) encoders such as wav2vec 2.0 and HuBERT to replace hand-engineered features like MFCCs. These encoders improve content preservation, especially when source and target speakers have dissimilar speaking styles or accents. Moreover, modern RVC models leverage vector quantization methods to discretize the acoustic space, improving both synthesis accuracy and generalization across unseen speakers. For example, retrieval-augmented VQ models can condition the synthesis stage on quantized speech tokens, which enhances controllability and style transfer. Despite its strengths, RVC still faces limitations related to database coverage, especially in real-time or few-shot settings. Inadequate diversity in the target voice corpus may lead to suboptimal retrieval or unnatural prosody. These advances demonstrate the viability of RVC as a strong alternative to conventional deep learning VC systems, balancing both flexibility and efficiency in diverse voice synthesis applications. == Training process == The training pipeline for retrieval-based voice conversion typically includes a preprocessing step where the target speaker's dataset is segmented and normalized. A pitch extractor such as librosa or DDSP-DDC may be used to obtain fundamental frequency (F0) features. During training, the model learns to map content features from the source speaker to the acoustic representation of the target speaker while maintaining pitch and prosody. The training objective often combines reconstruction loss with feature consistency loss across intermediate layers, and may incorporate cycle consistency loss to preserve speaker identity. Fine-tuning on small datasets is feasible due to the use of pre-trained models, particularly for the SSL encoder and content extractor components. This approach allows transfer learning to be applied effectively, enabling the model to converge faster and generalize better to unseen inputs. Most open implementations support batch training, gradient accumulation, and mixed-precision acceleration (e.g., FP16), especially when utilizing NVIDIA CUDA-enabled GPUs. == Real-time deployment == RVC systems can be deployed in real-time scenarios through WebUI interfaces and streaming audio frameworks. Optimizations include converting the inference graph to ONNX or TensorRT formats, reducing latency. Audio buffers are typically processed in chunks of 0.2–0.5 seconds to ensure minimal delay and seamless conversion. Cross-platform compatibility with tools such as OBS Studio and Voicemeeter enables integration into live streaming, video production, or virtual avatar environments. == Applications and concerns == The technology enables voice changing and mimicry, allowing users to create accurate models of others using only a negligible amount of minutes of clear audio samples. These voice models can be saved as .pth (PyTorch) files. While this capability facilitates numerous creative applications, it has also raised concerns about potential misuse as deepfake software for identity theft and malicious impersonation through voice calls. == Ethical and legal considerations == As with other deep generative models, the rise of RVC technology has led to increasing debate about copyright, consent, and authorship. While some jurisdictions may allow parody or fair use in creative contexts, impersonating living individuals without permission may infringe upon privacy and likeness rights. As a result, some platforms have begun issuing takedown notices against AI-generated voice content that closely mimics celebrities or musicians. === In pop culture === RVC inference has been used to create realistic depictions of song covers, such as replacing original vocals with characters like Twilight Sparkle and Mordecai to have them sing duets of popular music like "Airplanes" and "Somebody That I Used to Know." These AI-generated covers, which can sound strikingly similar to the voice imitated, have gained popularity on platforms like YouTube as humorous memes.
Read more →
Pinakes

The Pinakes (Ancient Greek: Πίνακες 'tables', plural of πίναξ pinax) is a lost bibliographic work composed by Callimachus (310/305–240 BCE) that is popularly considered to be the first library catalog in the West; its contents were based upon the holdings of the Library of Alexandria during Callimachus's tenure there during the third century BCE. == History == The Library of Alexandria had been founded by Ptolemy I Soter about 306 BCE. The first recorded librarian was Zenodotus of Ephesus. During Zenodotus' tenure, Callimachus, who was never the head librarian, compiled many catalogues/lists, each called Pinakes. His most famous one listed authors and their works; thus he became the first known bibliographer and the scholar who organized the library by authors and subjects about 245 BCE. His work was 120 volumes long. Apollonius of Rhodes was the successor to Zenodotus. Eratosthenes of Cyrene succeeded Apollonius in 235 BCE and compiled his tetagmenos epi teis megaleis bibliothekeis, the 'scheme of the great bookshelves'. In 195 BCE Aristophanes of Byzantium, Eratosthenes' successor, was the librarian and updated the Pinakes, although it is also possible that his work was not a supplement of Callimachus' Pinakes themselves, but an independent polemic against, or commentary upon, their contents. == Description == The collection at the Library of Alexandria contained nearly 500,000 papyrus scrolls, which were grouped together by subject matter and stored in bins. Each bin carried a label with painted tablets hung above the stored papyri. Pinakes was named after these tablets and are a set of index lists. The bins gave bibliographical information for every roll. A typical entry started with a title and also provided the author's name, birthplace, father's name, any teachers trained under, and educational background. It contained a brief biography of the author and a list of the author's publications. The entry had the first line of the work, a summary of its contents, the name of the author, and information about the origin of the roll, as well as any doubts about the genuineness of the ascription. Callimachus' system divided works into six genres of poetry and five sections of prose: rhetoric, law, epic, tragedy, comedy, lyric poetry, history, medicine, mathematics, natural science, and miscellanies. Each category was alphabetized by author. Callimachus composed two other works that were referred as pinakes and were probably somewhat similar in format to the Pinakes (of which they "may or may not be subsections"), but were concerned with individual topics. These are listed by the Suda as: A Chronological Pinax and Description of Didaskaloi from the Beginning and Pinax of the Vocabulary and Treatises of Democritus. == Later bibliographic pinakes == The term pinax was used for bibliographic catalogs beyond Callimachus. For example, Ptolemy-el-Garib's catalog of Aristotle's writings comes to us with the title Pinax (catalog) of Aristotle's writings. == Legacy == The Pinakes proved indispensable to librarians for centuries, and they became a model for organizing knowledge throughout the Mediterranean. Their later influence can be traced to medieval times, even to the Arabic counterpart of the tenth century: Ibn al-Nadim's Al-Fihrist ("Index"). Local variations for cataloging and library classification continued through the late 19th century, when Anthony Panizzi and Melvil Dewey paved the way for more shared and standardized approaches.
Read more →
Logic Theorist

Logic Theorist is a computer program completed in 1956 by Allen Newell, Herbert A. Simon, and Cliff Shaw. It was the first program deliberately engineered to perform automated reasoning, and has been described as "the first artificial intelligence program". Logic Theorist proved 38 of the first 52 theorems in chapter two of Whitehead and Bertrand Russell's Principia Mathematica, and found new and shorter proofs for some of them. == History == In 1955, when Newell and Simon began to work on the Logic Theorist, the field of artificial intelligence did not yet exist; the term "artificial intelligence" would not be coined until the following summer. Simon was a political scientist who had previously studied the way bureaucracies function as well as developing his theory of bounded rationality (for which he would later win the Nobel Memorial Prize in Economic Sciences in 1978). He believed the study of business organizations requires, like artificial intelligence, an insight into the nature of human problem solving and decision making. Simon has stated that when consulting at RAND Corporation in the early 1950s, he saw a printer typing out a map, using ordinary letters and punctuation as symbols. This led him to think that a machine that could manipulate symbols could simulate decision making and possibly even the process of human thought. The program that printed the map had been written by Newell, a RAND scientist studying logistics and organization theory. For Newell, the decisive moment was in 1954 when Oliver Selfridge came to RAND to describe his work on pattern matching. Watching the presentation, Newell suddenly understood how the interaction of simple, programmable units could accomplish complex behavior, including the intelligent behavior of human beings. "It all happened in one afternoon," he would later say. It was a rare moment of scientific epiphany. "I had such a sense of clarity that this was a new path, and one I was going to go down. I haven't had that sensation very many times. I'm pretty skeptical, and so I don't normally go off on a toot, but I did on that one. Completely absorbed in it—without existing with the two or three levels consciousness so that you're working, and aware that you're working, and aware of the consequences and implications, the normal mode of thought. No. Completely absorbed for ten to twelve hours." Newell and Simon began to talk about the possibility of teaching machines to think. Their first project was a program that could prove mathematical theorems like the ones used in Bertrand Russell and Alfred North Whitehead's Principia Mathematica. They enlisted the help of computer programmer Cliff Shaw, also from RAND, to develop the program. (Newell says "Cliff was the genuine computer scientist of the three".) The first version was hand-simulated: they wrote the program onto 3x5 cards and, as Simon recalled:In January 1956, we assembled my wife and three children together with some graduate students. To each member of the group, we gave one of the cards, so that each one became, in effect, a component of the computer program ... Here was nature imitating art imitating nature. They succeeded in showing that the program could successfully prove theorems as well as a talented mathematician. Eventually Shaw was able to run the program on the computer at RAND's Santa Monica facility. In the summer of 1956, John McCarthy, Marvin Minsky, Claude Shannon and Nathan Rochester organized a conference on the subject of what they called "artificial intelligence" (a term coined by McCarthy for the occasion). Newell and Simon proudly presented the group with the Logic Theorist. It was met with a lukewarm reception. Pamela McCorduck writes "the evidence is that nobody save Newell and Simon themselves sensed the long-range significance of what they were doing." Simon confides that "we were probably fairly arrogant about it all" and adds: They didn't want to hear from us, and we sure didn't want to hear from them: we had something to show them! ... In a way it was ironic because we already had done the first example of what they were after; and second, they didn't pay much attention to it. Logic Theorist soon proved 38 of the first 52 theorems in chapter 2 of the Principia Mathematica. The proof of theorem 2.85 was actually more elegant than the proof produced laboriously by hand by Russell and Whitehead (2026-03-20: What is called here Theorem 2.85 is, in fact, numbered as 2.53 in the page 107 of the 1963 Cambridge University Press edition (https://www.uhu.es/francisco.moreno/gii_mac/docs/Principia_Mathematica_vol1.pdf) and which appears, under the same 2.53 number, on page 112 of the 1910 CUP Edition, according to the digitalization on wikibooks (https://en.wikisource.org/wiki/Russell_%26_Whitehead%27s_Principia_Mathematica/Part_1/Section_A#Discussion_2)). Simon was able to show the new proof to Russell himself who "responded with delight". They attempted to publish the new proof in The Journal of Symbolic Logic, but it was rejected on the grounds that a new proof of an elementary mathematical theorem was not notable, apparently overlooking the fact that one of the authors was a computer program. Newell and Simon formed a lasting partnership, founding one of the first AI laboratories at the Carnegie Institute of Technology and developing a series of influential artificial intelligence programs and ideas, including the General Problem Solver, Soar, and their unified theory of cognition. == Architecture == The Logic Theorist is a program that performs logical processes on logical expressions. The Logic Theorist operates on the following principles: === Expressions === An expression is made of elements. There are two kinds of memories: working and storage. Each working memory contains a single element. The Logic Theorist usually uses 1 to 3 working memories. Each storage memory is a list representing a full expression or a set of elements. In particular, it contains all the axioms and proven logical theorems. An expression is an abstract syntax tree, each node being an element with up to 11 attributes. For example, the logical expression ¬ P → ( Q ∧ ¬ P ) {\displaystyle \neg P\to (Q\wedge \neg P)} is represented as a tree with a root element representing → {\displaystyle \to } . Among the attributes of the root element are pointers to the two elements representing the subexpressions ¬ P {\displaystyle \neg P} and Q ∧ ¬ P {\displaystyle Q\wedge \neg P} . === Processes === There are four kinds of processes, from the lowest to the highest level. Instruction: These are similar to assembly code. They may either perform a primitive operation on an expression in working memory, or perform a conditional jump to another instruction. An example is "put the right sub-element of working-memory 1 to working-memory 2" Elementary process: These are similar to subroutines. A sequence of instructions that can be called. Method: A sequence of elementary processes. There are 4 methods: substitution: given an expression, it attempts to transform it to a proven theorem or axiom by substitutions of variables and logical connectives. detachment: given expression B {\displaystyle B} , it attempts to find a proven theorem or axiom of form A → B ′ {\displaystyle A\to B'} , where B ′ {\displaystyle B'} yields B {\displaystyle B} after substitution, then attempts to prove A {\displaystyle A} by substitution. chaining forward: given expression A → C {\displaystyle A\to C} , it attempts to find for a proven theorem or axiom of form A → B {\displaystyle A\to B} , then attempt to prove B → C {\displaystyle B\to C} by substitution. chaining backward: given expression A → C {\displaystyle A\to C} , it attempts to find for a proven theorem or axiom of form B → C {\displaystyle B\to C} , then attempt to prove A → B {\displaystyle A\to B} by substitution. executive control method: This method applies each of the 4 methods in sequence to each theorem to be proved. == Logic Theorist's influence on AI == Logic Theorist introduced several concepts that would be central to AI research: Reasoning as search Logic Theorist explored a search tree: the root was the initial hypothesis, each branch was a deduction based on the rules of logic. Somewhere in the tree was the goal: the proposition the program intended to prove. The pathway along the branches that led to the goal was a proof – a series of statements, each deduced using the rules of logic, that led from the hypothesis to the proposition to be proved. Heuristics Newell and Simon realized that the search tree would grow exponentially and that they needed to "trim" some branches, using "rules of thumb" to determine which pathways were unlikely to lead to a solution. They called these ad hoc rules "heuristics", using a term introduced by George Pólya in his classic book on mathematical proof, How to Solve It. (Newell had taken courses from Pólya at Stanford). Heuristics would become an important area o
Read more →
ViBe

ViBe is a background subtraction algorithm which has been presented at the IEEE ICASSP 2009 conference and was refined in later publications. More precisely, it is a software module for extracting background information from moving images. It has been developed by Oliver Barnich and Marc Van Droogenbroeck of the Montefiore Institute, University of Liège, Belgium. ViBe is patented: the patent covers various aspects such as stochastic replacement, spatial diffusion, and non-chronological handling. ViBe is written in the programming language C, and has been implemented on CPU, GPU and FPGA. == Technical description == Source: === Pixel model and classification process === Many advanced techniques are used to provide an estimate of the temporal probability density function (pdf) of a pixel x. ViBe's approach is different, as it imposes the influence of a value in the polychromatic space to be limited to the local neighborhood. In practice, ViBe does not estimate the pdf, but uses a set of previously observed sample values as a pixel model. To classify a value pt(x), it is compared to its closest values among the set of samples. === Model update: Sample values lifespan policy === ViBe ensures a smooth exponentially decaying lifespan for the sample values that constitute the pixel models. This makes ViBe able to successfully deal with concomitant events with a single model of a reasonable size for each pixel. This is achieved by choosing, randomly, which sample to replace when updating a pixel model. Once the sample to be discarded has been chosen, the new value replaces the discarded sample. The pixel model that would result from the update of a given pixel model with a given pixel sample cannot be predicted since the value to be discarded is chosen at random. === Model update: Spatial Consistency === To ensure the spatial consistency of the whole image model and handle practical situations such as small camera movements or slowly evolving background objects, ViBe uses a technique similar to that developed for the updating process in which it chooses at random and update a pixel model in the neighborhood of the current pixel. By denoting NG(x) and p(x) respectively the spatial neighborhood of a pixel x and its value, and assuming that it was decided to update the set of samples of x by inserting p(x), then ViBe also use this value p(x) to update the set of samples of one of the pixels in the neighborhood NG(x), chosen at random. As a result, ViBe is able to produce spatially coherent results directly without the use of any post-processing method. === Model initialization === Although the model could easily recover from any type of initialization, for example by choosing a set of random values, it is convenient to get an accurate background estimate as soon as possible. Ideally a segmentation algorithm would like to be able to segment the video sequences starting from the second frame, the first frame being used to initialize the model. Since no temporal information is available prior to the second frame, ViBe populates the pixel models with values found in the spatial neighborhood of each pixel; more precisely, it initializes the background model with values taken randomly in each pixel neighborhood of the first frame. The background estimate is therefore valid starting from the second frame of a video sequence.
Read more →
OpenAI Five

OpenAI Five is a computer program by OpenAI that plays the five-on-five video game Dota 2. Its first public appearance occurred in 2017, where it was demonstrated in a live one-on-one game against the professional player Dendi, who lost to it. The following year, the system had advanced to the point of performing as a full team of five, and began playing against and showing the capability to defeat professional teams. By choosing a game as complex as Dota 2 to study machine learning, OpenAI thought they could more accurately capture the unpredictability and continuity seen in the real world, thus constructing more general problem-solving systems. The algorithms and code used by OpenAI Five were eventually borrowed by another neural network in development by the company, one which controlled a physical robotic hand. OpenAI Five has been compared to other similar cases of artificial intelligence (AI) playing against and defeating humans, such as AlphaStar in the video game StarCraft II, AlphaGo in the board game Go, Deep Blue in chess, and Watson on the television game show Jeopardy!. == History == Development on the algorithms used for the bots began in November 2016. OpenAI decided to use Dota 2, a competitive five-on-five video game, as a base due to it being popular on the live streaming platform Twitch, having native support for Linux, and had an application programming interface (API) available. Before becoming a team of five, the first public demonstration occurred at The International 2017 in August, the annual premiere championship tournament for the game, where Dendi, a Ukrainian professional player, lost against an OpenAI bot in a live one-on-one matchup. After the match, CTO Greg Brockman explained that the bot had learned by playing against itself for two weeks of real time, and that the learning software was a step in the direction of creating software that can handle complex tasks "like being a surgeon". OpenAI used a methodology called reinforcement learning, as the bots learn over time by playing against itself hundreds of times a day for months, in which they are rewarded for actions such as killing an enemy and destroying towers. By June 2018, the ability of the bots expanded to play together as a full team of five and were able to defeat teams of amateur and semi-professional players. At The International 2018, OpenAI Five played in two games against professional teams, one against the Brazilian-based paiN Gaming and the other against an all-star team of former Chinese players. Although the bots lost both matches, OpenAI still considered it a successful venture, stating that playing against some of the best players in Dota 2 allowed them to analyze and adjust their algorithms for future games. The bots' final public demonstration occurred in April 2019, where they won a best-of-three series against The International 2018 champions OG at a live event in San Francisco. A four-day online event to play against the bots, open to the public, occurred the same month. There, the bots played in 42,729 public games, winning 99.4% of those games. == Architecture == Each OpenAI Five bot is a neural network containing a single layer with a 4096-unit LSTM that observes the current game state extracted from the Dota developer's API. The neural network conducts actions via numerous possible action heads (no human data involved), and every head has meaning. For instance, the number of ticks to delay an action, what action to select – the X or Y coordinate of this action in a grid around the unit. In addition, action heads are computed independently. The AI system observes the world as a list of 20,000 numbers and takes an action by conducting a list of eight enumeration values. Also, it selects different actions and targets to understand how to encode every action and observe the world. OpenAI Five has been developed as a general-purpose reinforcement learning training system on the "Rapid" infrastructure. Rapid consists of two layers: it spins up thousands of machines and helps them 'talk' to each other and a second layer runs software. By 2018, OpenAI Five had played around 180 years worth of games in reinforcement learning running on 256 GPUs and 128,000 CPU cores, using Proximal Policy Optimization, a policy gradient method. == Comparisons with other game AI systems == Prior to OpenAI Five, other AI versus human experiments and systems have been successfully used before, such as Jeopardy! with Watson, chess with Deep Blue, and Go with AlphaGo. In comparison with other games that have used AI systems to play against human players, Dota 2 differs as explained below: Long run view: The bots run at 30 frames per second for an average match time of 45 minutes, which results in 80,000 ticks per game. OpenAI Five observes every fourth frame, generating 20,000 moves. By comparison, chess usually ends before 40 moves, while Go ends before 150 moves. Partially observed state of the game: Players and their allies can only see the map directly around them. The rest of it is covered in a fog of war which hides enemies units and their movements. Thus, playing Dota 2 requires making inferences based on this incomplete data, as well as predicting what their opponent could be doing at the same time. By comparison, Chess and Go are "full-information games", as they do not hide elements from the opposing player. Continuous action space: Each playable character in a Dota 2 game, known as a hero, can take dozens of actions that target either another unit or a position. The OpenAI Five developers allow the space into 170,000 possible actions per hero. Without counting the perpetual aspects of the game, there are an average of ~1,000 valid actions each tick. By comparison, the average number of actions in chess is 35 and 250 in Go. Continuous observation space: Dota 2 is played on a large map with ten heroes, five on each team, along with dozens of buildings and non-player character (NPC) units. The OpenAI system observes the state of a game through developers' bot API, as 20,000 numbers that constitute all information a human is allowed to get access to. A chess board is represented as about 70 lists, whereas a Go board has about 400 enumerations. == Reception == OpenAI Five have received acknowledgement from the AI, tech, and video game community at large. Microsoft founder Bill Gates called it a "big deal", as their victories "required teamwork and collaboration". Chess champion Garry Kasparov, who lost against the Deep Blue AI in 1997, stated that despite their losing performance at The International 2018, the bots would eventually "get there, and sooner than expected". In a conversation with MIT Technology Review, AI experts also considered OpenAI Five system as a significant achievement, as they noted that Dota 2 was an "extremely complicated game", so even beating non-professional players was impressive. PC Gamer wrote that their wins against professional players was a significant event in machine learning. In contrast, Motherboard wrote that the victory was "basically cheating" due to the simplified hero pools on both sides, as well as the fact that bots were given direct access to the API, as opposed to using computer vision to interpret pixels on the screen. The Verge wrote that the bots were evidence that the company's approach to reinforcement learning and its general philosophy about AI was "yielding milestones". In 2019, DeepMind unveiled a similar bot for StarCraft II, AlphaStar. Like OpenAI Five, AlphaStar used reinforcement learning and self-play. The Verge reported that "the goal with this type of AI research is not just to crush humans in various games just to prove it can be done. Instead, it's to prove that — with enough time, effort, and resources — sophisticated AI software can best humans at virtually any competitive cognitive challenge, be it a board game or a modern video game." They added that the DeepMind and OpenAI victories were also a testament to the power of certain uses of reinforcement learning. It was OpenAI's hope that the technology could have applications outside of the digital realm. In 2018, they were able to reuse the same reinforcement learning algorithms and training code from OpenAI Five for Dactyl, a human-like robot hand with a neural network built to manipulate physical objects. In 2019, Dactyl solved the Rubik's Cube.
Read more →
RevoScaleR

RevoScaleR is a machine learning package in R created by Microsoft. It is available as part of Machine Learning Server, Microsoft R Client, and Machine Learning Services in Microsoft SQL Server 2016. The package contains functions for creating linear model, logistic regression, random forest, decision tree and boosted decision tree, and K-means, in addition to some summary functions for inspecting and visualizing data. It has a Python package counterpart called revoscalepy. Another closely related package is MicrosoftML, which contains machine learning algorithms that RevoScaleR does not have, such as neural network and SVM. In June 2021, Microsoft announced to open source the RevoScaleR and revoscalepy packages, making them freely available under the MIT License. == Concepts == Many R packages are designed to analyze data that can fit in the memory of the machine and usually do not make use of parallel processing. RevoScaleR was designed to address these limitations. The functions in RevoScaleR orientate around three main abstraction concepts that users can specify to process large amount of data that might not fit in memory and exploit parallel resources to speed up the analysis. === Compute Contexts === A compute context refers to the location where the computation on the data happens. It could be "local" (on the client machine) or "remote" (on a data platform such as a SQL server, or Spark). Pushing the computation to a remote server allows people to take advantage of the greater compute resources that a remote machine may have. If the data being analyzed reside on the same machine, using a remote compute context also removes the need to pull data across the network onto the client machine. === Data source === Data source defines where the data comes from. There are various data sources available in RevoScaleR, such as text data, Xdf data, in-SQL data, and a spark dataframe. People can wrap their data in a data source object and use that as run analytics in different compute context. Different data sources are available in different compute context. For example, if the compute context is set to SQL server, then the only data source one can use would be an in-SQL data source. === Analytics === Analytic functions in RevoScaleR takes in data source object, a compute context, and the other parameters needed to build the specific model, such as formula for the logistic regression or the number of trees in a decision tree. In addition to those parameters, one can also specify the level of parallelism, such as the size of the data chunk for each process or number of processes to build the model. However, parallelism is only available in non-express edition. == Limitations == The package is mostly meant to be used with a SQL server or other remote machines. To fully leverage the abstractions it uses to process a large dataset, one needs a remote server and non-Express free edition of the package. It cannot be easily installed such as by running "install.packages("RevoScaleR")" like most open source R packages. It's available only through Microsoft R Client, a distribution of R for data science, or Microsoft Machine Learning Server (stand-alone with no SQL server attached), or Microsoft Machine Learning Services (a SQL server services). However, one can still use the analytics functions in an Express, free version of the package.
Read more →