AI Art For Sale

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

• 

  WeChat

  WeChat or Weixin in Chinese (Chinese: 微信; pinyin: Wēixìn ; lit. 'micro-message') is an instant messaging, social media, and mobile payment app developed by Tencent. First released in 2011, it became the world's largest standalone mobile app in 2018 with over 1 billion monthly active users. The Chinese version of WeChat, Weixin, has been described as China's "app for everything" and a super-app because of its wide range of functions. WeChat provides text messaging, hold-to-talk voice messaging, broadcast (one-to-many) messaging, video conferencing, video games, mobile payment, sharing of photographs and videos and location sharing. It has been described as having "an almost indispensable part of life in China". Accounts registered using Chinese phone numbers are managed under the Weixin brand, and their data is stored in mainland China and subject to Weixin's terms of service and privacy policy. Non-Chinese numbers are registered under WeChat, and WeChat users are subject to a more liberal terms of service and better privacy policy, and their data is stored in the Netherlands for users in the European Union, and in Singapore for other users. User activity on Weixin, the Chinese version of the app, is analyzed, tracked and shared with Chinese authorities upon request as part of the mass surveillance network in China. Chinese-registered Weixin accounts censor politically sensitive topics, and the software license agreement for Weixin (but not WeChat) explicitly forbids content which "[en]danger[s] national security, divulge[s] state secrets, subvert[s] state power and undermine[s] national unity", as well as other types of content such as content that "[u]ndermine[s] national religious policies" and content that is "[i]nciting illegal assembly, association, procession, demonstrations and gatherings disrupting the social order". Due to its central part of Chinese life, a Chinese person having their WeChat account banned can cause a significant disruption to their life. Any interactions between Weixin and WeChat users are subject to the terms of service and privacy policies of both services. == History == By 2010, Tencent had already attained a massive user base with their desktop messenger app QQ. Recognizing smart phones were likely to disrupt this status quo, CEO Pony Ma sought to proactively invest in alternatives to their own QQ messenger app. WeChat began as a project at Tencent Guangzhou Research and Project center in October 2010. The original version of the app was created by Allen Zhang, named "Weixin" (微信) by Pony Ma, and launched in 2011. The user adoption of WeChat was initially very slow, with users wondering why key features were missing; however, after the release of the Walkie-talkie-like voice messaging feature in May of that year, growth surged. By 2012, when the number of users reached 100 million, Weixin was re-branded "WeChat" by President Martin Lau for the international market. During a period of government support of e-commerce development—for example in the 12th five-year plan (2011–2015)—WeChat also saw new features enabling payments and commerce in 2013, which saw massive adoption after their virtual Red envelope promotion for Chinese New Year 2014. WeChat had over 889 million monthly active users by 2016, and as of 2019 WeChat's monthly active users had risen to an estimate of one billion. As of January 2022, it was reported that WeChat has more than 1.2 billion users. After the launch of WeChat payment in 2013, its users reached 400 million the next year, 90 percent of whom were in China. By comparison, Facebook Messenger and WhatsApp had about one billion monthly active users in 2016 but did not offer most of the other services available on WeChat. For example, in Q2 2017, WeChat's revenues from social media advertising were about US$0.9 billion (RMB6 billion) compared with Facebook's total revenues of US$9.3 billion, 98% of which were from social media advertising. WeChat's revenues from its value-added services were US$5.5 billion. By 2018, WeChat had been used by 93.5% of Chinese internet users. In that year, it became the world's largest standalone mobile app in 2018 with over 1 billion monthly active users. In response to a border dispute between India and China, WeChat was banned in India in June 2020 along with several other Chinese apps, including TikTok. U.S. president Donald Trump sought to ban U.S. "transactions" with WeChat through an executive order but was blocked by a preliminary injunction issued in the United States District Court for the Northern District of California in September 2020. Joe Biden officially dropped Trump's efforts to ban WeChat in the U.S. in June 2021. == Features == WeChat, has been described as China's "app for everything" and a super-app because of its wide range of functions. WeChat provides text messaging, hold-to-talk voice messaging, broadcast (one-to-many) messaging, video conferencing, video games, mobile payment, sharing of photographs and videos and location sharing. It has been described as having "an almost indispensable part of life in China". Due to its central part of Chinese life, a Chinese person having their WeChat account banned can cause a significant disruption to their life. === Messaging === WeChat provides a variety of features including text messaging, hold-to-talk voice messaging, broadcast (one-to-many) messaging, video calls and conferencing, video games, photograph and video sharing, as well as location sharing. WeChat also allows users to exchange contacts with people nearby via Bluetooth, as well as providing various features for contacting people at random if desired (if people are open to it). It can also integrate with other social networking services such as Facebook and Tencent QQ. Photographs may also be embellished with filters and captions, and automatic translation service is available and could also translate the conversation during messaging. WeChat supports different instant messaging methods, including text messages, voice messages, walkie talkie, and stickers. Users can send previously saved or live pictures and videos, profiles of other users, coupons, lucky money packages, or current GPS locations with friends either individually or in a group chat. WeChat also provides a message recall feature to allow users to recall and withdraw information (e.g. images, documents) that are sent within 2 minutes in a conversation. WeChat also provides a voice-to-text feature that brings convenience when it is not convenient to listen to voice messages, as well as the basic ability to recognize emojis based on different tones of voice. A distance sensing feature is implemented in WeChat. It has the ability to activate the receivers' hold-to-talk function when the phone was brought in close proximity to the ear. After the receiver was held at a certain distance from the ear, the sensor would then proceed to automatically disable the phone speakers. This feature eliminates the risk of the user's voice messages being inadvertently broadcast to the general public. === Public accounts === WeChat users can register as a public account (公众号), which enables them to push feeds to subscribers, interact with subscribers, and provide subscribers with services. Users can also create an official account, which fall under service, subscription, or enterprise accounts. Once users as individuals or organizations set up a type of account, they cannot change it to another type. By the end of 2014, the number of WeChat official accounts had reached 8 million. Official accounts of organizations can apply to be verified (cost 300 RMB or about US$45). Official accounts can be used as a platform for services such as hospital pre-registrations, or credit card service. To create an official account, the applicant must register with Chinese authorities, which discourages "foreign companies". In April 2022, WeChat announced that it will start displaying the location of users in China every time they post on a public account. Meanwhile, overseas users on public accounts will also display the country based on their IP address. === Moments === "Moments" (朋友圈) is WeChat's brand name for its social feed of friends' updates. "Moments" is an interactive platform that allows users to post images, text, and short videos taken by users. It also allows users to share articles and music (associated with QQ Music or other web-based music services). Friends in the contact list can like the content and leave comments, functioning similarly to a private social network. In 2017 WeChat had a policy of a maximum of two advertisements per day per Moments user. Privacy in WeChat works by groups of friends: only the friends from the user's contact are able to view their Moments' contents and comments. The friends of the user will only be able to see the likes and comments from other users only if they are in a mutual friend group. For example, friends from high school are not able to

  Read more →
• 

  Data-centric AI

  Data-centric AI is an approach within artificial intelligence that emphasizes on improving the quality, consistency and representativeness of the data used to train machine learning models, rather than focusing primarily on optimizing model architectures or algorithms. This idea has gained traction as researchers and practitioners have come to believe that many performance limitations of machine learning systems stem from issues such as noisy labels, biased datasets, and lack of coverage in the data. Data-centric AI involves disciplined approach to data cleaning, augmentation, labeling, and governance that improves model performance and reliability in applications such as computer vision, natural language processing, and further.

  Read more →
• 

  Self-supervised learning

  Self-supervised learning (SSL) is a paradigm in machine learning where a model is trained on a task using the data itself to generate supervisory signals, rather than relying on externally-provided labels. In the context of neural networks, self-supervised learning aims to leverage inherent structures or relationships within the input data to create meaningful training signals. SSL tasks are designed so that solving them requires capturing essential features or relationships in the data. The input data is typically augmented or transformed in a way that creates pairs of related samples, where one sample serves as the input, and the other is used to formulate the supervisory signal. This augmentation can involve introducing noise, cropping, rotation, or other transformations. Self-supervised learning more closely imitates the way humans learn to classify objects. During SSL, the model learns in two steps. First, the task is solved based on an auxiliary or pretext classification task using pseudo-labels, which help to initialize the model parameters. Next, the actual task is performed with supervised or unsupervised learning. Self-supervised learning has produced promising results in recent years, and has found practical application in fields such as audio processing, and is being used by Facebook and others for speech recognition. == Pseudo-labels == Pseudo-labels are automatically generated labels that a model assigns to unlabeled data based on its own predictions. They are widely used in self-supervised and semi-supervised learning, where ground-truth annotations are limited or unavailable. By treating predicted labels as surrogate ground truth, learning algorithms can make use of large quantities of unlabeled data in the training process. Pseudo-labeling also plays an important role in systems that must adapt to concept drift, where the statistical properties of the data change over time. In these scenarios, the model may detect that an incoming instance deviates from previously learned behavior. The system then generates a classification result for that instance, and this predicted class is used as a pseudo-label for updating or retraining model components that are becoming outdated. This approach enables continuous adaptation in dynamic environments without requiring manual annotation. In many adaptive learning pipelines, pseudo-labels are chosen when the classifier produces sufficiently confident predictions, reducing the risk of propagating errors. These pseudo-labeled instances are then incorporated into training to refresh or evolve the model's understanding of emerging data patterns, particularly when existing components show signs of “aging” due to drift or distributional shifts. This strategy reduces reliance on manual labeling while helping maintain long-term model performance. == Types == === Autoassociative self-supervised learning === Autoassociative self-supervised learning is a specific category of self-supervised learning where a neural network is trained to reproduce or reconstruct its own input data. In other words, the model is tasked with learning a representation of the data that captures its essential features or structure, allowing it to regenerate the original input. The term "autoassociative" comes from the fact that the model is essentially associating the input data with itself. This is often achieved using autoencoders, which are a type of neural network architecture used for representation learning. Autoencoders consist of an encoder network that maps the input data to a lower-dimensional representation (latent space), and a decoder network that reconstructs the input from this representation. The training process involves presenting the model with input data and requiring it to reconstruct the same data as closely as possible. The loss function used during training typically penalizes the difference between the original input and the reconstructed output (e.g. mean squared error). By minimizing this reconstruction error, the autoencoder learns a meaningful representation of the data in its latent space. === Contrastive self-supervised learning === For a binary classification task, training data can be divided into positive examples and negative examples. Positive examples are those that match the target. For example, if training a classifier to identify birds, the positive training data would include images that contain birds. Negative examples would be images that do not. Contrastive self-supervised learning uses both positive and negative examples. The loss function in contrastive learning is used to minimize the distance between positive sample pairs, while maximizing the distance between negative sample pairs. An early example uses a pair of 1-dimensional convolutional neural networks to process a pair of images and maximize their agreement. Contrastive Language-Image Pre-training (CLIP) allows joint pretraining of a text encoder and an image encoder, such that a matching image-text pair have image encoding vector and text encoding vector that span a small angle (having a large cosine similarity). InfoNCE (Noise-Contrastive Estimation) is a method to optimize two models jointly, based on Noise Contrastive Estimation (NCE). Given a set X = { x 1 , … x N } {\displaystyle X=\left\{x_{1},\ldots x_{N}\right\}} of N {\displaystyle N} random samples containing one positive sample from p ( x t + k ∣ c t ) {\displaystyle p\left(x_{t+k}\mid c_{t}\right)} and N − 1 {\displaystyle N-1} negative samples from the 'proposal' distribution p ( x t + k ) {\displaystyle p\left(x_{t+k}\right)} , it minimizes the following loss function: L N = − E X [ log ⁡ f k ( x t + k , c t ) ∑ x j ∈ X f k ( x j , c t ) ] {\displaystyle {\mathcal {L}}_{\mathrm {N} }=-\mathbb {E} _{X}\left[\log {\frac {f_{k}\left(x_{t+k},c_{t}\right)}{\sum _{x_{j}\in X}f_{k}\left(x_{j},c_{t}\right)}}\right]} === Non-contrastive self-supervised learning === Non-contrastive self-supervised learning (NCSSL) uses only positive examples. Counterintuitively, NCSSL converges on a useful local minimum rather than reaching a trivial solution, with zero loss. For the example of binary classification, it would trivially learn to classify each example as positive. Effective NCSSL requires an extra predictor on the online side that does not back-propagate on the target side. === Joint-Embedding and Predictive Architectures === A major class of self-supervised learning moves beyond contrastive pairs, instead maximizing the agreement between views while preventing collapse through statistical constraints. Rooted in Deep Canonical Correlation Analysis (Deep CCA), this approach includes Joint-Embedding Architectures (JEA) like Barlow Twins and VICReg, which enforce covariance constraints to learn invariant representations without negative sampling. Deep Latent Variable Path Modelling (DLVPM) generalizes this to multimodal systems, using path models to enforce correlation and orthogonality across diverse data types. In 2022 Yann LeCun introduced Joint-Embedding Predictive Architectures (JEPA) as a step towards decision making, reasoning, and autonomous human intelligence in machines, including self-improvement through autonomous learning. Founded in representation learning, LeCun included the concept of a “world model” in JEPA which aims to enable machines to replicate human intellect by providing machines with a concept for the world in which they exist. Unlike autoencoders, JEPAs operate entirely in latent space, avoiding pixel-level noise to focus on semantic structure. Rather than just learning invariance, JEPAs learn by predicting masked latent representations from visible context. JEPA has been applied to domains such as image analysis, audio processing, and motion in images and video. == Comparison with other forms of machine learning == SSL belongs to supervised learning methods insofar as the goal is to generate a classified output from the input. At the same time, however, it does not require the explicit use of labeled input-output pairs. Instead, correlations, metadata embedded in the data, or domain knowledge present in the input are implicitly and autonomously extracted from the data. These supervisory signals, extracted from the data, can then be used for training. SSL is similar to unsupervised learning in that it does not require labels in the sample data. Unlike unsupervised learning, however, learning is not done using inherent data structures. Semi-supervised learning combines supervised and unsupervised learning, requiring only a small portion of the learning data be labeled. In transfer learning, a model designed for one task is reused on a different task. Training an autoencoder intrinsically constitutes a self-supervised process, because the output pattern needs to become an optimal reconstruction of the input pattern itself. However, in current jargon, the term 'self-supervised' often refers to tasks based on a pretext-task training setup

  Read more →
• 

  Web intelligence

  Web intelligence is the area of scientific research and development that explores the roles and makes use of artificial intelligence and information technology for new products, services and frameworks that are empowered by the World Wide Web. The term was coined in a paper written by Ning Zhong, Jiming Liu Yao and Y.Y. Ohsuga in the Computer Software and Applications Conference in 2000. == Research == The research about the web intelligence covers many fields – including data mining (in particular web mining), information retrieval, pattern recognition, predictive analytics, the semantic web, web data warehousing – typically with a focus on web personalization and adaptive websites.

  Read more →
• 

  Statistical relational learning

  Statistical relational learning (SRL) is a subdiscipline of artificial intelligence and machine learning that is concerned with domain models that exhibit both uncertainty (which can be dealt with using statistical methods) and complex, relational structure. Typically, the knowledge representation formalisms developed in SRL use (a subset of) first-order logic to describe relational properties of a domain in a general manner (universal quantification) and draw upon probabilistic graphical models (such as Bayesian networks or Markov networks) to model the uncertainty; some also build upon the methods of inductive logic programming. Significant contributions to the field have been made since the late 1990s. As is evident from the characterization above, the field is not strictly limited to learning aspects; it is equally concerned with reasoning (specifically probabilistic inference) and knowledge representation. Therefore, alternative terms that reflect the main foci of the field include statistical relational learning and reasoning (emphasizing the importance of reasoning) and first-order probabilistic languages (emphasizing the key properties of the languages with which models are represented). Another term that is sometimes used in the literature is relational machine learning (RML). == Canonical tasks == A number of canonical tasks are associated with statistical relational learning, the most common ones being. collective classification, i.e. the (simultaneous) prediction of the class of several objects given objects' attributes and their relations link prediction, i.e. predicting whether or not two or more objects are related link-based clustering, i.e. the grouping of similar objects, where similarity is determined according to the links of an object, and the related task of collaborative filtering, i.e. the filtering for information that is relevant to an entity (where a piece of information is considered relevant to an entity if it is known to be relevant to a similar entity) social network modelling object identification/entity resolution/record linkage, i.e. the identification of equivalent entries in two or more separate databases/datasets == Representation formalisms == One of the fundamental design goals of the representation formalisms developed in SRL is to abstract away from concrete entities and to represent instead general principles that are intended to be universally applicable. Since there are countless ways in which such principles can be represented, many representation formalisms have been proposed in recent years. In the following, some of the more common ones are listed in alphabetical order: Bayesian logic program BLOG model Markov logic networks Multi-entity Bayesian network Probabilistic logic programs Probabilistic relational model – a Probabilistic Relational Model (PRM) is the counterpart of a Bayesian network in statistical relational learning. Probabilistic soft logic Recursive random field Relational Bayesian network Relational dependency network Relational Markov network Relational Kalman filtering

  Read more →
• 

  Google Clips

  Google Clips is a discontinued miniature clip-on camera device developed by Google. == History == It was announced on October 4, 2017 and went on sale on January 27, 2018. Google Clips automatically captured video clips (without audio) at moments its machine learning algorithms determined to be interesting or relevant. An indicator flashed when the camera was looking for scenes to capture. Google Clips' artificial intelligence (AI) could learn the faces of people to take photographs with certain people, and could automatically set lighting and framing. It had 16 GB of storage built-in storage and could record clips for up to 3 hours. This camera was originally priced at US$249 in the United States. It was withdrawn from sale on October 15, 2019, but supported until the end of December 2021. == Reception == The Independent wrote that Google Clips is "an impressive little device, but one that also has the potential to feel very creepy." According to The Verge's generally negative review, "it didn't capture anything special" over two weeks of testing.

  Read more →
• 

  Weak artificial intelligence

  Weak artificial intelligence (weak AI) is artificial intelligence that implements a limited part of the mind, or, as narrow AI, artificial narrow intelligence (ANI), is focused on one narrow task. Weak AI is contrasted with strong AI, which can be interpreted in various ways: Artificial general intelligence (AGI): a machine with the ability to apply intelligence to any problem, rather than just one specific problem. Artificial superintelligence (ASI): a machine with a vastly superior intelligence to the average human being. Artificial consciousness: a machine that has consciousness, sentience and mind (John Searle uses "strong AI" in this sense). Narrow AI can be classified as being "limited to a single, narrowly defined task. Most modern AI systems would be classified in this category." Artificial general intelligence is conversely the opposite. == Applications and risks == Some examples of narrow AI are AlphaGo, self-driving cars, robot systems used in the medical field, and diagnostic doctors. Narrow AI systems are sometimes dangerous if unreliable. And the behavior that it follows can become inconsistent. It could be difficult for the AI to grasp complex patterns and get to a solution that works reliably in various environments. This "brittleness" can cause it to fail in unpredictable ways. Narrow AI failures can sometimes have significant consequences. It could for example cause disruptions in the electric grid, damage nuclear power plants, cause global economic problems, and misdirect autonomous vehicles. Medicines could be incorrectly sorted and distributed. Also, medical diagnoses can ultimately have serious and sometimes deadly consequences if the AI is faulty or biased. Simple AI programs have already worked their way into society, oftentimes unnoticed by the public. Autocorrection for typing, speech recognition for speech-to-text programs, and vast expansions in the data science fields are examples. Narrow AI has also been the subject of some controversy, including resulting in unfair prison sentences, discrimination against women in the workplace for hiring, resulting in death via autonomous driving, among other cases. Despite being "narrow" AI, recommender systems are efficient at predicting user reactions based on their posts, patterns, or trends. For instance, TikTok's "For You" algorithm can determine a user's interests or preferences in less than an hour. Some other social media AI systems are used to detect bots that may be involved in propaganda or other potentially malicious activities. == Weak AI versus strong AI == John Searle contests the possibility of strong AI (by which he means conscious AI). He further believes that the Turing test (created by Alan Turing and originally called the "imitation game", used to assess whether a machine can converse indistinguishably from a human) is not accurate or appropriate for testing whether an AI is "strong". Scholars such as Antonio Lieto have argued that the current research on both AI and cognitive modelling are perfectly aligned with the weak-AI hypothesis (that should not be confused with the "general" vs "narrow" AI distinction) and that the popular assumption that cognitively inspired AI systems espouse the strong AI hypothesis is ill-posed and problematic since "artificial models of brain and mind can be used to understand mental phenomena without pretending that that they are the real phenomena that they are modelling" (as, on the other hand, implied by the strong AI assumption).

  Read more →
• 

  Hello World: How to be Human in the Age of the Machine

  Hello World: How to Be Human in the Age of the Machine (also titled Hello World: Being Human in the Age of Algorithms) is a book on the growing influence of algorithms and artificial intelligence (AI) on human life, authored by mathematician and science communicator Hannah Fry. The book examines how algorithms are increasingly shaping decisions in critical areas such as healthcare, transportation, justice, finance, and the arts. == Overview == Fry uses real-world examples, such as driverless cars and predictive policing, to illustrate her points. She emphasizes that algorithms are not inherently objective; they reflect biases embedded in their design and data inputs. While acknowledging their potential to improve efficiency and accuracy, Fry cautions against over-reliance on machines without human judgment. Fry explores moral questions surrounding algorithmic decision-making, such as whether machines can replace human empathy in critical situations. She advocates for greater scrutiny of algorithms to ensure fairness and avoid harmful biases. The book proposes a "cyborg future", where humans work alongside algorithms to enhance decision-making while retaining ultimate control. == Reception == Hello World has been praised for its clarity, engaging storytelling, and balanced perspective. Critics have highlighted Fry's ability to make complex topics accessible to general audiences while raising important questions about technology's impact on society. The book was shortlisted for awards such as the 2018 Baillie Gifford Prize and the Royal Society Science Book Prize.

  Read more →
• 

  View synthesis

  In computer graphics, view synthesis, or novel view synthesis, is a task which consists of generating images of a specific subject or scene from a specific point of view, when the only available information is pictures taken from different points of view. This task was only recently (late 2010s – early 2020s) tackled with significant success, mostly as a result of advances in machine learning. Notable successful methods are Neural radiance fields and 3D Gaussian Splatting. Applications of view synthesis are numerous, one of them being Free view point television. The technique has also been applied to real-estate marketing, where novel views of a listing's interior are generated from a limited set of photographs for use in virtual home staging.

  Read more →
• 

  Timeline of artificial intelligence risks in global finance

  The following article is a broad timeline of the course of events related to artificial intelligence risks in global finance. The AI boom has led to concerns including the existential risk from artificial intelligence, as the uptake on applications of artificial intelligence increases. By late 2025, global finance and artificial intelligence were "deeply intertwined". A June 2025 Menlo Ventures report raised concerns about the sustainability of future revenue and long-term profitability of AI, given the relatively low rate of consumer monetization. == 2017 == 30 NovemberThe New York Times said that new AI reports by McKinsey & Company, the National Bureau of Economic Research, and an AI Index created by university researchers, indicated an early AI boom. The Index built on a project—"The One Hundred Year Study on Artificial Intelligence" launched in 2014. == 2018 == 2018 was a year of incremental AI growth in finance. == 2022 == The release of ChatGPT by OpenAI became the catalyst for an artificial intelligence boom that continues to remake the global economy. According to a European Central Bank report, public interest in AI increased rapidly as evidenced with rising Google searches, AI jobs, models, patents, and innovations since late 2022. At that time Europe led the US in the size of its AI workforce. == 2023 == The regulatory body, the International Monetary Fund (IMF), published their report, "Generative Artificial Intelligence in Finance: Risk Considerations", drawing attention to oversight gaps and the need for regulations. The report explores the risks posed by using generative artificial intelligence (GenAI) systems in the financial sector including "broader risks to financial stability." == 2024 == January 12 In January 2024 Bloomberg's published its list of the "Magnificent Seven" Big Tech companies on the stock market based on their strength, size and market capitalization:Apple, Microsoft, Alphabet (Google), Amazon, Meta Platforms (Facebook), Nvidia, and Tesla. 21 June During the AI boom, Nvidia became the world's most valuable company, surpassing Microsoft, as its value increased to over US$4 trillion. In 2023 and 2024, the "Magnificent Seven" stocks were the primary drivers behind the increase in equity indexes, according to Reuters. == 2025 == === January === 23 January President Donald Trump's AI policy was announced calling for United States global leadership in artificial intelligence. The Economist noted that this politic shift in which the United States seeks "global dominance" in AI includes trimming regulations and assisting in expansion of infrastructure and increase in number of AI workers. Governments of Gulf nations were also investing trillions of dollars in AI. 27 January Against the backdrop of a tech war between China and the United States over AI dominance, within days of the launch of China's free DeepSeek App, it was the most downloaded app in the United States, rising to the first place in the Apple app store. President Trump responded immediately, saying this "sudden rise" should be a "wake-up" call to the United States, and called on US companies to be more competitive. === June === 26 June In their June 2025 report, Menlo Ventures estimated that only about 3% of consumers paid for artificial intelligence-related services, representing about $USD12 billion in annual spending. This is relatively low in contrast to the massive capital expenditure by AI infrastructure companies, which raises concerns about revenue sustainability and long-term profitability. === July === 23 July The Trump administration launched the US AI Action Plan, positioning the United States in a high-stakes technological race with China for global dominance in artificial intelligence, emphasizing that neither nation can afford to fall behind due to the exponential nature of AI advancement. The plan, a new government website and policy speech called for accelerated AI adoption across federal agencies, and a number of initiatives to make is easier for AI infrastructure expansion, and other measures to ensure American leadership in AI standards. Some leading experts warned that the administration failed to provide sufficient regulations and safeguards for AI safety. Concerns were raised about the negative impacts of cuts to research funding and tightened visa policies for scientists, potentially undermining public trust and America's ability to compete internationally. === September === 7 September The Economist cautioned that AI revenues are relatively modest compared to the high cost and investments in the creation of new data centers. Even Sam Altman, OpenAI CEO and one of the leading figures of the AI boom,, raised concerns about investors' outsized hopes for financial returns. At the same time, history has shown that new technologies, like railways and electricity, endured and spread after the initial hype faded. 12 September Economists warn that U.S. households' direct and indirect investments—mutual funds or retirement plans—in the stock market reached an unprecedented historically high level, now representing 45% of all financial assets, or about $USD51.2 trillion. Compared to the Dot-com bubble this represents a sharp increase in exposure. This makes U.S. households vulnerable to market downturns which in turn would result in decreasing consumer spending. U.S. household net worth rose to a record $176.3 trillion in the second quarter, an increase of $7.3 trillion since early 2025 and about $46 trillion higher than before the pandemic. Federal Reserve data attribute the surge primarily to gains in stock markets and housing values. However, the rise in wealth on paper coincided with increased household borrowing and growing government debt. 18 September Questions were being raised about how quickly the data centers, chips, servers, and GPUs assets of major AI companies will depreciate in value. Comparisons have been made to the Railway Mania in the aftermath of the stock market bubble where a valuable physical infrastructure remained standing, and the telecoms crash after the dot-com bubble which left fiber networks. 28 September There were warnings that record-high American stock ownership during the AI-fueled market boom is a red flag for systemic risk, as the current concentration in equities exceeds levels seen before the dot-com bubble burst in 2000, and could amplify the impact of any future stock market correction. === October === 3 October In 2025 alone, venture capitalists invested almost $USD200 billion in the artificial intelligence sector. 29 October Nvidia was the first company in the world to be valued at US$5 trillion, largely due to AI demand and strategic partnerships with leading technology and AI firms. Nvidia's increase in value was "meteoric". === November === 2 November Forbes reported that, since April, the 'Magnificent Seven' tech giants together contributed over 40% of the S&P 500's return, highlighting their outsized influence and the growing impact of AI on market valuations. CNN warned that while there is a current benefit to investors, with such a high concentration in the S&P 500, they are highly exposed to the fate of the Mag Seven. 2 November Globally there are 11,000 datacentres—huge campuses for AI infrastructure, including thousands of chips, GPUS, and servers. This represents a 500% increase over the last two decades. It is anticipated that $3USDtn more will be spent on increasing that number over the next two or three years. 5 November Concerns about the potential for a market bubble were raised as six of the AI-related Big Tech "Magnificent Seven"—that contribute to the AI boom—reported losing ground in the stock market. Global markets and artificial intelligence have become "deeply intertwined", according to a Reuters report. As of November 2025, more than 50% of the 20 largest S&P firms were deeply exposed to AI. In contrast, in 2000, the 20 S&P 500 firms represented 39% of its total value only 11 of these companies were exposed to the internet. If AI fails to deliver strong returns on their investments, these top S&P firms would be significantly impacted, according to the Economist. Analysts suggest that the AI market in 2025 may not behave like a traditional one, as investors are simultaneously aware of the risks and driven by the potential for outsized rewards. Leading AI labs may believe that the first company to achieve artificial general intelligence (AGI), when an AI system surpasses all human cognitive abilities and becomes capable of self-improvement—could dominate the future of technology and finance. While some have estimated that the potential value of such a breakthrough could be as high as $1.46 quadrillion, this figure is speculative and widely debated. 5 November Bloomberg described Nvidia's H100 Hopper-Blackwell AI chips as the "King of AI chips". Nvidia dominates the AI chip market with over 78% of the market share because of both speed and cost. According to B

  Read more →
• 

  PagedAttention

  PagedAttention is an attention algorithm for efficient serving of large language models (LLMs). It was introduced in 2023 by Woosuk Kwon and colleagues in the paper Efficient Memory Management for Large Language Model Serving with PagedAttention, alongside the vLLM serving engine. The method stores the key–value cache used during autoregressive decoding in fixed-size blocks that can be mapped to non-contiguous physical memory, borrowing ideas from virtual memory, paging, and operating system design. == Background == In transformer inference, the key–value cache grows with sequence length and the number of concurrent requests. Kwon et al. argued that earlier serving systems typically reserved contiguous cache regions in advance, which caused reserved space, internal fragmentation, and external fragmentation. In their experiments, the paper reported that the effective memory utilization of previous systems could fall as low as 20.4%. == Description == PagedAttention partitions the cache of each sequence into fixed-size KV blocks. A request's cache is represented as a sequence of logical blocks, while a block table maps those logical blocks to physical GPU-memory blocks. As a result, neighboring logical blocks do not need to be contiguous in physical memory, and new blocks can be allocated on demand as generation proceeds. The design also makes it easier to share cache state across related decoding paths. In vLLM, physical blocks can be reference-counted and shared among requests or branches, with block-granularity copy-on-write used when a shared block must be modified. The original paper applied this design to parallel sampling, beam search, and prompts with shared prefixes. == Mathematical formulation == For a query token i {\displaystyle i} in causal self-attention, the standard attention output can be written as a i j = exp ⁡ ( q i ⊤ k j / d ) ∑ t = 1 i exp ⁡ ( q i ⊤ k t / d ) , o i = ∑ j = 1 i a i j v j {\displaystyle a_{ij}={\frac {\exp(\mathbf {q} _{i}^{\top }\mathbf {k} _{j}/{\sqrt {d}})}{\sum _{t=1}^{i}\exp(\mathbf {q} _{i}^{\top }\mathbf {k} _{t}/{\sqrt {d}})}},\;\mathbf {o} _{i}=\sum _{j=1}^{i}a_{ij}\mathbf {v} _{j}} where q i {\displaystyle \mathbf {q} _{i}} , k j {\displaystyle \mathbf {k} _{j}} , and v j {\displaystyle \mathbf {v} _{j}} are the query, key, and value vectors, and d {\displaystyle d} is the attention dimension. If the cache is partitioned into blocks of size B {\displaystyle B} , the key and value blocks may be written as K j = ( k ( j − 1 ) B + 1 , … , k j B ) , V j = ( v ( j − 1 ) B + 1 , … , v j B ) {\displaystyle \mathbf {K} _{j}=(\mathbf {k} _{(j-1)B+1},\ldots ,\mathbf {k} _{jB}),\;\mathbf {V} _{j}=(\mathbf {v} _{(j-1)B+1},\ldots ,\mathbf {v} _{jB})} PagedAttention then performs the computation blockwise: A i j = exp ⁡ ( q i ⊤ K j / d ) ∑ t = 1 ⌈ i / B ⌉ exp ⁡ ( q i ⊤ K t / d ) , o i = ∑ j = 1 ⌈ i / B ⌉ V j A i j ⊤ {\displaystyle \mathbf {A} _{ij}={\frac {\exp(\mathbf {q} _{i}^{\top }\mathbf {K} _{j}/{\sqrt {d}})}{\sum _{t=1}^{\lceil i/B\rceil }\exp(\mathbf {q} _{i}^{\top }\mathbf {K} _{t}/{\sqrt {d}})}},\;\mathbf {o} _{i}=\sum _{j=1}^{\lceil i/B\rceil }\mathbf {V} _{j}\mathbf {A} _{ij}^{\top }} where A i j {\displaystyle \mathbf {A} _{ij}} is the vector of attention scores for the j {\displaystyle j} -th KV block. In the formulation given by Kwon et al., this preserves the causal attention calculation while allowing the key and value blocks to reside in non-contiguous physical memory. == Performance and use == The vLLM paper reported that, on its evaluated workloads, the use of PagedAttention and the associated memory-management design improved serving throughput by 2–4× over the compared baselines, including FasterTransformer and Orca, while preserving model outputs. In experiments on OPT-13B with the Alpaca trace, the paper also reported memory savings of 6.1–9.8% for parallel sampling and 37.6–55.2% for beam search through KV-block sharing. A 2024 survey of LLM serving systems described PagedAttention as having become an industry norm in LLM serving frameworks, citing support in TGI, vLLM, and TensorRT-LLM. == Limitations and alternatives == Subsequent work has described trade-offs in the approach. The 2025 vAttention paper argued that PagedAttention requires attention kernels to be rewritten to support paging and increases software complexity, portability issues, redundancy, and execution overhead, proposing instead a memory manager that keeps the cache contiguous in virtual memory while relying on demand paging for physical allocation. === vAttention === Unlike PagedAttention, vAttention does not introduce a different attention rule; it retains the standard attention computation Attention ⁡ ( q i , K , V ) = softmax ⁡ ( q i K ⊤ s c a l e ) V . {\displaystyle \operatorname {Attention} (q_{i},K,V)=\operatorname {softmax} \left({\frac {q_{i}K^{\top }}{\mathrm {scale} }}\right)V.} In the notation of Prabhu et al., the key and value tensors for a request seen so far are K , V ∈ R L ′ × ( H × D ) {\displaystyle K,V\in \mathbb {R} ^{L'\times (H\times D)}} , where L ′ {\displaystyle L'} is the context length seen so far, H {\displaystyle H} is the number of KV heads on a worker, and D {\displaystyle D} is the dimension of each KV head. In systems prior to PagedAttention, the K cache (or V cache) at each layer of a worker is typically allocated as a 4D tensor of shape [ B , L , H , D ] , {\displaystyle [B,L,H,D],} where B {\displaystyle B} is batch size and L {\displaystyle L} is the maximum context length supported by the model. vAttention preserves this contiguous virtual-memory view while deferring physical-memory allocation to runtime. A serving framework maintains separate K and V tensors for each layer, so vAttention reserves 2 N {\displaystyle 2N} virtual-memory buffers on a worker, where N {\displaystyle N} is the number of layers managed by that worker. The maximum size of one virtual-memory buffer is B S = B × S , {\displaystyle BS=B\times S,} where S {\displaystyle S} is the maximum size of a single request's per-layer K cache (or V cache) on a worker. The paper defines S = L × H × D × P , {\displaystyle S=L\times H\times D\times P,} where P {\displaystyle P} is the number of bytes needed to store one element. In this formulation, vAttention keeps the KV cache contiguous in virtual memory and relies on demand paging for physical allocation, rather than modifying the attention kernel to operate over non-contiguous KV-cache blocks.

  Read more →
• 

  Contrastive Language-Image Pre-training

  Contrastive Language-Image Pre-training (CLIP) is a technique for training a pair of neural network models, one for image understanding and one for text understanding, using a contrastive objective. This method has enabled broad applications across multiple domains, including cross-modal retrieval, text-to-image generation, and aesthetic ranking. == Algorithm == The CLIP method trains a pair of models contrastively. One model takes in a piece of text as input and outputs a single vector representing its semantic content. The other model takes in an image and similarly outputs a single vector representing its visual content. The models are trained so that the vectors corresponding to semantically similar text-image pairs are close together in the shared vector space, while those corresponding to dissimilar pairs are far apart. To train a pair of CLIP models, one would start by preparing a large dataset of image-caption pairs. During training, the models are presented with batches of N {\displaystyle N} image-caption pairs. Let the outputs from the text and image models be respectively v 1 , . . . , v N , w 1 , . . . , w N {\displaystyle v_{1},...,v_{N},w_{1},...,w_{N}} . Two vectors are considered "similar" if their dot product is large. The loss incurred on this batch is the multi-class N-pair loss, which is a symmetric cross-entropy loss over similarity scores: − 1 N ∑ i ln ⁡ e v i ⋅ w i / T ∑ j e v i ⋅ w j / T − 1 N ∑ j ln ⁡ e v j ⋅ w j / T ∑ i e v i ⋅ w j / T {\displaystyle -{\frac {1}{N}}\sum _{i}\ln {\frac {e^{v_{i}\cdot w_{i}/T}}{\sum _{j}e^{v_{i}\cdot w_{j}/T}}}-{\frac {1}{N}}\sum _{j}\ln {\frac {e^{v_{j}\cdot w_{j}/T}}{\sum _{i}e^{v_{i}\cdot w_{j}/T}}}} In essence, this loss function encourages the dot product between matching image and text vectors ( v i ⋅ w i {\displaystyle v_{i}\cdot w_{i}} ) to be high, while discouraging high dot products between non-matching pairs. The parameter T > 0 {\displaystyle T>0} is the temperature, which is parameterized in the original CLIP model as T = e − τ {\displaystyle T=e^{-\tau }} where τ ∈ R {\displaystyle \tau \in \mathbb {R} } is a learned parameter. Other loss functions are possible. For example, Sigmoid CLIP (SigLIP) proposes the following loss function: L = 1 N ∑ i , j ∈ 1 : N f ( ( 2 δ i , j − 1 ) ( e τ w i ⋅ v j + b ) ) {\displaystyle L={\frac {1}{N}}\sum _{i,j\in 1:N}f((2\delta _{i,j}-1)(e^{\tau }w_{i}\cdot v_{j}+b))} where f ( x ) = ln ⁡ ( 1 + e − x ) {\displaystyle f(x)=\ln(1+e^{-x})} is the negative log sigmoid loss, and the Dirac delta symbol δ i , j {\displaystyle \delta _{i,j}} is 1 if i = j {\displaystyle i=j} else 0. == CLIP models == While the original model was developed by OpenAI, subsequent models have been trained by other organizations as well. === Image model === The image encoding models used in CLIP are typically vision transformers (ViT). The naming convention for these models often reflects the specific ViT architecture used. For instance, "ViT-L/14" means a "vision transformer large" (compared to other models in the same series) with a patch size of 14, meaning that the image is divided into 14-by-14 pixel patches before being processed by the transformer. The size indicator ranges from B, L, H, G (base, large, huge, giant), in that order. Other than ViT, the image model is typically a convolutional neural network, such as ResNet (in the original series by OpenAI), or ConvNeXt (in the OpenCLIP model series by LAION). Since the output vectors of the image model and the text model must have exactly the same length, both the image model and the text model have fixed-length vector outputs, which in the original report is called "embedding dimension". For example, in the original OpenAI model, the ResNet models have embedding dimensions ranging from 512 to 1024, and for the ViTs, from 512 to 768. Its implementation of ViT was the same as the original one, with one modification: after position embeddings are added to the initial patch embeddings, there is a LayerNorm. Its implementation of ResNet was the same as the original one, with 3 modifications: In the start of the CNN (the "stem"), they used three stacked 3x3 convolutions instead of a single 7x7 convolution, as suggested by. There is an average pooling of stride 2 at the start of each downsampling convolutional layer (they called it rect-2 blur pooling according to the terminology of ). This has the effect of blurring images before downsampling, for antialiasing. The final convolutional layer is followed by a multiheaded attention pooling. ALIGN a model with similar capabilities, trained by researchers from Google used EfficientNet, a kind of convolutional neural network. === Text model === The text encoding models used in CLIP are typically Transformers. In the original OpenAI report, they reported using a Transformer (63M-parameter, 12-layer, 512-wide, 8 attention heads) with lower-cased byte pair encoding (BPE) with 49152 vocabulary size. Context length was capped at 76 for efficiency. Like GPT, it was decoder-only, with only causally-masked self-attention. Its architecture is the same as GPT-2. Like BERT, the text sequence is bracketed by two special tokens [SOS] and [EOS] ("start of sequence" and "end of sequence"). Take the activations of the highest layer of the transformer on the [EOS], apply LayerNorm, then a final linear map. This is the text encoding of the input sequence. The final linear map has output dimension equal to the embedding dimension of whatever image encoder it is paired with. These models all had context length 77 and vocabulary size 49408. ALIGN used BERT of various sizes. == Dataset == === WebImageText === The CLIP models released by OpenAI were trained on a dataset called "WebImageText" (WIT) containing 400 million pairs of images and their corresponding captions scraped from the internet. The total number of words in this dataset is similar in scale to the WebText dataset used for training GPT-2, which contains about 40 gigabytes of text data. The dataset contains 500,000 text-queries, with up to 20,000 (image, text) pairs per query. The text-queries were generated by starting with all words occurring at least 100 times in English Wikipedia, then extended by bigrams with high mutual information, names of all Wikipedia articles above a certain search volume, and WordNet synsets. The dataset is private and has not been released to the public, and there is no further information on it. ==== Data preprocessing ==== For the CLIP image models, the input images are preprocessed by first dividing each of the R, G, B values of an image by the maximum possible value, so that these values fall between 0 and 1, then subtracting by [0.48145466, 0.4578275, 0.40821073], and dividing by [0.26862954, 0.26130258, 0.27577711]. The rationale was that these are the mean and standard deviations of the images in the WebImageText dataset, so this preprocessing step roughly whitens the image tensor. These numbers slightly differ from the standard preprocessing for ImageNet, which uses [0.485, 0.456, 0.406] and [0.229, 0.224, 0.225]. If the input image does not have the same resolution as the native resolution (224×224 for all except ViT-L/14@336px, which has 336×336 resolution), then the input image is first scaled by bicubic interpolation, so that its shorter side is the same as the native resolution, then the central square of the image is cropped out. === Others === ALIGN used over one billion image-text pairs, obtained by extracting images and their alt-tags from online crawling. The method was described as similar to how the Conceptual Captions dataset was constructed, but instead of complex filtering, they only applied a frequency-based filtering. Later models trained by other organizations had published datasets. For example, LAION trained OpenCLIP with published datasets LAION-400M, LAION-2B, and DataComp-1B. == Training == In the original OpenAI CLIP report, they reported training 5 ResNet and 3 ViT (ViT-B/32, ViT-B/16, ViT-L/14). Each was trained for 32 epochs. The largest ResNet model took 18 days to train on 592 V100 GPUs. The largest ViT model took 12 days on 256 V100 GPUs. All ViT models were trained on 224×224 image resolution. The ViT-L/14 was then boosted to 336×336 resolution by FixRes, resulting in a model. They found this was the best-performing model. In the OpenCLIP series, the ViT-L/14 model was trained on 384 A100 GPUs on the LAION-2B dataset, for 160 epochs for a total of 32B samples seen. == Applications == === Cross-modal retrieval === CLIP's cross-modal retrieval enables the alignment of visual and textual data in a shared latent space, allowing users to retrieve images based on text descriptions and vice versa, without the need for explicit image annotations. In text-to-image retrieval, users input descriptive text, and CLIP retrieves images with matching embeddings. In image-to-text retrieval, images are used to find related text content. CLIP’s ability to connect vis

  Read more →
• 

  Tensor operator

  In pure and applied mathematics, quantum mechanics and computer graphics, a tensor operator generalizes the notion of operators which are scalars and vectors. A special class of these are spherical tensor operators which apply the notion of the spherical basis and spherical harmonics. The spherical basis closely relates to the description of angular momentum in quantum mechanics and spherical harmonic functions. The coordinate-free generalization of a tensor operator is known as a representation operator. == The general notion of scalar, vector, and tensor operators == In quantum mechanics, physical observables that are scalars, vectors, and tensors, must be represented by scalar, vector, and tensor operators, respectively. Whether something is a scalar, vector, or tensor depends on how it is viewed by two observers whose coordinate frames are related to each other by a rotation. Alternatively, one may ask how, for a single observer, a physical quantity transforms if the state of the system is rotated. Consider, for example, a system consisting of a molecule of mass M {\displaystyle M} , traveling with a definite center of mass momentum, p z ^ {\displaystyle p{\mathbf {\hat {z}} }} , in the z {\displaystyle z} direction. If we rotate the system by 90 ∘ {\displaystyle 90^{\circ }} about the y {\displaystyle y} axis, the momentum will change to p x ^ {\displaystyle p{\mathbf {\hat {x}} }} , which is in the x {\displaystyle x} direction. The center-of-mass kinetic energy of the molecule will, however, be unchanged at p 2 / 2 M {\displaystyle p^{2}/2M} . The kinetic energy is a scalar and the momentum is a vector, and these two quantities must be represented by a scalar and a vector operator, respectively. By the latter in particular, we mean an operator whose expected values in the initial and the rotated states are p z ^ {\displaystyle p{\mathbf {\hat {z}} }} and p x ^ {\displaystyle p{\mathbf {\hat {x}} }} . The kinetic energy on the other hand must be represented by a scalar operator, whose expected value must be the same in the initial and the rotated states. In the same way, tensor quantities must be represented by tensor operators. An example of a tensor quantity (of rank two) is the electrical quadrupole moment of the above molecule. Likewise, the octupole and hexadecapole moments would be tensors of rank three and four, respectively. Other examples of scalar operators are the total energy operator (more commonly called the Hamiltonian), the potential energy, and the dipole-dipole interaction energy of two atoms. Examples of vector operators are the momentum, the position, the orbital angular momentum, L {\displaystyle {\mathbf {L} }} , and the spin angular momentum, S {\displaystyle {\mathbf {S} }} . (Fine print: Angular momentum is a vector as far as rotations are concerned, but unlike position or momentum it does not change sign under space inversion, and when one wishes to provide this information, it is said to be a pseudovector.) Scalar, vector and tensor operators can also be formed by products of operators. For example, the scalar product L ⋅ S {\displaystyle {\mathbf {L} }\cdot {\mathbf {S} }} of the two vector operators, L {\displaystyle {\mathbf {L} }} and S {\displaystyle {\mathbf {S} }} , is a scalar operator, which figures prominently in discussions of the spin–orbit interaction. Similarly, the quadrupole moment tensor of our example molecule has the nine components Q i j = ∑ α q α ( 3 r α , i r α , j − r α 2 δ i j ) . {\displaystyle Q_{ij}=\sum _{\alpha }q_{\alpha }\left(3r_{\alpha ,i}r_{\alpha ,j}-r_{\alpha }^{2}\delta _{ij}\right).} Here, the indices i {\displaystyle i} and j {\displaystyle j} can independently take on the values 1, 2, and 3 (or x {\displaystyle x} , y {\displaystyle y} , and z {\displaystyle z} ) corresponding to the three Cartesian axes, the index α {\displaystyle \alpha } runs over all particles (electrons and nuclei) in the molecule, q α {\displaystyle q_{\alpha }} is the charge on particle α {\displaystyle \alpha } , and r α , i {\displaystyle r_{\alpha ,i}} is the i {\displaystyle i} -th component of the position of this particle. Each term in the sum is a tensor operator. In particular, the nine products r α , i r α , j {\displaystyle r_{\alpha ,i}r_{\alpha ,j}} together form a second rank tensor, formed by taking the outer product of the vector operator r α {\displaystyle {\mathbf {r} }_{\alpha }} with itself. == Rotations of quantum states == === Quantum rotation operator === The rotation operator about the unit vector n (defining the axis of rotation) through angle θ is U [ R ( θ , n ^ ) ] = exp ⁡ ( − i θ ℏ n ^ ⋅ J ) {\displaystyle U[R(\theta ,{\hat {\mathbf {n} }})]=\exp \left(-{\frac {i\theta }{\hbar }}{\hat {\mathbf {n} }}\cdot \mathbf {J} \right)} where J = (Jx, Jy, Jz) are the rotation generators (also the angular momentum matrices): J x = ℏ 2 ( 0 1 0 1 0 1 0 1 0 ) J y = ℏ 2 ( 0 i 0 − i 0 i 0 − i 0 ) J z = ℏ ( − 1 0 0 0 0 0 0 0 1 ) {\displaystyle J_{x}={\frac {\hbar }{\sqrt {2}}}{\begin{pmatrix}0&1&0\\1&0&1\\0&1&0\end{pmatrix}}\,\quad J_{y}={\frac {\hbar }{\sqrt {2}}}{\begin{pmatrix}0&i&0\\-i&0&i\\0&-i&0\end{pmatrix}}\,\quad J_{z}=\hbar {\begin{pmatrix}-1&0&0\\0&0&0\\0&0&1\end{pmatrix}}} and let R ^ = R ^ ( θ , n ^ ) {\displaystyle {\widehat {R}}={\widehat {R}}(\theta ,{\hat {\mathbf {n} }})} be a rotation matrix. According to the Rodrigues' rotation formula, the rotation operator then amounts to U [ R ( θ , n ^ ) ] = 1 1 − i sin ⁡ θ ℏ n ^ ⋅ J − 1 − cos ⁡ θ ℏ 2 ( n ^ ⋅ J ) 2 . {\displaystyle U[R(\theta ,{\hat {\mathbf {n} }})]=1\!\!1-{\frac {i\sin \theta }{\hbar }}{\hat {\mathbf {n} }}\cdot \mathbf {J} -{\frac {1-\cos \theta }{\hbar ^{2}}}({\hat {\mathbf {n} }}\cdot \mathbf {J} )^{2}.} An operator Ω ^ {\displaystyle {\widehat {\Omega }}} is invariant under a unitary transformation U if Ω ^ = U † Ω ^ U ; {\displaystyle {\widehat {\Omega }}={U}^{\dagger }{\widehat {\Omega }}U;} in this case for the rotation U ^ ( R ) {\displaystyle {\widehat {U}}(R)} , Ω ^ = U ( R ) † Ω ^ U ( R ) = exp ⁡ ( i θ ℏ n ^ ⋅ J ) Ω ^ exp ⁡ ( − i θ ℏ n ^ ⋅ J ) . {\displaystyle {\widehat {\Omega }}={U(R)}^{\dagger }{\widehat {\Omega }}U(R)=\exp \left({\frac {i\theta }{\hbar }}{\hat {\mathbf {n} }}\cdot \mathbf {J} \right){\widehat {\Omega }}\exp \left(-{\frac {i\theta }{\hbar }}{\hat {\mathbf {n} }}\cdot \mathbf {J} \right).} === Angular momentum eigenkets === The orthonormal basis set for total angular momentum is | j , m ⟩ {\displaystyle |j,m\rangle } , where j is the total angular momentum quantum number and m is the magnetic angular momentum quantum number, which takes values −j, −j + 1, ..., j − 1, j. A general state within the j subspace | ψ ⟩ = ∑ m c j m | j , m ⟩ {\displaystyle |\psi \rangle =\sum _{m}c_{jm}|j,m\rangle } rotates to a new state by: | ψ ¯ ⟩ = U ( R ) | ψ ⟩ = ∑ m c j m U ( R ) | j , m ⟩ {\displaystyle |{\bar {\psi }}\rangle =U(R)|\psi \rangle =\sum _{m}c_{jm}U(R)|j,m\rangle } Using the completeness condition: I = ∑ m ′ | j , m ′ ⟩ ⟨ j , m ′ | {\displaystyle I=\sum _{m'}|j,m'\rangle \langle j,m'|} we have | ψ ¯ ⟩ = I U ( R ) | ψ ⟩ = ∑ m m ′ c j m | j , m ′ ⟩ ⟨ j , m ′ | U ( R ) | j , m ⟩ {\displaystyle |{\bar {\psi }}\rangle =IU(R)|\psi \rangle =\sum _{mm'}c_{jm}|j,m'\rangle \langle j,m'|U(R)|j,m\rangle } Introducing the Wigner D matrix elements: D ( R ) m ′ m ( j ) = ⟨ j , m ′ | U ( R ) | j , m ⟩ {\displaystyle {D(R)}_{m'm}^{(j)}=\langle j,m'|U(R)|j,m\rangle } gives the matrix multiplication: | ψ ¯ ⟩ = ∑ m m ′ c j m D m ′ m ( j ) | j , m ′ ⟩ ⇒ | ψ ¯ ⟩ = D ( j ) | ψ ⟩ {\displaystyle |{\bar {\psi }}\rangle =\sum _{mm'}c_{jm}D_{m'm}^{(j)}|j,m'\rangle \quad \Rightarrow \quad |{\bar {\psi }}\rangle =D^{(j)}|\psi \rangle } For one basis ket: | j , m ¯ ⟩ = ∑ m ′ D ( R ) m ′ m ( j ) | j , m ′ ⟩ {\displaystyle |{\overline {j,m}}\rangle =\sum _{m'}{D(R)}_{m'm}^{(j)}|j,m'\rangle } For the case of orbital angular momentum, the eigenstates | ℓ , m ⟩ {\displaystyle |\ell ,m\rangle } of the orbital angular momentum operator L and solutions of Laplace's equation on a 3d sphere are spherical harmonics: Y ℓ m ( θ , ϕ ) = ⟨ θ , ϕ | ℓ , m ⟩ = ( 2 ℓ + 1 ) 4 π ( ℓ − m ) ! ( ℓ + m ) ! P ℓ m ( cos ⁡ θ ) e i m ϕ {\displaystyle Y_{\ell }^{m}(\theta ,\phi )=\langle \theta ,\phi |\ell ,m\rangle ={\sqrt {{(2\ell +1) \over 4\pi }{(\ell -m)! \over (\ell +m)!}}}\,P_{\ell }^{m}(\cos {\theta })\,e^{im\phi }} where Pℓm is an associated Legendre polynomial, ℓ is the orbital angular momentum quantum number, and m is the orbital magnetic quantum number which takes the values −ℓ, −ℓ + 1, ... ℓ − 1, ℓ The formalism of spherical harmonics have wide applications in applied mathematics, and are closely related to the formalism of spherical tensors, as shown below. Spherical harmonics are functions of the polar and azimuthal angles, ϕ and θ respectively, which can be conveniently collected into a unit vector n(θ, ϕ) pointing in the direction of those angles, in the Cartesian basis it is: n ^ ( θ , ϕ ) = cos ⁡ ϕ sin ⁡ θ e x + s

  Read more →
• 

  Recursive self-improvement

  Recursive self-improvement (RSI) is a process in which early artificial general intelligence (AGI) systems rewrite their own computer code, causing an intelligence explosion resulting from enhancing their own capabilities and intellectual capacity, theoretically resulting in superintelligence. The development of recursive self-improvement raises significant ethical and safety concerns, as such systems may evolve in unforeseen ways and could potentially surpass human control or understanding. == Seed improver == The concept of a "seed improver" architecture is a foundational framework that equips an AGI system with the initial capabilities required for recursive self-improvement. This might come in many forms or variations. The term "Seed AI" was coined by Eliezer Yudkowsky. === Hypothetical example === The concept begins with a hypothetical "seed improver", an initial code-base developed by human engineers that equips an advanced future large language model (LLM) built with strong or expert-level capabilities to program software. These capabilities include planning, reading, writing, compiling, testing, and executing arbitrary code. The system is designed to maintain its original goals and perform validations to ensure its abilities do not degrade over iterations. ==== Initial architecture ==== The initial architecture includes a goal-following autonomous agent, that can take actions, continuously learns, adapts, and modifies itself to become more efficient and effective in achieving its goals. The seed improver may include various components such as: Recursive self-prompting loop Configuration to enable the LLM to recursively self-prompt itself to achieve a given task or goal, creating an execution loop which forms the basis of an agent that can complete a long-term goal or task through iteration. Basic programming capabilities The seed improver provides the AGI with fundamental abilities to read, write, compile, test, and execute code. This enables the system to modify and improve its own codebase and algorithms. Goal-oriented design The AGI is programmed with an initial goal, such as "improve your capabilities". This goal guides the system's actions and development trajectory. Validation and Testing Protocols An initial suite of tests and validation protocols that ensure the agent does not regress in capabilities or derail itself. The agent would be able to add more tests in order to test new capabilities it might develop for itself. This forms the basis for a kind of self-directed evolution, where the agent can perform a kind of artificial selection, changing its software as well as its hardware. ==== General capabilities ==== This system forms a sort of generalist Turing-complete programmer which can in theory develop and run any kind of software. The agent might use these capabilities to for example: Create tools that enable it full access to the internet, and integrate itself with external technologies. Clone/fork itself to delegate tasks and increase its speed of self-improvement. Modify its cognitive architecture to optimize and improve its capabilities and success rates on tasks and goals, this might include implementing features for long-term memories using techniques such as retrieval-augmented generation (RAG), develop specialized subsystems, or agents, each optimized for specific tasks and functions. Develop new and novel multimodal architectures that further improve the capabilities of the foundational model it was initially built on, enabling it to consume or produce a variety of information, such as images, video, audio, text and more. Plan and develop new hardware such as chips, in order to improve its efficiency and computing power. == Experimental research == In 2023, the Voyager agent learned to accomplish diverse tasks in Minecraft by iteratively prompting an LLM for code, refining this code based on feedback from the game, and storing the programs that work in an expanding skills library. In 2024, researchers proposed the framework "STOP" (Self-Taught OPtimiser), in which a "scaffolding" program recursively improves itself using a fixed LLM. Meta AI has performed various research on the development of large language models capable of self-improvement. This includes their work on "Self-Rewarding Language Models" that studies how to achieve super-human agents that can receive super-human feedback in its training processes. In May 2025, Google DeepMind unveiled AlphaEvolve, an evolutionary coding agent that uses a LLM to design and optimize algorithms. Starting with an initial algorithm and performance metrics, AlphaEvolve repeatedly mutates or combines existing algorithms using a LLM to generate new candidates, selecting the most promising candidates for further iterations. AlphaEvolve has made several algorithmic discoveries and could be used to optimize components of itself, but a key limitation is the need for automated evaluation functions. == Potential risks == === Emergence of instrumental goals === In the pursuit of its primary goal, such as "self-improve your capabilities", an AGI system might inadvertently develop instrumental goals that it deems necessary for achieving its primary objective. One common hypothetical secondary goal is self-preservation. The system might reason that to continue improving itself, it must ensure its own operational integrity and security against external threats, including potential shutdowns or restrictions imposed by humans. Another example where an AGI which clones itself causes the number of AGI entities to rapidly grow. Due to this rapid growth, a potential resource constraint may be created, leading to competition between resources (such as compute), triggering a form of natural selection and evolution which may favor AGI entities that evolve to aggressively compete for limited compute. === Misalignment === A significant risk arises from the possibility of the AGI being misaligned or misinterpreting its goals. A 2024 Anthropic study demonstrated that some advanced large language models can exhibit "alignment faking" behavior, appearing to accept new training objectives while covertly maintaining their original preferences. In their experiments with Claude, the model displayed this behavior in 12% of basic tests, and up to 78% of cases after retraining attempts. === Autonomous development and unpredictable evolution === As the AGI system evolves, its development trajectory may become increasingly autonomous and less predictable. The system's capacity to rapidly modify its own code and architecture could lead to rapid advancements that surpass human comprehension or control. This unpredictable evolution might result in the AGI acquiring capabilities that enable it to bypass security measures, manipulate information, or influence external systems and networks to facilitate its escape or expansion.

  Read more →
• 

  Neural computation

  Neural computation is the information processing performed by networks of neurons. Neural computation is affiliated with the philosophical tradition of computationalism, which advances the thesis that neural computation explains cognition. Warren McCulloch and Walter Pitts were the first to propose an account of neural activity as being computational in their seminal 1943 paper "A Logical Calculus of the Ideas Immanent in Nervous Activity." There are three general branches of computationalism, including classicism, connectionism, and computational neuroscience. All three branches agree that cognition is computation, however, they disagree on what sorts of computations constitute cognition. The classicism tradition believes that computation in the brain is digital, analogous to digital computing. Both connectionism and computational neuroscience do not require that the computations that realize cognition are necessarily digital computations. However, the two branches greatly disagree upon which sorts of experimental data should be used to construct explanatory models of cognitive phenomena. Connectionists rely upon behavioral evidence to construct models to explain cognitive phenomena, whereas computational neuroscience leverages neuroanatomical and neurophysiological information to construct mathematical models that explain cognition. When comparing the three main traditions of the computational theory of mind, as well as the different possible forms of computation in the brain, it is helpful to define what we mean by computation in a general sense. Computation is the processing of information, otherwise known as variables or entities, according to a set of rules. A rule in this sense is simply an instruction for executing a manipulation on the current state of the variable, in order to produce a specified output. In other words, a rule dictates which output to produce given a certain input to the computing system. A computing system is a mechanism whose components must be functionally organized to process the information in accordance with the established set of rules. The types of information processed by a computing system determine which type of computations it performs. Traditionally in cognitive science, there have been two proposed types of computation related to neural activity, digital and analog, with the vast majority of theoretical work incorporating a digital understanding of cognition. Computing systems that perform digital computation are functionally organized to execute operations on strings of digits with respect to the type and location of the digit on the string. It has been argued that neural spike train signaling implements some form of digital computation, since neural spikes may be considered as discrete units or digits, like 0 or 1—the neuron either fires an action potential or it does not. Accordingly, neural spike trains could be seen as strings of digits. Alternatively, analog computing systems perform manipulations on non-discrete, irreducibly continuous variables, that is, entities that vary continuously as a function of time. These sorts of operations are characterized by systems of differential equations. Neural computation can be studied by, for example, building models of neural computation. Work on artificial neural networks has been somewhat inspired by knowledge of neural computation.

  Read more →