Ni1000

Ni1000

The Ni1000 is an artificial neural network chip developed by Nestor Corporation and Intel, developed in the 1990s. It is Intel's second-generation neural network chip, but the first all-digital chip. The chip is aimed at image analysis applications– containing more than 3 million transistors – and can analyze 40,000 patterns per second. Prototypes running Nestor's OCR software in 1994 were capable of recognizing around 100 handwritten characters per second. The development was funded with money from DARPA and Office of Naval Research.

AI nationalism

AI nationalism is the idea that nations should develop and control their own artificial intelligence technologies to advance their own interests and ensure technological sovereignty. This concept is gaining traction globally, leading countries to implement new laws, form strategic alliances, and invest significantly in domestic AI capabilities. == Global trends and national strategies == In 2018, British technology investor Ian Hogarth published an influential essay titled AI Nationalism. He argued that as AI gains more power and its economic and military significance expands, governments will take measures to bolster their own domestic AI industries, and predicted that the advancement of machine learning systems would lead to what he termed "AI nationalism." He anticipated that this rise in AI would accelerate a global arms race, resulting in more closed economies, restrictions on foreign acquisitions, and limitations on the movement of talent. Hogarth predicted that AI policy would become a central focus of government agendas. He also criticized Britain’s approach to AI strategy, citing the sale of London-based DeepMind—one of the leading AI laboratories, acquired by Google for a relatively modest £400 million in 2014—as a significant misstep. AI nationalism is chiefly reflected in the escalating rhetoric of an artificial intelligence arms race, portraying AI development as a zero-sum game where the winner gains significant economic, political, and military advantages. This mindset, as highlighted in a 2017 Pentagon report, warns that sharing AI technology could erode technological supremacy and enhance rivals' capabilities. The winner-takes-all mentality of AI nationalism poses risks including unsafe AI development, increased geopolitical tension, and potential military aggression (such as cyberattacks or targeting AI professionals). Several countries, including Canada, France, and India, have formulated national strategies to advance their positions in AI. In the United States, a leading player in the global AI arena, trade policies have been enacted to restrict China's access to critical microchips, reflecting a strategic effort to maintain a technological edge. The United States’ National Security Commission on Artificial Intelligence (NSCAI) frames AI development as a critical aspect of a broader technology competition crucial for national success. It emphasizes the need to outpace China in AI to maintain strategic advantage, reflecting AI nationalism by linking geopolitical power directly to advancements in AI. France has seen notable governmental support for local AI startups, particularly those specializing in language technologies that cater to French and other non-English languages. In Saudi Arabia, Crown Prince Mohammed bin Salman is investing billions in AI research and development. The country has actively collaborated with major technology firms such as Amazon, IBM, and Microsoft to establish itself as a prominent AI hub. == Historical and cultural context == AI nationalism is seen as deeply connected to historical racism and imperialism. It is viewed not merely as a technological competition but as a contest over racial and civilizational superiority. Historically, technological achievements were often used to justify colonialism and racial hierarchies, with Western societies perceiving their advancements as evidence of superiority. In the context of AI, this historical context continues to shape views on intelligence and development. Some argue that AI nationalism reinforces the idea of fundamental civilizational divides, especially between the Western world and China. This perspective often frames China's progress in AI as a direct challenge to Western values, presenting the AI competition as a struggle over values. AI nationalism is said to draw from long-standing anti-Asian stereotypes, such as the "Yellow Peril," which portray Asian nations as threats to Western civilization. This viewpoint links Asian technological advances with dehumanization and artificiality, reflecting persistent anxieties about China's growing role in the global tech landscape. == Implications == AI nationalism is seen as a component of a broader trend towards the fragmentation of the internet, where digital services are increasingly influenced by local regulations and national interests. This shift is creating a new technological landscape in which the impact of artificial intelligence on individuals' lives can vary significantly depending on their geographic location. J. Paul Goode argues that AI nationalism may exacerbate existing societal divisions by promoting the development of systems that embed cultural biases, thereby privileging certain groups while disadvantaging others.

Brand networking

Brand networking is the engagement of a social networking service around a brand by providing consumers with a platform of relevant content, elements of participation, and a currency, score, or ranking. Brand networking creates communities that serve as interactive destinations to encourage brand participation online and off. This evolved level of user participation with the brand facilitates strong relationships with consumers, leverages sales, and generates fan equity. The concept builds on the marketing literature on brand communities, which describes specialized, non-geographically bound groups of consumers organized around shared interest in a brand, and on subsequent research on social-media-based brand communities that examines how such groups operate when embedded in general-purpose networking platforms. == History == The development and growth of social networking in the early 2000s gave birth to brand networking. Brands saw the immediate potential to reach and interact with consumers through online platforms like Facebook and MySpace. At first, the ability to reach consumers through these platforms was inadequate; brands had the option to join as members or simply advertise on these sites. The potential existed to not only display advertisements to consumers, but to encourage them to interact with the brand. This is when brands made the shift to create their own networking platforms. Less evolved attempts to connect brands with consumers via networking are typically built as online platforms meant only to complement a product/service and are limited in functionality. Typically these sites offer consumers the opportunity to interact through discussion boards and group pages. The Guiding Light Community was built to complement the popular CBS television soap opera. The site offers members reward points for contributing content to discussion boards and blogs (which is all geared toward the show). == Structure == Brand networking is more than the utilization of a social networking platform; it is connecting consumers together and constructing relationships directly with the brand. Three key elements, in unity, create effective brand networking: relevant content, elements of participation, and a competitive currency. Websites in conjunction with other media types (television, radio, print) present content around a vertical industry, sector of interest, or cultural and social issues for a brand. This can be in areas such as health, marketing, or business, or any content relevant to the brand message. Such content is not only provided by the brand but also in the form of consumer-generated media. Research on brand-related user-generated content across major platforms suggests that the form and tone of consumer contributions vary by platform, with promotional content more common on some networks and response-oriented content on others. A brand provides participation with consumers online and offline. This is accomplished through the combination of typical social networking features online, such as personalised pages, friend lists, groups, and messaging, alongside elements of involvement offline. This is not simply connecting an online platform with mobile devices, but providing separate mobile features jointly with a secondary media type to drive online usage and build relationships with the brand on the go. By participating in mobile campaigns, users are interacting with the brand outside of traditional brick and mortar or e-commerce destinations. Empirical work on consumer brand engagement in social media frames such participation along cognitive, affective, and behavioural dimensions. The final element of brand networking involves incentivising participation with the other two elements. The addition of a currency or point system acts as an anchor to the brand and network and creates a competitive dynamic between consumers. These points are distributed for activity carried out outside of the networking site. By incentivising usage offline, the brand image is reinforced for the consumer and strengthens the relationship. Consumers are turned into promoters for both the brand and the users' benefit. The use of points, badges, leaderboards, and similar mechanics is described in the marketing literature as gamification, and has been linked to higher participation rates in mobile and loyalty programmes. == Fan equity == Fan equity is the idea that by locking in consumers to a brand, they are turned into fans of the brand. As fans, they promote, interact, and consume on a daily basis and become assets. Apple Inc. is one example of a company often cited as possessing fan equity. Customers of Apple are extremely brand loyal and are assets to the company. Creating a fan-generated brand is a difficult but effective method of business. Through the use of brand networking, a company is able to build a consumer or fan base that provides a strong relationship between business and consumers. The trust is formed and fans do a lot of work for the brand by word of mouth. Peer-to-peer channels are the strongest means of communication for a brand, but also one in which the brand can only influence and not control. Subsequent research links community engagement with brand trust, identifying community engagement as a mediator between social-media brand community participation and trust. This method of business is argued to be a relationship handled by the brand generally for its own gain. Many fans do not realise the work they are doing for companies by using their product or service. Facebook is a fan-based brand that has become a global phenomenon through customer use, with social media features such as sharing and commenting. With the growth of social media, marketing and advertising through social media has continued to expand. Brands can display and promote their products or services at a fast rate, with consumers sharing and contributing to the brand on a global scale. This can also be seen as online word of mouth exposure that can produce positive or negative feedback for brands. Once consumers become fans they are typically loyal, which can create positive word of mouth for a brand. Fans become a valuable asset, boosting the status and reputation of a brand. Different perceptions of brands can be linked to a person's origin or religion, which creates a difficulty when trying to enter a market or gain market share. Businesses need to be aware of the types of products or services they introduce to a specific market, ensuring they are culturally sensitive. Fan pages are created on social media to maintain the relationship between brands and consumers. By engaging and interacting with consumers, brands obtain fans and produce positive imaging. Some fans become attached to brands and are often encouraged to remain as fans through the use of celebrities endorsing the brand. Research on parasocial interaction in social-media environments suggests that one-sided emotional bonds that consumers form with endorsers and brand personae help convert ordinary followers into engaged fans.

Kimchi (software)

Kimchi is a web management tool to manage Kernel-based Virtual Machine (KVM) infrastructure. Developed with HTML5, Kimchi is developed to intuitively manage KVM guests, create storage pools, manage network interfaces (bridges, VLANs, NAT), and perform other related tasks. The name is an extended acronym for KVM infrastructure management. It is an Apache-licensed project hosted on GitHub, and incubated by oVirt.org.

Industry Dive

Industry Dive is a United States-based business-to-business news organization with an estimated 18 million readers in more than 25 industries, such as banking and waste management. Since 2022, it has been owned by Informa plc. Industry Dive aims to serve business executives who read news on their mobile phones. The company had an estimated revenue of more than of more than $110 million in 2023. As of 2020, it has more than 300 employees, including 80 journalists and 12 engineers. Its headquarters is in Washington, D.C. == History == Industry Dive was formed in 2012 by Sean Griffey (president), Eli Dickinson (chief technology officer), and Ryan Willumson (chief revenue officer). It was funded with $900,000 from private investors in 2012 and 2013. The company covered five industries: construction, education, marketing, utility, and waste. In 2016, it began its Dive Awards. Industry Dive's revenues quadrupled from 2015 to 2018, putting it in the top half of the Deloitte Technology Fast 500 and the top 20 percent of the Inc. Top 5000 list. In 2019, Falfurrias Capital Partners acquired a majority stake in the company. ID's content marketing clients included IBM, Siemens, and UPS. In 2020, DCA Live named Industry Dive to its "Red Hot Companies" list, which recognizes the D.C. area's 'fastest-growing' companies. In the same year, Industry Dive acquired CFO. In 2021, Industry Dive acquired PharmaVOICE. In 2022, it was purchased by Informa plc, which bought its majority stake from Falfurrias Capital Partners for about $530 million. == Publications == Industry Dive provides news coverage of a variety of industries including agriculture, banking, construction, education, fashion, healthcare, and manufacturing, each using a different website: == Awards == Industry Dive publications have received several national and regional Awards of Excellence from the American Society of Business Publication Editors, including for a series of 2020 articles about Big Pharma and the race for the coronavirus vaccine. The Washington Post recognized Industry Dive as a top place to work for four consecutive years, from 2016 to 2020.

XLNet

The XLNet was an autoregressive Transformer designed as an improvement over BERT, with 340M parameters and trained on 33 billion words. It was released on 19 June 2019, under the Apache 2.0 license. It achieved state-of-the-art results on a variety of natural language processing tasks, including language modeling, question answering, and natural language inference. == Architecture == The main idea of XLNet is to model language autoregressively like the GPT models, but allow for all possible permutations of a sentence. Concretely, consider the following sentence:My dog is cute.In standard autoregressive language modeling, the model would be tasked with predicting the probability of each word, conditioned on the previous words as its context: We factorize the joint probability of a sequence of words x 1 , … , x T {\displaystyle x_{1},\ldots ,x_{T}} using the chain rule: Pr ( x 1 , … , x T ) = Pr ( x 1 ) Pr ( x 2 | x 1 ) Pr ( x 3 | x 1 , x 2 ) … Pr ( x T | x 1 , … , x T − 1 ) . {\displaystyle \Pr(x_{1},\ldots ,x_{T})=\Pr(x_{1})\Pr(x_{2}|x_{1})\Pr(x_{3}|x_{1},x_{2})\ldots \Pr(x_{T}|x_{1},\ldots ,x_{T-1}).} For example, the sentence "My dog is cute" is factorized as: Pr ( My , dog , is , cute ) = Pr ( My ) Pr ( dog | My ) Pr ( is | My , dog ) Pr ( cute | My , dog , is ) . {\displaystyle \Pr({\text{My}},{\text{dog}},{\text{is}},{\text{cute}})=\Pr({\text{My}})\Pr({\text{dog}}|{\text{My}})\Pr({\text{is}}|{\text{My}},{\text{dog}})\Pr({\text{cute}}|{\text{My}},{\text{dog}},{\text{is}}).} Schematically, we can write it as → My → My dog → My dog is → My dog is cute . {\displaystyle {\texttt {}}{\texttt {}}{\texttt {}}{\texttt {}}\to {\text{My }}{\texttt {}}{\texttt {}}{\texttt {}}\to {\text{My dog }}{\texttt {}}{\texttt {}}\to {\text{My dog is }}{\texttt {}}\to {\text{My dog is cute}}.} However, for XLNet, the model is required to predict the words in a randomly generated order. Suppose we have sampled a randomly generated order 3241, then schematically, the model is required to perform the following prediction task: is dog is dog is cute → My dog is cute {\displaystyle {\texttt {}}{\texttt {}}{\texttt {}}{\texttt {}}\to {\texttt {}}{\texttt {}}{\text{is }}{\texttt {}}\to {\texttt {}}{\text{dog is }}{\texttt {}}\to {\texttt {}}{\text{dog is cute}}\to {\text{My dog is cute}}} By considering all permutations, XLNet is able to capture longer-range dependencies and better model the bidirectional context of words. === Two-Stream Self-Attention === To implement permutation language modeling, XLNet uses a two-stream self-attention mechanism. The two streams are: Content stream: This stream encodes the content of each word, as in standard causally masked self-attention. Query stream: This stream encodes the content of each word in the context of what has gone before. In more detail, it is a masked cross-attention mechanism, where the queries are from the query stream, and the key-value pairs are from the content stream. The content stream uses the causal mask M causal = [ 0 − ∞ − ∞ … − ∞ 0 0 − ∞ … − ∞ 0 0 0 … − ∞ ⋮ ⋮ ⋮ ⋱ ⋮ 0 0 0 … 0 ] {\displaystyle M_{\text{causal}}={\begin{bmatrix}0&-\infty &-\infty &\dots &-\infty \\0&0&-\infty &\dots &-\infty \\0&0&0&\dots &-\infty \\\vdots &\vdots &\vdots &\ddots &\vdots \\0&0&0&\dots &0\end{bmatrix}}} permuted by a random permutation matrix to P M causal P − 1 {\displaystyle PM_{\text{causal}}P^{-1}} . The query stream uses the cross-attention mask P ( M causal − ∞ I ) P − 1 {\displaystyle P(M_{\text{causal}}-\infty I)P^{-1}} , where the diagonal is subtracted away specifically to avoid the model "cheating" by looking at the content stream for what the current masked token is. Like the causal masking for GPT models, this two-stream masked architecture allows the model to train on all tokens in one forward pass. == Training == Two models were released: XLNet-Large, cased: 110M parameters, 24-layer, 1024-hidden, 16-heads XLNet-Base, cased: 340M parameters, 12-layer, 768-hidden, 12-heads. It was trained on a dataset that amounted to 32.89 billion tokens after tokenization with SentencePiece. The dataset was composed of BooksCorpus, and English Wikipedia, Giga5, ClueWeb 2012-B, and Common Crawl. It was trained on 512 TPU v3 chips, for 5.5 days. At the end of training, it still under-fitted the data, meaning it could have achieved lower loss with more training. It took 0.5 million steps with an Adam optimizer, linear learning rate decay, and a batch size of 8192.

Video game

A video game, computer game, or simply game is an electronic game that involves interaction with a user interface or input device (such as a joystick, controller, keyboard, or motion sensing device) to generate visual feedback from a display device, most commonly shown in a video format on a television set, computer monitor, flat-panel display or touchscreen on handheld devices, or a virtual reality headset. Most modern video games are audiovisual, with audio complement delivered through speakers or headphones, and sometimes also with other types of sensory feedback (e.g., haptic technology that provides tactile sensations). Some video games also allow microphone and webcam inputs for in-game chatting and livestreaming. Video games are typically categorized according to their hardware platform, which traditionally includes arcade video games, console games, and computer games (which includes LAN games, online games, and browser games). More recently, the video game industry has expanded onto mobile gaming through mobile devices (such as smartphones and tablet computers), virtual and augmented reality systems, and remote cloud gaming. Video games are also classified into a wide range of genres based on their style of gameplay and target audience. The first video game prototypes in the 1950s and 1960s were simple extensions of electronic games using video-like output from large, room-sized mainframe computers. The first consumer video game was the arcade video game Computer Space in 1971, which took inspiration from the earlier 1962 computer game Spacewar!. In 1972 came the now-iconic video game Pong and the first home console, the Magnavox Odyssey. The industry grew quickly during the "golden age" of arcade video games from the late 1970s to early 1980s but suffered from the crash of the North American video game market in 1983 due to loss of publishing control and saturation of the market. Following the crash, the industry matured, was dominated by Japanese companies such as Nintendo, Sega, and Sony, and established practices and methods around the development and distribution of video games to prevent a similar crash in the future, many of which continue to be followed. In the 2000s, the core industry centered on "AAA" games, leaving little room for riskier experimental games. Coupled with the availability of the Internet and digital distribution, this gave room for independent video game development (or "indie games") to gain prominence into the 2010s. Since then, the commercial importance of the video game industry has been increasing. The emerging Asian markets and proliferation of smartphone games in particular are altering player demographics towards casual and cozy gaming, and increasing monetization by incorporating games as a service. Today, video game development requires numerous skills, vision, teamwork, and liaisons between different parties, including developers, publishers, distributors, retailers, hardware manufacturers, and other marketers, to successfully bring a game to its consumers. As of 2020, the global video game market had estimated annual revenues of US$159 billion across hardware, software, and services, which is three times the size of the global music industry and four times that of the film industry in 2019, making it a formidable heavyweight across the modern entertainment industry. The video game market is also a major influence behind the electronics industry, where personal computer component, console, and peripheral sales, as well as consumer demands for better game performance, have been powerful driving factors for hardware design and innovation. == Origins == Early video games used interactive electronic devices with various display formats. The earliest example dates to 1947—a "cathode-ray tube amusement device" was filed for a patent on 25 January 1947, by Thomas T. Goldsmith Jr. and Estle Ray Mann, and issued on 14 December 1948, as U.S. Patent 2455992. Inspired by radar display technology, it consisted of an analog device allowing a user to control the parabolic arc of a dot on the screen to simulate a missile being fired at targets, which were paper drawings fixed to the screen. Other early examples include the Nimrod computer at the 1951 Festival of Britain; Christopher Strachey's Checkers, possibly the first game to display visuals on an electronic screen in 1952; OXO, a tic-tac-toe computer game by Alexander S. Douglas for the EDSAC in 1952; Tennis for Two, an electronic interactive game engineered by William Higinbotham in 1958; and Spacewar!, written by Massachusetts Institute of Technology students Martin Graetz, Steve Russell, and Wayne Wiitanen's on a DEC PDP-1 computer in 1962. Each game had different means of display: NIMROD had a panel of lights to play the game of Nim, OXO had a graphical display to play tic-tac-toe, Tennis for Two had an oscilloscope to display a side view of a tennis court, and Spacewar! had the DEC PDP-1's vector display to have two spaceships battle each other. These inventions laid the foundation for modern video games. In 1966, while working at Sanders Associates, Ralph H. Baer devised a system to play a basic table tennis game on a television screen. With the company's approval, Baer created the prototype known as the "Brown Box". Sanders patented Baer's innovations and licensed them to Magnavox, which commercialized the technology as the first home video game console, the Magnavox Odyssey, released in 1972. Separately, Nolan Bushnell and Ted Dabney, inspired by seeing Spacewar! running at Stanford University, devised a similar version running in a smaller coin-operated arcade cabinet using a less expensive computer. This was released as Computer Space, the first arcade video game, in 1971. Bushnell and Dabney went on to form Atari, Inc., and with Allan Alcorn, created their second arcade game in 1972, the hit ping pong-style Pong, which was directly inspired by the table tennis game on the Odyssey. Atari made a home version of Pong, which was released by Christmas 1975. The success of the Odyssey and Pong, both as an arcade game and home machine, launched the video game industry. Both Baer and Bushnell have been titled "Father of Video Games" for their contributions. == Terminology == The term "video game" was developed to describe electronic games played on a video display rather than on a teletype printer, audio speaker, or similar device. This also distinguished from handheld electronic games such as Merlin, which commonly used LED lights for indicators not in combination for imaging purposes. "Computer game" may also be used as a descriptor, as all these types of games essentially require the use of a computer processor; in some cases, it is used interchangeably with "video game". Particularly in the United Kingdom and Western Europe, this is common due to the historic relevance of domestically produced microcomputers. Other terms used include digital game, for example, by the Australian Bureau of Statistics. The term "computer game" can also refer to PC games, which are played primarily on personal computers or other flexible hardware systems, to distinguish them from console games, arcade games, or mobile games. Other terms, such as "television game", "telegame", or "TV game", had been used in the 1970s and early 1980s, particularly for home gaming consoles that rely on connection to a television set. However, these terms were also used interchangeably with "video game" in the 1970s, primarily due to "video" and "television" being synonymous. In Japan, where consoles like the Odyssey were first imported and then made within the country by the large television manufacturers such as Toshiba and Sharp Corporation, such games are known as "TV games", "TV geemu", or "terebi geemu". The term "TV game" is still commonly used into the 21st century. "Electronic game" may also be used to refer to video games, but this also incorporates devices like early handheld electronic games that lack any video output. The first appearance of the term "video game" emerged around 1973. The Oxford English Dictionary cited a 10 November 1973 BusinessWeek article as the first printed use of the term. Though Bushnell believed the term came from a vending magazine review of Computer Space in 1971, a review of the major vending magazines Vending Times and Cashbox showed that the term may have come even earlier, appearing first in a letter dated July 10, 1972. In the letter, Bushnell uses the term "video game" twice. Per video game historian Keith Smith, the sudden appearance suggested that the term had been proposed and readily adopted by those in the field. Around March 1973, Ed Adlum, who ran Cashbox's coin-operated section until 1972 and then later founded RePlay Magazine, covering the coin-op amusement field, in 1975, used the term in an article in March 1973. In a September 1982 issue of RePlay, Adlum is credited with first naming these games as "video games": "RePlay