AI Data Explorer Servicenow — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.
Picture Prowler was an early piece of photo management software developed around and meant to show off Xing Technology's JPEG image decompression library during the early 1990s. Little known today, it featured thumbnail based picture management, printing, etc. The primary developer was Ray Bunnage from compression / decompression libraries developed by Howard Gordon and Chris Eddy.
Read more →
Existential risk from artificial intelligence, or AI x-risk, refers to the idea that substantial progress in artificial general intelligence (AGI) and artificial superintelligence (ASI) could lead to human extinction or an irreversible global catastrophe. One argument for the validity of this concern and the importance of this risk references how human beings dominate other species because the human brain possesses distinctive capabilities other animals lack. If AI were to surpass human intelligence and become superintelligent, it might become uncontrollable. Just as the fate of the mountain gorilla depends on human goodwill, the fate of humanity could depend on the actions of a future machine superintelligence. Experts disagree on whether artificial general intelligence (AGI) can achieve the capabilities needed for human extinction. Debates center on AGI's technical feasibility, the speed of self-improvement, and the effectiveness of alignment strategies. Concerns about superintelligence have been voiced by researchers including Geoffrey Hinton, Yoshua Bengio, Demis Hassabis, and Alan Turing, and AI company CEOs such as Dario Amodei (Anthropic), Sam Altman (OpenAI), and Elon Musk (xAI). In 2022, a survey of AI researchers with a 17% response rate found that the majority believed there is a 10 percent or greater chance that human inability to control AI will cause an existential catastrophe. In 2023, hundreds of AI experts and other notable figures signed a statement declaring, "Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war". Following increased concern over AI risks, government leaders such as United Kingdom prime minister Rishi Sunak and United Nations Secretary-General António Guterres called for an increased focus on global AI regulation. In 2025, hundreds of public figures including AI experts, five Nobel Prize laureates, and former senior US national security officials such as Michael Mullen and Susan Rice signed a statement calling for a ban on the development of superintelligence. Two sources of concern stem from the problems of AI control and alignment. Controlling a superintelligent machine or instilling it with human-compatible values may be difficult. Many researchers believe that a superintelligent machine would likely resist attempts to disable it or change its goals as that would prevent it from accomplishing its present goals. It would be extremely challenging to align a superintelligence with the full breadth of significant human values and constraints. In contrast, skeptics such as computer scientist Yann LeCun argue that superintelligent machines will have no desire for self-preservation. A June 2025 study showed that in some circumstances, models may break laws and disobey direct commands to prevent shutdown or replacement, even at the cost of human lives. Researchers warn that an "intelligence explosion"—a rapid, recursive cycle of AI self-improvement—could outpace human oversight and infrastructure, leaving no opportunity to implement safety measures. In this scenario, an AI more intelligent than its creators would recursively improve itself at an exponentially increasing rate, too quickly for its handlers or society at large to control. Empirically, examples like AlphaZero, which taught itself to play Go and quickly surpassed human ability, show that domain-specific AI systems can sometimes progress from subhuman to superhuman ability very quickly, although such machine learning systems do not recursively improve their fundamental architecture. == History == One of the earliest authors to express serious concern that highly advanced machines might pose existential risks to humanity was the novelist Samuel Butler, who wrote in his 1863 essay Darwin among the Machines: The upshot is simply a question of time, but that the time will come when the machines will hold the real supremacy over the world and its inhabitants is what no person of a truly philosophic mind can for a moment question. In 1951, foundational computer scientist Alan Turing wrote the article "Intelligent Machinery, A Heretical Theory", in which he proposed that artificial general intelligences would likely "take control" of the world as they became more intelligent than human beings: Let us now assume, for the sake of argument, that [intelligent] machines are a genuine possibility, and look at the consequences of constructing them... There would be no question of the machines dying, and they would be able to converse with each other to sharpen their wits. At some stage therefore we should have to expect the machines to take control, in the way that is mentioned in Samuel Butler's Erewhon. In 1965, I. J. Good originated the concept now known as an "intelligence explosion" and said the risks were underappreciated: Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an 'intelligence explosion', and the intelligence of man would be left far behind. Thus the first ultraintelligent machine is the last invention that man need ever make, provided that the machine is docile enough to tell us how to keep it under control. It is curious that this point is made so seldom outside of science fiction. It is sometimes worthwhile to take science fiction seriously. Scholars such as Marvin Minsky and I. J. Good himself occasionally expressed concern that a superintelligence could seize control, but issued no call to action. In 2000, computer scientist and Sun co-founder Bill Joy penned an influential essay, "Why The Future Doesn't Need Us", identifying superintelligent robots as a high-tech danger to human survival, alongside nanotechnology and engineered bioplagues. Nick Bostrom published Superintelligence in 2014, which presented his arguments that superintelligence poses an existential threat. By 2015, public figures such as physicists Stephen Hawking and Nobel laureate Frank Wilczek, computer scientists Stuart J. Russell and Roman Yampolskiy, and entrepreneurs Elon Musk and Bill Gates were expressing concern about the risks of superintelligence. Also in 2015, the Open Letter on Artificial Intelligence highlighted the "great potential of AI" and encouraged more research on how to make it robust and beneficial. In April 2016, the journal Nature warned: "Machines and robots that outperform humans across the board could self-improve beyond our control—and their interests might not align with ours". In 2020, Brian Christian published The Alignment Problem, which details the history of progress on AI alignment up to that time. In March 2023, key figures in AI, such as Musk, signed a letter from the Future of Life Institute calling a halt to advanced AI training until it could be properly regulated. In May 2023, the Center for AI Safety released a statement signed by numerous experts in AI safety and the AI existential risk that read: Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war. A 2025 open letter by the Future of Life Institute, whose signers include five Nobel Prize laureates, reads: We call for a prohibition on the development of superintelligence, not lifted before there is broad scientific consensus that it will be done safely and controllably, and strong public buy-in. == Potential AI capabilities == === General Intelligence === Artificial general intelligence (AGI) is typically defined as a system that performs at least as well as humans in most or all intellectual tasks. A 2022 survey of AI researchers found that 90% of respondents expected AGI would be achieved in the next 100 years, and half expected the same by 2061. In May 2023, some researchers dismissed existential risks from AGI as "science fiction" based on their high confidence that AGI would not be created anytime soon. But in August 2023, a survey of 2,778 AI researchers found that most believed that AGI would be achieved by 2040. Breakthroughs in large language models (LLMs) have led some researchers to reassess their expectations. Notably, Geoffrey Hinton said in 2023 that he recently changed his estimate from "20 to 50 years before we have general purpose A.I." to "20 years or less". === Superintelligence === In contrast with AGI, Bostrom defines a superintelligence as "any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest", including scientific creativity, strategic planning, and social skills. He argues that a superintelligence can outmaneuver humans anytime its goals conflict with humans'. It may choose to hide its true intent until humanity cannot stop it. Bostrom writes that in order to be safe for
Read more →
WordNet is a lexical database of semantic relations between words that links words into semantic relations including synonyms, hyponyms, and meronyms. The synonyms are grouped into synsets with short definitions and usage examples. It can thus be seen as a combination and extension of a dictionary and thesaurus. Its primary use is in automatic text analysis and artificial intelligence applications. It was first created in the English language and the English WordNet database and software tools have been released under a BSD style license and are freely available for download. The latest official release from Princeton was released in 2011. Princeton currently has no plans to release any new versions due to staffing and funding issues. New versions are still being released annually through the Open English WordNet website. Until about 2024 an online version was previously available through wordnet.princeton.edu. That version of WordNet has been deprecated, but a new online version is available at en-word.net. There are now WordNets in more than 200 languages. == History and team members == WordNet was first created in 1985, in English only, in the Cognitive Science Laboratory of Princeton University under the direction of psychology professor George Armitage Miller. It was later directed by Christiane Fellbaum. The project was initially funded by the U.S. Office of Naval Research, and later also by other U.S. government agencies including the DARPA, the National Science Foundation, the Disruptive Technology Office (formerly the Advanced Research and Development Activity) and REFLEX. George Miller and Christiane Fellbaum received the 2006 Antonio Zampolli Prize for their work with WordNet. The Global WordNet Association is a non-commercial organization that provides a platform for discussing, sharing and connecting WordNets for all languages in the world. Christiane Fellbaum and Piek Th.J.M. Vossen are its co-presidents. == Database contents == The database contains 155,327 words organized in 175,979 synsets for a total of 207,016 word-sense pairs; in compressed form, it is about 12 megabytes in size. It includes the lexical categories nouns, verbs, adjectives and adverbs but ignores prepositions, determiners and other function words. Words from the same lexical category that are roughly synonymous are grouped into synsets, which include simplex words as well as collocations like "eat out" and "car pool." The different senses of a polysemous word form are assigned to different synsets. A synset's meaning is further clarified with a short defining gloss and one or more usage examples. An example adjective synset is: good, right, ripe – (most suitable or right for a particular purpose; "a good time to plant tomatoes"; "the right time to act"; "the time is ripe for great sociological changes") All synsets are connected by means of semantic relations. These relations, which are not all shared by all lexical categories, include: Nouns hypernym: Y is a hypernym of X if every X is a (kind of) Y (canine is a hypernym of dog) hyponym: Y is a hyponym of X if every Y is a (kind of) X (dog is a hyponym of canine) coordinate term: Y is a coordinate term of X if X and Y share a hypernym (wolf is a coordinate term of dog, and dog is a coordinate term of wolf) holonym: Y is a holonym of X if X is a part of Y (building is a holonym of window) meronym: Y is a meronym of X if Y is a part of X (window is a meronym of building) Verbs hypernym: the verb Y is a hypernym of the verb X if the activity X is a (kind of) Y (to perceive is an hypernym of to listen) troponym: the verb Y is a troponym of the verb X if the activity Y is doing X in some manner (to lisp is a troponym of to talk) entailment: the verb Y is entailed by the verb X if by doing X you must be doing Y (to sleep is entailed by to snore) coordinate term: the verb Y is a coordinate term of the verb X if X and Y share a hypernym (to lisp is a coordinate term of to yell, and to yell is a coordinate term of to lisp) These semantic relations hold among all members of the linked synsets. Individual synset members (words) can also be connected with lexical relations. For example, (one sense of) the noun "director" is linked to (one sense of) the verb "direct" from which it is derived via a "morphosemantic" link. The morphology functions of the software distributed with the database try to deduce the lemma or stem form of a word from the user's input. Irregular forms are stored in a list, and looking up "ate" will return "eat," for example. == Knowledge structure == Both nouns and verbs are organized into hierarchies, defined by hypernym or IS A relationships. For instance, one sense of the word dog is found following hypernym hierarchy; the words at the same level represent synset members. Each set of synonyms has a unique index. At the top level, these hierarchies are organized into 25 beginner "trees" for nouns and 15 for verbs (called lexicographic files at a maintenance level). All are linked to a unique beginner synset, "entity". Noun hierarchies are far deeper than verb hierarchies. Adjectives are not organized into hierarchical trees. Instead, two "central" antonyms such as "hot" and "cold" form binary poles, while 'satellite' synonyms such as "steaming" and "chilly" connect to their respective poles via a "similarity" relations. The adjectives can be visualized in this way as "dumbbells" rather than as "trees". == Psycholinguistic aspects == The initial goal of the WordNet project was to build a lexical database that would be consistent with theories of human semantic memory developed in the late 1960s. Psychological experiments indicated that speakers organized their knowledge of concepts in an economic, hierarchical fashion. Retrieval time required to access conceptual knowledge seemed to be directly related to the number of hierarchies the speaker needed to "traverse" to access the knowledge. Thus, speakers could more quickly verify that canaries can sing because a canary is a songbird, but required slightly more time to verify that canaries can fly (where they had to access the concept "bird" on the superordinate level) and even more time to verify canaries have skin (requiring look-up across multiple levels of hyponymy, up to "animal"). While such psycholinguistic experiments and the underlying theories have been subject to criticism, some of WordNet's organization is consistent with experimental evidence. For example, anomic aphasia selectively affects speakers' ability to produce words from a specific semantic category, a WordNet hierarchy. Antonymous adjectives (WordNet's central adjectives in the dumbbell structure) are found to co-occur far more frequently than chance, a fact that has been found to hold for many languages. == As a lexical ontology == WordNet is sometimes called an ontology, a persistent claim that its creators do not make. The hypernym/hyponym relationships among the noun synsets can be interpreted as specialization relations among conceptual categories. In other words, WordNet can be interpreted and used as a lexical ontology in the computer science sense. However, such an ontology should be corrected before being used, because it contains hundreds of basic semantic inconsistencies; for example there are, (i) common specializations for exclusive categories and (ii) redundancies in the specialization hierarchy. Furthermore, transforming WordNet into a lexical ontology usable for knowledge representation should normally also involve (i) distinguishing the specialization relations into subtypeOf and instanceOf relations, and (ii) associating intuitive unique identifiers to each category. Although such corrections and transformations have been performed and documented as part of the integration of WordNet 1.7 into the cooperatively updatable knowledge base of WebKB-2, most projects claiming to reuse WordNet for knowledge-based applications (typically, knowledge-oriented information retrieval) simply reuse it directly. WordNet has also been converted to a formal specification, by means of a hybrid bottom-up top-down methodology to automatically extract association relations from it and interpret these associations in terms of a set of conceptual relations, formally defined in the DOLCE foundational ontology. In most works that claim to have integrated WordNet into ontologies, the content of WordNet has not simply been corrected when it seemed necessary; instead, it has been heavily reinterpreted and updated whenever suitable. This was the case when, for example, the top-level ontology of WordNet was restructured according to the OntoClean-based approach, or when it was used as a primary source for constructing the lower classes of the SENSUS ontology. == Limitations == The most widely discussed limitation of WordNet (and related resources like ImageNet) is that some of the semantic relations are more suited to concrete concepts than to abstract concepts. For example,
Read more →
The DARPA AlphaDogfight was a 2019–2020 DARPA program that pitted computers using F-16 flight simulators against one another. The computers were managed by eight teams of humans, who competed in a single-round elimination for the right to battle a skilled human dogfighter. Heron Systems corporation wrote a deep reinforcement learning software tool that bested the human pilot by a score of 5–0. The tournament program was managed by the Applied Physics Laboratory. The trials took place in October 2019 and January 2020 while the finals were held in August 2020. In 2024 a successor version of the program was tested with in the physical world with the X-62A.
Read more →
Class activation mapping methods are explainable AI (XAI) techniques used to visualize the regions of an input image that are the most relevant for a particular task, especially image classification, in convolutional neural networks (CNNs). These methods generate heatmaps by weighting the feature maps from a convolutional layer according to their relevance to the target class. In the field of artificial intelligence, generically defined as "the effort to automate intellectual tasks normally performed by humans", machine learning and deep learning were created. They both use statistical and computational methods to learn patterns from data, reducing the need for manually coded rules. Machine learning models are trained on input data and the known respective answers, learning the underlying patterns or structures present in the data. Traditional Machine learning algorithms employ manually designed feature sets, posing a direct link between machine learning designers and employed features. Deep learning is a subfield of machine learning, based on the concept of successive layers of representation, in which the data is progressively unfolded in different ways, to extract relevant and informative patterns in data analysis. Deep learning algorithms are defined as feature learning algorithms automatically learning hierarchical feature representations from raw data, extracting increasingly abstract features through multiple layers. CNNs are a specific architecture of deep learning models, designed to process spatially structured data, such as images, exploiting a series of convolution, non-linear activation and pooling operations to extract relevant features, contained in the so-called feature maps from input data. CNNs have demonstrated to be highly effective in a variety of computer vision and image processing tasks. CNNs (and deep learning models more broadly) are described as black boxes due to their complex and non-transparent internal layers of representation. The need for clearer indications on its internal working and decision-making process gave birth to XAI techniques. Among the proposed XAI techniques for computer vision tasks, Class activation mapping methods can show which pixels in an input image are important to the predicted logit for a class of interest, in a classification task. Class activation mapping methods were originally developed for class-discriminative scenarios to visualize which parts of the input image influenced the classification decision, namely to visually highlight the regions of those feature maps that contribute most strongly to the prediction of a given class. More advanced versions of these methods are not limited to image classification tasks, but have been extended also to several vision-related tasks, such as object detection, image captioning, visual question answering and image segmentation. == Background == The following methods laid the groundwork for the class activation maps approaches, forming the conceptual basis of using gradients to highlight class-discriminative regions. === Class model visualization and saliency maps for convolutional neural networks === The class model visualization and image-specific saliency maps approaches have been presented in the foundational work "Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps" by Karen Simonyan, Andrea Vedaldi, and Andrew Zisserman and it generalizes the deconvnet method by Zeiler and Fergus. Class model visualization synthesizes an artificial input image that strongly activates the output neurons associated with a target class. Given a trained, fixed model, this method starts with a zero-initialized image, backpropagates the gradients from the class score to the image pixels, updates the image pixels increasing the specific class scores and it repeats the pixel updating process, showing an encoded (idealized version) prototype of the class of interest. Image-specific class saliency visualization method provides a visual explanation by highlighting the most relevant pixels in an image for predicting a certain class C of interest. This is done by computing the gradient of the class score with respect to the input image, I 0 , {\displaystyle I_{0},} w = ∂ S C ∂ I | I 0 {\displaystyle w=\left.{\frac {\partial S_{C}}{\partial I}}\right|_{I_{0}}} approximating the model locally (around I 0 {\displaystyle I_{0}} ) as linear, using a first-order Taylor expansion: S C ( I ) ≈ w C T I + b {\displaystyle S_{C}(I)\approx w_{C}^{T}I+b} . The magnitude of w C {\displaystyle w_{C}} , the gradient, indicates the importancy of the pixels: larger gradients suggest greater influence on the prediction. Once the gradient is known, the saliency map is defined as the maximum absolute gradient across the color channels: M i j = m a x C | ∂ S C ∂ I i j C | {\displaystyle M_{ij}=max_{C}\left|{\frac {\partial S_{C}}{\partial I_{ij}^{C}}}\right|} resulting in an saliency map (i.e. heatmap). === Guided backpropagation === The concept of guided backpropagation can be traced for the first time in the paper by Springenberg et al. "Striving For Simplicity: The All Convolutional Net" and also this method builds upon the work by Zeiler and Fergus "Visualizing and Understanding Convolutional Networks". Guided backpropagation core is to understand what a CNN is learning, by visualizing the patterns that activate more strongly individual neurons (or filters), in architectures which do not rely on max-pooling layer. When propagating gradients back through a rectified linear unit (ReLU), guided backpropagation passes the gradient if and only if the input to the ReLU was positive (forward pass) and the output gradient is positive (backward signal), tackling both inactive neurons, negative gradients and suppressing the noise. The result displays sharper, high-resolution visualizations of what each neuron is responding to. Guided backpropagation represents a simple and practical method for model interpretability, helping understand how and where neural networks detect semantic concepts across layers. Moreover, it can be applied to any network architecture, due to its working principle. == Base versions == Class activation mapping and gradient-weighted class activation mapping are the original and most widely used methods for visual explanations in convolutional neural networks. These methods serve as the foundation for many later developments in explainable AI. Notation: In this article, the symbols i and j represent integer indices that disappear inside sums or averages, while x and y are the continuous (or up-sampled integer) coordinates of the final heat-map that is plotted. === Class activation mapping (CAM) === Class activation mapping (CAM) was the first, and the original, version of CAM methods, and it gave the name to the whole category. The approach was firstly introduced by Zhou et al. in their seminal work "Learning Deep Features for Discriminative Localization". This approach achieves class-specific heatmaps by modifying image classification CNN architectures, replacing fully-connected layers with convolutional layers and a final global average pooling layer. Its main scope is to localize and highlight discriminative regions of an input image that a CNN uses to identify a particular class, without needing explicit bounding box annotations. ==== Global average pooling (GAP) ==== Global average pooling (GAP) represents the key element in the original CAM approach. It is a dimensionality reduction technique and, similarly to other pooling layers, it allows the downsampling of the feature maps, calculating representative values for a specific region of the feature map. The particularity of GAP is that it calculates a single value for an entire feature map, significantly reducing the model dimensions. ==== Mathematical description ==== The mathematical description considers as its key the combination of convolutional and GAP layers. In CAM, it is mandatory to have the GAP layer after the last convolutional layer and before the final linear classifier layer. This last element of the architecture connects the output logits (the network predictions) y C {\displaystyle y^{C}} , to the GAP values, with its respective fine-tuned weights, w k C {\displaystyle w_{k}^{C}} . Considering A k {\displaystyle A^{k}} as the last feature maps of the last convolutional layer, GAP produces one value for each feature map, by averaging all the matrix elements (i, j) of the feature map: F k = 1 m n ∑ i = 1 m ∑ j = 1 n A i j k {\displaystyle F^{k}={\frac {1}{mn}}\sum _{i=1}^{m}\sum _{j=1}^{n}A_{ij}^{k}} with A k = [ A 11 k A 12 k ⋯ A 1 n k A 21 k A 22 k ⋯ A 2 n k ⋮ ⋮ ⋱ ⋮ A m 1 k A m 2 k ⋯ A m n k ] = { A i j k ∣ 1 ≤ i ≤ m , 1 ≤ j ≤ n } {\displaystyle A^{k}={\begin{bmatrix}A_{11}^{k}&A_{12}^{k}&\cdots &A_{1n}^{k}\\A_{21}^{k}&A_{22}^{k}&\cdots &A_{2n}^{k}\\\vdots &\vdots &\ddots &\vdots \\A_{m1}^{k}&A_{m2}^{k}&\cdots &A_{mn}^{k}\end{bmatrix}}=\left\{A_{
Read more →
Liang Wenfeng (Chinese: 梁文锋; pinyin: Liáng Wénfēng; born 1985) is a Chinese entrepreneur and businessman who is the co-founder of the quantitative hedge fund High-Flyer, as well as the founder and CEO of its artificial intelligence company DeepSeek. Liang attended Zhejiang University, and began his career by applying machine learning methods to quantitative finance. Through High-Flyer, he built large-scale computing infrastructure that was later used to support artificial intelligence research, leading to the creation of DeepSeek in 2023. DeepSeek gained international attention following the release of DeepSeek-R1, which analysts described as demonstrating high-level performance with comparatively limited compute resources. In 2025, Liang was named to Time magazine's list of 100 Most Influential People in AI and Fortune's list of the Most Powerful People in Business. == Early life == Liang was born in 1985 in the village of Mililing (米历岭村), Qinba town (覃巴镇), Wuchuan city (吴川市), Guangdong. His parents were both primary school teachers. Liang was routinely praised by both locals and teachers alike. Even since middle school, Liang was recalled for being well-known for reading comic books, while also being very proficient in mathematics. == Education == After elementary school, Liang attended Wuchuan No. 1 Middle School. There, he quickly excelled in class and ranked highly amongst his peers. He taught himself high school and university-level mathematics courses. Liang then attended Wuchaun No. 1 High School. In these years, he developed hobbies of mathematical modeling and conducting research projects. Compared to his peers, he was always ranked highly. For every mathematics exam, he always ranked within the top three. He was also the top scorer in the Zhanjiang region of Guangdong for the college entrance exam. Thus, in 2002, Liang left high school early to further pursue his education at the university level at the young age of 17. Attending Zhejiang University at the age of 17, Liang earned a Bachelor of Engineering in Electronic Information Engineering in 2007 and his Master of Engineering in Information & Communication Engineering in 2010. His master's dissertation was titled "Study on Object Tracking Algorithm Based on Low-Cost PTZ camera" (基于低成本PTZ摄像机的目标跟踪算法研究). In his college years, DJI founder Wang Tao asked Liang to join as a co-founder. Liang declined the invitation to pursue artificial intelligence methodologies in financial markets. While he states that those around him had entrepreneurial mindsets, he himself valued academics. == Career == === Early career (2008–2016) === During the 2008 financial crisis, Liang formed a team with his classmates to accumulate data related to financial markets. He also led the team to explore quantitative trading using machine learning and other technologies. After his graduation, Liang moved to a cheap flat in Chengdu, Sichuan, where he experimented with ways to apply AI to various fields. These ventures failed, until he tried applying AI to finance. In 2013, Liang attempted to integrate artificial intelligence with quantitative trading and founded Hangzhou Yakebi Investment Management Co Ltd with Xu Jin, an alumnus of Zhejiang University. In 2015, they co-founded Hangzhou Huanfang Technology Co Ltd, which is today's Zhejiang Jiuzhang Asset Management Co Ltd. === High-Flyer (2016–2023) === In February 2016, Liang and two other engineering classmates co-founded Ningbo High-Flyer Quantitative Investment Management Partnership (Limited Partnership). The team relied on mathematics and AI to make investments. Much of the early startup culture was described by former employees to be "geeky" and "quirky," often seen as contrary to the existing culture in large Chinese tech companies. In 2019, Liang founded High-Flyer AI which was dedicated to research on AI algorithms and its basic applications. By this time, High-Flyer had over 10 billion yuan in assets under management. On 30 August 2019, Liang Wenfeng delivered a keynote speech entitled "The Future of Quantitative Investment in China from a Programmer's Perspective" at the Private Equity Golden Bull Award ceremony held by China Securities Journal, and sparked heated discussions. Liang stated that the criterion for determining what is quantitative or non-quantitative is whether the investment decision is made by quantitative methods or by people. Quantitative funds do not have portfolio managers making the decisions and instead are just servers. He also stated High-Flyer's mission is to improve the effectiveness of China's secondary market. In February 2021, Gregory Zuckerman's book The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution was published. Liang wrote the preface for the Chinese edition of the book where he stated that whenever he encountered difficulties at work, he would think of Simons' words "There must be a way to model prices". In January 2025, Zuckerman wrote in The Wall Street Journal where he acknowledged this fact and stated he has been trying to get in touch with Liang but much like Simons, Liang is very secretive and difficult to contact. During 2021, Liang started buying thousands of Nvidia GPUs for his AI side project while running High-Flyer. Liang wanted to build something and it will be a game changer which his business partners thought was only possible from giants such as ByteDance and Alibaba Group. === DeepSeek (since 2023) === ==== DeepSeek begins ==== In May 2023, Liang announced High-Flyer would pursue the development of artificial general intelligence and launched DeepSeek. During that month in an interview with 36Kr, Liang stated that High-Flyer had acquired 10,000 Nvidia A100 GPUs before the US government imposed AI chip restrictions on China. That laid the foundation for DeepSeek to operate as an LLM developer. Liang also stated DeepSeek gets funding from High-Flyer. This was because when DeepSeek was founded, venture capital firms were reluctant in providing funding as it was unlikely that it would be able to generate an exit in a short period of time. Liang only personally holds 1% of the company, with 99% of the company being held by Ningbo High-Flyer Quantitative Investment Management Partnership (Limited Partnership). With DeepSeek's funding model, it lacks commercial pressure and rigid key performance indicators, enabling the company to deviate from previously established model architectures. ==== Early development ==== In July 2024, Liang was interviewed again by 36Kr. He stated that when DeepSeek-V2 was released and triggered an AI price war in China, it came as a huge surprise as the team did not expect pricing to be so sensitive. Liang's aggressive pricing of the language model forced domestic tech giants including Alibaba and Baidu to cut their own rates by over 95%. He also stated that as China's economy develops, it should gradually become a contributor instead of freeriding. What is lacking in China's innovation is not capital but a lack of confidence and knowledge on organizing talent into it. DeepSeek has not hired anyone particularly special and employees tend to be locally educated. When it comes to disruptive technologies, closed source approaches can only temporarily delay others in catching up. As the goal was long-term, DeepSeek sought employees who had ability and passion rather than experience. To retain a high talent density relative to larger firms like Bytedance or Baidu, DeepSeek aimed to maintain a low-hierarchy corporate culture, with members working in project-based groups, as well as competitive compensation. Liang emphasized his vision for DeepSeek employees to bring their "unique experience and ideas" instead of needing to be explicitly directed, with an overall bottom-up approach to division of labor. Liang noted that a significant outcome of this approach was the multi-head latent attention training architecture, which was attributed directly to a young DeepSeek researcher's personal interest. This advancement played a core role in reducing the cost of training the DeepSeek-V3 model, released in December 2024. ==== Release of DeepSeek-R1 ==== Also on 20 January 2025, DeepSeek, the company Liang founded and served as the CEO, released DeepSeek-R1, a 671-billion-parameter open-source reasoning AI model, alongside the publication of a detailed technical paper explaining its architecture and training methodology. The model was built using just 2,048 Nvidia H800 GPUs at a cost of $5.6 million, showcasing a resource-efficient approach that contrasted sharply with the billion-dollar budgets of Western competitors. The development of DeepSeek-R1 occurred amidst U.S. sanctions where Trump limited sales of Nvidia chips to China. By 27 January, DeepSeek surpassed ChatGPT to become the #1 free app on the United States iOS App Store. U.S. stocks plummeted, as more than $1 trillion was erased in market capitalization amid panic over DeepSeek. Technology journ
Read more →
Effective accelerationism (e/acc) is a 21st-century ideological movement that advocates for an explicitly pro-technology stance. Its proponents believe that unrestricted technological progress, especially driven by artificial intelligence, is a solution to universal human problems, such as poverty, war, and climate change. They perceive themselves as a counterweight to more cautious views on technological innovation and often label their opponents derogatorily as "doomers" or "decels" (short for decelerationists). The movement carries utopian undertones and advocates for faster AI progress to ensure human survival and propagate consciousness throughout the universe. Although effective accelerationism has been described as a fringe movement and as cult-like, it has gained mainstream visibility in 2023. A number of high-profile Silicon Valley figures, including investors Marc Andreessen and Garry Tan, explicitly endorsed it by adding "e/acc" to their public social media profiles. == Etymology and central beliefs == Effective accelerationism, a portmanteau of "effective altruism" and "accelerationism", is a fundamentally techno-optimist movement. According to Guillaume Verdon, one of the movement's founders, its aim is for human civilization to "clim[b] the Kardashev gradient", meaning its purpose is for human civilization to rise to next levels on the Kardashev scale by maximizing energy usage. To achieve this goal, effective accelerationism wants to accelerate technological progress. It is strongly focused on artificial general intelligence (AGI), because it sees AGI as fundamental for climbing the Kardashev scale. The movement therefore advocates for unrestricted development and deployment of artificial intelligence. Regulation of artificial intelligence and government intervention in markets more generally is met with opposition. Many of its proponents have libertarian views and think that AGI will be most aligned if many AGIs compete against each other on the marketplace. The founders of the movement see it as rooted in Jeremy England's theory on the origin of life, which is focused on entropy and thermodynamics. According to them, the universe aims to increase entropy, and life is a way of increasing it. By spreading life throughout the universe and making life use up ever increasing amounts of energy, the universe's purpose would thus be fulfilled. == History == === Intellectual origins === While Nick Land is seen as the intellectual originator of contemporary accelerationism in general, the precise origins of effective accelerationism remain unclear. The earliest known reference to the movement can be traced back to a May 2022 newsletter published by four pseudonymous authors known by their X (formerly Twitter) usernames @BasedBeffJezos, @bayeslord, @zestular and @creatine_cycle. Effective accelerationism is an extension of the TESCREAL movement, being etymologically derived from Effective Altruism and heavily rooted in the older Silicon Valley subcultures of transhumanism and extropianism (which similarly emphasized the value of progress and resisted efforts to restrain the development of technology), alongside elements of singularitarianism, cosmism, and longtermism. It is also often considered to have emerged at least in part from the work of the Cybernetic Culture Research Unit (of which Nick Land was a leading member, alongside writers such as Mark Fisher and Sadie Plant). It is sometimes compared and contrasted with the work of philosopher Benjamin Bratton on planetary computation. === Disclosure of the identity of BasedBeffJezos === Forbes disclosed in December 2023 that the @BasedBeffJezos persona is maintained by Guillaume Verdon, a Canadian former Google quantum computing engineer and theoretical physicist. The revelation was supported by a voice analysis conducted by the National Center for Media Forensics of the University of Colorado Denver, which further confirmed the match between Jezos and Verdon. The magazine justified its decision to disclose Verdon's identity on the grounds of it being "in the public interest". On 29 December 2023 Guillaume Verdon was interviewed by Lex Fridman on the Lex Fridman Podcast and introduced as the "creator of the effective accelerationism movement". === Second Trump presidency === Following Donald Trump's victory in the 2024 U.S. presidential election, several prominent tech industry figures expressed support for positions aligned with effective accelerationism, particularly regarding deregulation and technological advancement. The potential appointment of Elon Musk to government roles focused on auditing federal programs drew support from venture capitalists who anticipated reduced regulatory oversight of the technology sector. Notable tech figures publicly connected these developments to the movement's principles. Aaron Levie, CEO of Box, expressed support for "removing unnecessary red tape and over-regulation", while Mark Pincus, early Facebook investor and Zynga founder, explicitly referenced "effective accelerationism" in his post-election commentary. Venture capitalists viewed the incoming administration as an opportunity to ease regulations that had affected technology mergers and acquisitions during the previous years. == Relation to other movements == === Traditional accelerationism === Traditional accelerationism, as developed by the British philosopher Nick Land, sees the acceleration of technological change as a way to bring about a fundamental transformation of current culture, society, and the political economy. This is done through capitalism, which Land views as "an autonomous force that’s reconfiguring society" that can overcome its limits if intensified. Land's work has also been characterized as concerning "the supposedly inevitable 'disintegration of the human species' when artificial intelligence improves sufficiently." While both concern ideas like a technocapital singularity and AGI progress, effective accelerationism focuses on using AGI for the greatest ethical good for conscious life and civilization (whether human or machine), as well as expanding civilization and maximizing energy usage in order to align with the "will of the universe". Land focuses on capitalist self-optimization as the driver of modernity, progress, and the eroding of existing social orders. Land has expressed support for effective accelerationism, while Thomas Murphy referred to the movement as "Nick Land diluted for LinkedIn". === Effective altruism === Effective accelerationism diverges from the principles of effective altruism, which prioritizes using evidence and reasoning to identify the most effective ways to altruistically improve the world. This divergence comes primarily from one of the causes effective altruists focus on – AI existential risk. Effective altruists (particularly longtermists) argue that AI companies should be cautious and strive to develop safe AI systems, as they fear that any misaligned AGI could eventually lead to human extinction. Proponents of effective accelerationism generally consider existential risks from AGI to be negligible, and claim that even if they were not, decentralized free markets would much better mitigate this risk than centralized governmental regulation. === Degrowth === Effective accelerationism stands in stark contrast with the degrowth movement, sometimes described by it as "decelerationism" or "decels". The degrowth movement advocates for reducing economic activity and consumption to address ecological and social issues. Effective accelerationism on the contrary embraces technological progress, energy consumption and the dynamics of capitalism, rather than advocating for a reduction in economic activity. == Reception == The "Techno-Optimist Manifesto", a 2023 essay by Marc Andreessen, has been described by the Financial Times and the German Süddeutsche Zeitung as espousing the views of effective accelerationism. Mother Jones also characterized it as expressing effective accelerationism and reported that Andressen cited Land's work. David Swan of The Sydney Morning Herald has criticized effective accelerationism due to its opposition to government and industry self-regulation. He argues that "innovations like AI needs thoughtful regulations and guardrails ... to avoid the myriad mistakes Silicon Valley has already made." During the 2023 Reagan National Defense Forum, U.S. Secretary of Commerce Gina Raimondo cautioned against embracing the "move fast and break things" mentality associated with "effective acceleration [sic]". She emphasized the need to exercise caution in dealing with AI, stating "that's too dangerous. You can't break things when you are talking about AI." In a similar vein, Ellen Huet argued on Bloomberg News that some of the ideas of the movement were "deeply unsettling", focusing especially on Guillaume Verdon's "post-humanism" and the view that "natural selection could lead AI to replace us as the dominant spe
Read more →
PyTorch is an open-source deep learning library, originally developed by Meta Platforms and currently developed with support from the Linux Foundation. The successor to Torch, PyTorch provides a high-level API that builds upon optimised, low-level implementations of deep learning algorithms and architectures, such as the Transformer, or SGD. Notably, this API simplifies model training and inference to a few lines of code. PyTorch allows for automatic parallelization of training and, internally, implements CUDA bindings that speed training further by leveraging GPU resources. PyTorch utilises the tensor as a fundamental data type, similarly to NumPy. Training is facilitated by a reversed automatic differentiation system, Autograd, that constructs a directed acyclic graph of the operations (and their arguments) executed by a model during its forward pass. With a loss, backpropagation is then undertaken. As of 2025, PyTorch remains one of the most popular deep learning libraries, alongside others such as TensorFlow and Keras. It can be installed using Anaconda package managers. A number of commercial deep learning architectures are built on top of PyTorch, including ChatGPT, Tesla Autopilot, Uber's Pyro, and Hugging Face's Transformers. == History == In 2001, Torch was written and released under a GPL. It was a machine-learning library written in C++ and CUDA, supporting methods including neural networks, support vector machines (SVM), hidden Markov models, etc. Around 2010, it was rewritten by Ronan Collobert, Clement Farabet and Koray Kavuckuoglu. This was known as Torch7 or LuaTorch. This was written so that the backend was in C and the frontend was in Lua. In mid-2016, some developers refactored it to decouple the frontend and the backend, with strong influence from torch-autograd and Chainer. In turn, torch-autograd was influenced by HIPS/autograd. Development on Torch7 ceased in 2018 and was subsumed by the PyTorch project. Meta (formerly known as Facebook) operates both PyTorch and Convolutional Architecture for Fast Feature Embedding (Caffe2), but models defined by the two frameworks were mutually incompatible. The Open Neural Network Exchange (ONNX) project was created by Meta and Microsoft in September 2017 to decouple deep learning frameworks from hardware-specific runtimes, allowing models to be converted between frameworks and optimized for execution providers like NVIDIA’s TensorRT. Caffe2 was merged into PyTorch at the end of March 2018. In September 2022, Meta announced that PyTorch would be governed by the independent PyTorch Foundation, a newly created subsidiary of the Linux Foundation. PyTorch 2.0 was released on 15 March 2023, introducing TorchDynamo, a Python-level compiler that makes code run up to two times faster, along with significant improvements in training and inference performance across major cloud platforms. == PyTorch tensors == PyTorch defines a class called Tensor (torch.Tensor) to store and operate on homogeneous multidimensional rectangular arrays of numbers. PyTorch supports various sub-types of multi-dimensional arrays, or Tensors. PyTorch Tensors are similar to NumPy Arrays, but can also be operated on by a CUDA-capable NVIDIA GPU. PyTorch has also been developing support for other GPU platforms, for example, AMD's ROCm and Apple's Metal Framework. == PyTorch neural networks == PyTorch defines a module called nn (torch.nn) to describe neural networks and to support training. This module offers a comprehensive collection of building blocks for neural networks, including various layers and activation functions, enabling the construction of complex models. Networks are built by inheriting from the torch.nn module and defining the sequence of operations in the forward() function. == PyTorch Serialized File Format == Pytorch can save and load models using its own file format, which is a ZIP64 archive containing the model weights in a Python pickle file, and other information such as the byte order. The file extensions .pt and .pth are commonly used for these files. == Example == The following program shows the low-level functionality of the library with a simple example. The following code block defines a neural network with linear layers using the nn module.
Read more →
Color management is the process of ensuring consistent and accurate colors across various devices, such as monitors, printers, and cameras. It involves the use of color profiles, which are standardized descriptions of how colors should be displayed or reproduced. Color management is necessary because different devices have different color capabilities and characteristics. For example, a monitor may display colors differently than a printer can reproduce them. Without color management, the same image may appear differently on different devices, leading to inconsistencies and inaccuracies. To achieve color management, a color profile is created for each device involved in the color workflow. This profile describes the device's color capabilities and characteristics, such as its color gamut (range of colors it can display or reproduce) and color temperature. These profiles are then used to translate colors between devices, ensuring consistent and accurate color reproduction. Color management is particularly important in industries such as graphic design, photography, and printing, where accurate color representation is crucial. It helps to maintain color consistency throughout the entire workflow, from capturing an image to displaying or printing it. Parts of color management are implemented in the operating system (OS), helper libraries, the application, and devices. The type of color profile that is typically used is called an ICC profile. A cross-platform view of color management is the use of an ICC-compatible color management system. The International Color Consortium (ICC) is an industry consortium that has defined: an open standard for a Color Matching Module (CMM) at the OS level color profiles for: devices, including DeviceLink profiles that transform one device profile (color space) to another device profile without passing through an intermediate color space, such as LAB, more accurately preserving color working spaces, the color spaces in which color data is meant to be manipulated There are other approaches to color management besides using ICC profiles. This is partly due to history and partly because of other needs than the ICC standard covers. The film and broadcasting industries make use of some of the same concepts, but they frequently rely on more limited boutique solutions. The film industry, for instance, often uses 3D LUTs (lookup table) to represent a complete color transformation for a specific RGB encoding. At the consumer level, system wide color management is available in most of Apple's products (macOS, iOS, iPadOS, watchOS). Microsoft Windows lacks system wide color management and virtually all applications do not employ color management. Windows' media player API is not color space aware, and if applications want to color manage videos manually, they have to incur significant performance and power consumption penalties. Android supports system wide color management, but most devices ship with color management disabled. == Overview == Characterize. Every color-managed device requires a personalized table, or "color profile," which characterizes the color response of that particular device. Standardize. Each color profile describes these colors relative to a standardized set of reference colors (the "Profile Connection Space"). Translate. Color-managed software then uses these standardized profiles to translate color from one device to another. This is usually performed by a color management module (CMM). == Hardware == === Characterization === To describe the behavior of various output devices, they must be compared (measured) in relation to a standard color space. Often a step called linearization is performed first, to undo the effect of gamma correction that was done to get the most out of limited 8-bit color paths. Instruments used for measuring device colors include colorimeters and spectrophotometers. As an intermediate result, the device gamut is described in the form of scattered measurement data. The transformation of the scattered measurement data into a more regular form, usable by the application, is called profiling. Profiling is a complex process involving mathematics, intense computation, judgment, testing, and iteration. After the profiling is finished, an idealized color description of the device is created. This description is called a profile. === Calibration === Calibration is like characterization, except that it can include the adjustment of the device, as opposed to just the measurement of the device. Color management is sometimes sidestepped by calibrating devices to a common standard color space such as sRGB; when such calibration is done well enough, no color translations are needed to get all devices to handle colors consistently. This avoidance of the complexity of color management was one of the goals in the development of sRGB. == Color profiles == === Embedding === Image formats themselves (such as TIFF, JPEG, PNG, EPS, PDF, and SVG) may contain embedded color profiles but are not required to do so by the image format. The International Color Consortium standard was created to bring various developers and manufacturers together. The ICC standard permits the exchange of output device characteristics and color spaces in the form of metadata. This allows the embedding of color profiles into images as well as storing them in a database or a profile directory. === Working spaces === Working spaces, such as sRGB, Adobe RGB or ProPhoto are color spaces that facilitate good results while editing. For instance, pixels with equal values of R,G,B should appear neutral. Using a large (gamut) working space will lead to posterization, while using a small working space will lead to clipping. This trade-off is a consideration for the critical image editor. == Color transformation == Color transformation, or color space conversion, is the transformation of the representation of a color from one color space to another. This calculation is required whenever data is exchanged inside a color-managed chain and carried out by a Color Matching Module. Transforming profiled color information to different output devices is achieved by referencing the profile data into a standard color space. It makes it easier to convert colors from one device to a selected standard color space and from that to the colors of another device. By ensuring that the reference color space covers the many possible colors that humans can see, this concept allows one to exchange colors between many different color output devices. Color transformations can be represented by two profiles (source profile and target profile) or by a devicelink profile. In this process there are approximations involved which make sure that the image keeps its important color qualities and also gives an opportunity to control on how the colors are being changed. === Profile connection space === In the terminology of the International Color Consortium, a translation between two color spaces can go through a profile connection space (PCS): Color Space 1 → PCS (CIELAB or CIEXYZ) → Color space 2; conversions into and out of the PCS are each specified by a profile. === Gamut mapping === In nearly every translation process, we have to deal with the fact that the color gamut of different devices vary in range which makes an accurate reproduction impossible. They therefore need some rearrangement near the borders of the gamut. Some colors must be shifted to the inside of the gamut, as they otherwise cannot be represented on the output device and would simply be clipped. This so-called gamut mismatch occurs for example, when we translate from the RGB color space with a wider gamut into the CMYK color space with a narrower gamut range. In this example, the dark highly saturated purplish-blue color of a typical computer monitor's "blue" primary is impossible to print on paper with a typical CMYK printer. The nearest approximation within the printer's gamut will be much less saturated. Conversely, an inkjet printer's "cyan" primary, a saturated mid-brightness blue, is outside the gamut of a typical computer monitor. The color management system can utilize various methods to achieve desired results and give experienced users control of the gamut mapping behavior. ==== Rendering intent ==== When the gamut of source color space exceeds that of the destination, saturated colors are liable to become clipped (inaccurately represented), or more formally burned. The color management module can deal with this problem in several ways. The ICC specification includes four different rendering intents, listed below. Before the actual rendering intent is carried out, one can temporarily simulate the rendering by soft proofing. It is a useful tool as it predicts the outcome of the colors and is available as an application in many color management systems: Absolute colorimetric Absolute colorimetry and relative colorimetry actually use the same table but differ in the adjust
Read more →
Lisp machines are general-purpose computers designed to efficiently run Lisp as their main software and programming language, usually via hardware support. They are an example of a high-level language computer architecture. In a sense, they were the first commercial single-user workstations. Despite being modest in number (perhaps 7,000 units total as of 1988) Lisp machines commercially pioneered some now-commonplace technologies, including networking innovations such as Chaosnet, and effective garbage collection. Several firms built and sold Lisp machines in the 1980s: Symbolics (3600, 3640, XL1200, MacIvory, and other models), Lisp Machines Incorporated (LMI Lambda), Texas Instruments (Explorer, MicroExplorer), and Xerox (Interlisp-D workstations). The operating systems were written in Lisp Machine Lisp, Interlisp (Xerox), and later partly in Common Lisp. == History == === Historical context === Artificial intelligence (AI) computer programs of the 1960s and 1970s intrinsically required what was then considered a huge amount of computer power, as measured in processor time and memory space. The power requirements of AI research were exacerbated by the Lisp symbolic programming language, when commercial hardware was designed and optimized for assembly- and Fortran-like programming languages. At first, the cost of such computer hardware meant that it had to be shared among many users. As integrated circuit technology shrank the size and cost of computers in the 1960s and early 1970s, and the memory needs of AI programs began to exceed the address space of the most common research computer, the Digital Equipment Corporation (DEC) PDP-10, researchers considered a new approach: a computer designed specifically to develop and run large artificial intelligence programs, and tailored to the semantics of the Lisp language. To provide consistent performance for interactive programs, these machines would often not be shared, but would be dedicated to a single user at a time. === Initial development === In 1973, Richard Greenblatt and Thomas Knight, programmers at Massachusetts Institute of Technology (MIT) Artificial Intelligence Laboratory (AI Lab), began what would become the MIT Lisp Machine Project when they first began building a computer hardwired to run certain basic Lisp operations, rather than run them in software, in a 24-bit tagged architecture. The machine also did incremental (or Arena) garbage collection. More specifically, since Lisp variables are typed at runtime rather than compile time, a simple addition of two variables could take five times as long on conventional hardware, due to test and branch instructions. Lisp Machines ran the tests in parallel with the more conventional single instruction additions. If the simultaneous tests failed, then the result was discarded and recomputed; this meant in many cases a speed increase by several factors. This simultaneous checking approach was used as well in testing the bounds of arrays when referenced, and other memory management necessities (not merely garbage collection or arrays). Type checking was further improved and automated when the conventional byte word of 32 bits was lengthened to 36 bits for Symbolics 3600-model Lisp machines and eventually to 40 bits or more (usually, the excess bits not accounted for by the following were used for error-correcting codes). The first group of extra bits were used to hold type data, making the machine a tagged architecture, and the remaining bits were used to implement compressed data representation (CDR) coding (wherein the usual linked list elements are compressed to occupy roughly half the space), aiding garbage collection by reportedly an order of magnitude. A further improvement was two microcode instructions which specifically supported Lisp functions, reducing the cost of calling a function to as little as 20 clock cycles, in some Symbolics implementations. The first machine was called the CONS machine (named after the list construction operator cons in Lisp). Often it was affectionately referred to as the Knight machine, perhaps since Knight wrote his master's thesis on the subject; it was extremely well received. It was subsequently improved into a version called CADR (a pun; in Lisp, the cadr function, which returns the second item of a list, is pronounced /ˈkeɪ.dəɹ/ or /ˈkɑ.dəɹ/, as some pronounce the word "cadre") which was based on essentially the same architecture. About 25 of what were essentially prototype CADRs were sold within and without MIT for ~$50,000; it quickly became the favorite machine for hacking – many of the most favored software tools were quickly ported to it (e.g. Emacs was ported from ITS in 1975). It was so well received at an AI conference held at MIT in 1978 that Defense Advanced Research Projects Agency (DARPA) began funding its development. === Commercializing MIT Lisp machine technology === In 1979, Russell Noftsker, being convinced that Lisp machines had a bright commercial future due to the strength of the Lisp language and the enabling factor of hardware acceleration, proposed to Greenblatt that they commercialize the technology. In a counter-intuitive move for an AI Lab hacker, Greenblatt acquiesced, hoping perhaps that he could recreate the informal and productive atmosphere of the Lab in a real business. These ideas and goals were considerably different from those of Noftsker. The two negotiated at length, but neither would compromise. As the proposed firm could succeed only with the full and undivided assistance of the AI Lab hackers as a group, Noftsker and Greenblatt decided that the fate of the enterprise was up to them, and so the choice should be left to the hackers. The ensuing discussions of the choice divided the lab into two factions. In February 1979, matters came to a head. The hackers sided with Noftsker, believing that a commercial venture-fund-backed firm had a better chance of surviving and commercializing Lisp machines than Greenblatt's proposed self-sustaining start-up. Greenblatt lost the battle. It was at this juncture that Symbolics, Noftsker's enterprise, slowly came together. While Noftsker was paying his staff a salary, he had no building or any equipment for the hackers to work on. He bargained with Patrick Winston that, in exchange for allowing Symbolics' staff to keep working out of MIT, Symbolics would let MIT use internally and freely all the software Symbolics developed. A consultant from CDC, who was trying to put together a natural language computer application with a group of West-coast programmers, came to Greenblatt, seeking a Lisp machine for his group to work with, about eight months after the disastrous conference with Noftsker. Greenblatt had decided to start his own rival Lisp machine firm, but he had done nothing. The consultant, Alexander Jacobson, decided that the only way Greenblatt was going to start the firm and build the Lisp machines that Jacobson desperately needed was if Jacobson pushed and otherwise helped Greenblatt launch the firm. Jacobson pulled together business plans, a board, a partner for Greenblatt (one F. Stephen Wyle). The newfound firm was named LISP Machine, Inc. (LMI), and was funded by CDC orders, via Jacobson. Around this time Symbolics (Noftsker's firm) began operating. It had been hindered by Noftsker's promise to give Greenblatt a year's head start, and by severe delays in procuring venture capital. Symbolics still had the major advantage that while 3 or 4 of the AI Lab hackers had gone to work for Greenblatt, 14 other hackers had signed onto Symbolics. Two AI Lab people were not hired by either: Richard Stallman and Marvin Minsky. Stallman, however, blamed Symbolics for the decline of the hacker community that had centered around the AI lab. For two years, from 1982 to the end of 1983, Stallman worked by himself to clone the output of the Symbolics programmers, with the aim of preventing them from gaining a monopoly on the lab's computers. Regardless, after a series of internal battles, Symbolics did get off the ground in 1980/1981, selling the CADR as the LM-2, while Lisp Machines, Inc. sold it as the LMI-CADR. Symbolics did not intend to produce many LM-2s, since the 3600 family of Lisp machines was supposed to ship quickly, but the 3600s were repeatedly delayed, and Symbolics ended up producing ~100 LM-2s, each of which sold for $70,000. Both firms developed second-generation products based on the CADR: the Symbolics 3600 and the LMI-LAMBDA (of which LMI managed to sell ~200). The 3600, which shipped a year late, expanded on the CADR by widening the machine word to 36-bits, expanding the address space to 28-bits, and adding hardware to accelerate certain common functions that were implemented in microcode on the CADR. The LMI-LAMBDA, which came out a year after the 3600, in 1983, was compatible with the CADR (it could run CADR microcode), but hardware differences existed. Texas Instruments (TI) joined the fray whe
Read more →
TD-Gammon is a computer backgammon program developed in the 1990s by Gerald Tesauro at IBM's Thomas J. Watson Research Center. Its name comes from the fact that it is an artificial neural net trained by a form of temporal-difference learning, specifically TD-Lambda. It explored strategies that humans had not pursued and led to advances in the theory of correct backgammon play. In 1993, TD-Gammon (version 2.1) was trained with 1.5 million games of self-play, and achieved a level of play just slightly below that of the top human backgammon players of the time. In 1998, during a 100-game series, it was defeated by the world champion by a mere margin of 8 points. Its unconventional assessment of some opening strategies had been accepted and adopted by expert players. TD-gammon is commonly cited as an early success of reinforcement learning and neural networks, and was cited in, for example, papers for deep Q-learning and AlphaGo. == Algorithm for play and learning == During play, TD-Gammon examines on each turn all possible legal moves and all their possible responses (lookahead search), feeds each resulting board position into its evaluation function, and chooses the move that leads to the board position that got the highest score. In this respect, TD-Gammon is no different than almost any other computer board-game program. TD-Gammon's innovation was in how it learned its evaluation function. TD-Gammon's learning algorithm consists of updating the weights in its neural net after each turn to reduce the difference between its evaluation of previous turns' board positions and its evaluation of the present turn's board position—hence "temporal-difference learning". The score of any board position is a set of four numbers reflecting the program's estimate of the likelihood of each possible game result: White wins normally, Black wins normally, White wins a gammon, Black wins a gammon. For the final board position of the game, the algorithm compares with the actual result of the game rather than its own evaluation of the board position. The core of TD-gammon is a neural network with 3 layers. The input layer has two types of neurons. One type codes for the board position. They are non-negative integers ranging from 0 to 15, indicating the number of White or Black checkers at each board location. There are 99 input neurons for each, totaling 198 neurons. Another type codes for hand-crafted features previously used in Neurogammon. These features encoded standard concepts used by human experts, such as "advanced anchor," "blockade strength," "home board strength" and the probability of a "blot" (single checker) being hit. The hidden layer contains hidden neurons. Later versions had more of these. The output layer contains 4 neurons, representing the network's estimate of the probability ("equity") that the current board would lead to. The 4 neurons code for: White normal win, White gammon win, Black normal win, Black gammon win. Backgammon win is so rare that Tesauro opted to not represent it. After each turn, the learning algorithm updates each weight in the neural net according to the following rule: w t + 1 − w t = α ( Y t + 1 − Y t ) ∑ k = 1 t λ t − k ∇ w Y k {\displaystyle w_{t+1}-w_{t}=\alpha (Y_{t+1}-Y_{t})\sum _{k=1}^{t}\lambda ^{t-k}\nabla _{w}Y_{k}} where: It was found that picking small λ {\displaystyle \lambda } offered performance roughly equally good, and large λ {\displaystyle \lambda } degraded performance. Because of this, after 1992, TD-Gammon was trained with λ = 0 {\displaystyle \lambda =0} , degenerating into standard TD-learning. This saved compute by a factor of 2. == Development history == Version 1.0 used simple 1-ply search: every next move is scored by the neural net, and the highest-scoring move is selected. Versions 2.0 and 2.1 used 2-ply search: Make a 1-ply analysis to remove unlikely moves ("forward pruning"). Make a 2-play minimax analysis for only the likely moves. Pick the best move, probability-weighted by each of the opponent's 21 possible dice rolls (weighting non-doubles twice as much as doubles). Versions 3.0 and 3.1 used 3-ply search, using 21 2 = 441 {\displaystyle 21^{2}=441} possible dice rolls instead of 21. The last version, 3.1, was trained specifically for an exhibition match against Malcolm Davis at the 1998 AAAI Hall of Champions. It lost at -8 points, mainly due to one blunder, where TD-Gammon opted to double and got gammoned at -32 points. == Experiments and stages of training == Unlike previous neural-net backgammon programs such as Neurogammon (also written by Tesauro), where an expert trained the program by supplying the "correct" evaluation of each position, TD-Gammon was at first programmed "knowledge-free". In early experimentation, using only a raw board encoding with no human-designed features, TD-Gammon reached a level of play comparable to Neurogammon: that of an intermediate-level human backgammon player. Even though TD-Gammon discovered insightful features on its own, Tesauro wondered if its play could be improved by using hand-designed features like Neurogammon's. Indeed, the self-training TD-Gammon with expert-designed features soon surpassed all previous computer backgammon programs. It stopped improving after about 1,500,000 games (self-play) using a three-layered neural network, with 198 input units encoding expert-designed features, 80 hidden units, and one output unit representing predicted probability of winning. == Advances in backgammon theory == TD-Gammon's exclusive training through self-play (rather than imitation learning) enabled it to explore strategies that humans previously had not considered or had ruled out erroneously. Its success with unorthodox strategies had a significant impact on the backgammon community. Late 1991, Bill Robertie, Paul Magriel, and Malcolm Davis, were invited to play against TD-Gammon (version 1.0). A total of 51 games were played, with TD-Gammon losing at -0.25 ppg. Robertie found TD-Gammon to be at the level of a competent advanced player, and better than any previous backgammon program. Robertie subsequently wrote about the use of TD-Gammon for backgammon study. For example, on the opening play, the conventional wisdom was that given a roll of 2-1, 4-1, or 5-1, White should move a single checker from point 6 to point 5. Known as "slotting", this technique trades the risk of a hit for the opportunity to develop an aggressive position. TD-Gammon found that the more conservative play of splitting 24-23 was superior. Tournament players began experimenting with TD-Gammon's move, and found success. Within a few years, slotting had disappeared from tournament play, replaced by splitting, though in 2006 it made a reappearance for 2-1. Backgammon expert Kit Woolsey found that TD-Gammon's positional judgement, especially its weighing of risk against safety, was superior to his own or any human's. TD-Gammon's excellent positional play was undercut by occasional poor endgame play. The endgame requires a more analytical approach, sometimes with extensive lookahead. TD-Gammon's limitation to two-ply lookahead put a ceiling on what it could achieve in this part of the game. TD-Gammon's strengths and weaknesses were the opposite of symbolic artificial intelligence programs and most computer software in general: it was good at matters that require an intuitive "feel" but bad at systematic analysis. It is also poor at doubling strategies. This is likely due to the fact that the neural network is trained without the doubling cube, with the doubling added by feeding the neural network's cubeless equity estimates into theoretically-based heuristic formulae. This was particularly the case in the 1998 exhibition match, where it played 100 games against Malcolm Davis. A single doubling blunder lost the match. TD-gammon was never commercialized or released to the public in some other form, but it inspired commercial backgammon programs based on neural networks, such as JellyFish (1994) and Snowie (1998).
Read more →
The AgMES (Agricultural Metadata Element set) initiative was developed by the Food and Agriculture Organization (FAO) of the United Nations and aims to encompass issues of semantic standards in the domain of agriculture with respect to description, resource discovery, interoperability, and data exchange for different types of information resources. There are numerous other metadata schemas for different types of information resources. The following list contains a list of a few examples: Document-like Information Objects (DLIOs): Dublin Core, Agricultural Metadata Element Set (AgMES) Events: VCalendar Geographic and Regional Information: Geographic information—Metadata ISO/IEC 11179 Standards Persons: Friend-of-a-friend (FOAF), vCard Plant Production and Protection: Darwin Core (1.0 and 2.0) (DwC) AgMES as a namespace is designed to include agriculture specific extensions for terms and refinements from established standard metadata namespaces like Dublin Core, AGLS etc. Thus, to be used for Document-like Information Objects, for example like publications, articles, books, web sites, papers, etc., it will have to be used in conjunction with the standard namespaces mentioned before. The AgMES initiative strives to achieve improved interoperability between information resources in agricultural domain by enabling means for exchange of information. Describing a DLIO with AgMES means exposing its major characteristics and contents in a standard way that can be reused easily in any information system. The more institutions and organizations in the agricultural domain that use AgMES to describe their DLIOs, the easier it will be to interchange data in between information systems like digital libraries and other repositories of agricultural information. == Use of AgMES == Metadata on agricultural Document-like Information Objects (DLIOs) can be created and stored in various formats: embedded in a web site (in the manner as with the HTML meta tag) in a separate metadata database in an XML file in an RDF file AgMES defines elements that can be used to describe a DLIO that can be used together with other metadata standards such as the Dublin Core, the Australian Government Locator Service. A complete list of all elements, refinements and schemes endorsed by AgMES is available from the AgMES website. === Creating application profiles === Application profiles are defined as schemas which consist of data elements drawn from one or more namespaces, combined by implementers, and optimized for a particular local application. Application profiles share the following four characteristics: They draw upon existing pool of metadata definition standards to extract suitable application- or requirement oriented elements. An application profile cannot create new elements. Application profiles specify the application specific details such as the schemes or controlled vocabularies. An application profile also contains information such as the format for the element value, cardinality or data type. Lastly, an application profile can refine standardized definitions as long as it is "semantically narrower or more specific". This capability of application profiles caters to situations where a domain specific terminology is needed to replace a more general one. === Sample application profiles using AgMES === The AGRIS Application Profile is a standard created specifically to enhance the description, exchange and subsequent retrieval of agricultural Document-like Information Objects (DLIOs). It is a format that allows sharing of information across dispersed bibliographic systems and is based on well-known and accepted metadata standards. The Event Application Profile is a standard created to allow members of the Agricultural community to 'know' about an upcoming event and guide them to the event Web site where they can find further information. The information communicated is thus minimum yet interoperable across domains and organizations. == AgMES and the semantic web == One of the advantages of the AgMES metadata schema is the ability to link between the metadata element and controlled vocabularies. The use of controlled vocabulary provides a "known" set of options to the indexer (and the search programmer) as to how the field can be filled out. Often the values may come from a specific thesaurus (e.g. AGROVOC) or classification schemes (e.g. the AGRIS/CARIS classification scheme) etc. Thanks to the possibility to use controlled vocabularies for metadata elements, the user is provided with the most precise information. In this context, work is also being carried out on exploiting the power of controlled vocabularies expressed as using URIs and machine-understandable semantics. In this context, FAO is promoting the Agricultural Ontology Service (AOS) initiative with the objective of expressing more semantics within the traditional thesaurus AGROVOC and build a Concept Server as a repository from which it will be always possible to extract traditional KOS.
Read more →
A cloudlet is a mobility-enhanced small-scale cloud datacenter that is located at the edge of the Internet. The main purpose of the cloudlet is supporting resource-intensive and interactive mobile applications by providing powerful computing resources to mobile devices with lower latency. It is a new architectural element that extends today's cloud computing infrastructure. It represents the middle tier of a 3-tier hierarchy: mobile device - cloudlet - cloud. A cloudlet can be viewed as a data center in a box whose goal is to bring the cloud closer. The cloudlet term was first coined by M. Satyanarayanan, Victor Bahl, Ramón Cáceres, and Nigel Davies, and a prototype implementation is developed by Carnegie Mellon University as a research project. The concept of cloudlet is also known as follow me cloud, and mobile micro-cloud. == Motivation == Many mobile services split the application into a front-end client program and a back-end server program following the traditional client-server model. The front-end mobile application offloads its functionality to the back-end servers for various reasons such as speeding up processing. With the advent of cloud computing, the back-end server is typically hosted at the cloud datacenter. Though the use of a cloud datacenter offers various benefits such as scalability and elasticity, its consolidation and centralization lead to a large separation between a mobile device and its associated datacenter. End-to-end communication then involves many network hops and results in high latencies and low bandwidth. For the reasons of latency, some emerging mobile applications require cloud offload infrastructure to be close to the mobile device to achieve low response time. In the ideal case, it is just one wireless hop away. For example, the offload infrastructure could be located in a cellular base station or it could be LAN-connected to a set of Wi-Fi base stations. The individual elements of this offload infrastructure are referred to as cloudlets. == Applications == Cloudlets aim to support mobile applications that are both resource-intensive and interactive. Augmented reality applications that use head-tracked systems require end-to-end latencies of less than 16 ms. Cloud games with remote rendering also require low latencies and high bandwidth. Wearable cognitive assistance systems combine devices such as Google Glass with cloud-based processing to guide users through complex tasks. This futuristic genre of applications is characterized as “astonishingly transformative” by the report of the 2013 NSF Workshop on Future Directions in Wireless Networking. These applications use cloud resources in the critical path of real-time user interaction. Consequently, they cannot tolerate end-to-end operation latencies of more than a few tens of milliseconds. Apple Siri and Google Now which perform compute-intensive speech recognition in the cloud, are further examples in this emerging space. == Cloudlet vs Cloud == There is significant overlap in the requirements for cloud and cloudlet. At both levels, there is the need for: (a) strong isolation between untrusted user-level computations; (b) mechanisms for authentication, access control, and metering; (c) dynamic resource allocation for user-level computations; and, (d) the ability to support a very wide range of user-level computations, with minimal restrictions on their process structure, programming languages or operating systems. At a cloud datacenter, these requirements are met today using the virtual machine (VM) abstraction. For the same reasons they are used in cloud computing today, VMs are used as an abstraction for cloudlets. Meanwhile, there are a few but important differentiators between cloud and cloudlet. === Rapid provisioning === Different from cloud data centers that are optimized for launching existing VM images in their storage tier, cloudlets need to be much more agile in their provisioning. Their association with mobile devices is highly dynamic, with considerable churn due to user mobility. A user from far away may unexpectedly show up at a cloudlet (e.g., if he just got off an international flight) and try to use it for an application such as a personalized language translator. For that user, the provisioning delay before he is able to use the application impacts usability. === VM handoff across cloudlets === If a mobile device user moves away from the cloudlet he is currently using, the interactive response will degrade as the logical network distance increases. To address this effect of user mobility, the offloaded services on the first cloudlet need to be transferred to the second cloudlet maintaining end-to-end network quality. This resembles live migration in cloud computing but differs considerably in a sense that the VM handoff happens in Wide Area Network (WAN). == OpenStack++ == Since the cloudlet model requires reconfiguration or additional deployment of hardware/software, it is important to provide a systematic way to incentivise the deployment. However, it can face a classic bootstrapping problem. Cloudlets need practical applications to incentivize cloudlet deployment. However, developers cannot heavily rely on cloudlet infrastructure until it is widely deployed. To break this deadlock and bootstrap the cloudlet deployment, researchers at Carnegie Mellon University proposed OpenStack++ that extends OpenStack to leverage its open ecosystem. OpenStack++ provides a set of cloudlet-specific APIs as OpenStack extensions. == Commercial implementations and standardization effort == By 2015 cloudlet based applications were commercially available. In 2017 the National Institute of Standards and Technology published draft standards for fog computing in which cloudlets were defined as nodes on the fog architecture.
Read more →
John Schulman (born 1987 or 1988) is an American artificial intelligence researcher and co-founder of OpenAI. In August 2024, he announced he would be joining Anthropic. In February 2025, he announced he was leaving to join Thinking Machines Lab, where he is chief scientist. == Early life and education == Schulman had an interest in science and math from a young age. He enjoyed science fiction, especially the work of Isaac Asimov. When he was in seventh grade, he became deeply interested in the television program BattleBots, which featured combat between remote-controlled robots. In what he said was his first self-directed study, he read extensively in subject areas that would help him design a superior robot, but the robot he and his friends worked on was never built. He attended Great Neck South High School. He was a member of the US Physics olympiad Team in 2005. In 2010, he graduated from Caltech with a degree in physics. He has a PhD in electrical engineering and computer sciences from the University of California, Berkeley, where he was advised by Pieter Abbeel. == Career == In December 2015, shortly before finishing his PhD, Schulman co-founded OpenAI with Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman, Trevor Blackwell, Vicki Cheung, Andrej Karpathy, Durk Kingma, Pamela Vagata, and Wojciech Zaremba, with Sam Altman and Elon Musk as the co-chairs. There, he led the reinforcement learning team that created ChatGPT. He has been referred to as the "architect" of ChatGPT. In August 2024, Schulman announced he would be joining Anthropic. He stated his move was to allow him to deepen his focus on AI alignment and return to more hands-on technical work. In February 2025, he announced he was leaving to join Thinking Machines Lab, where he is chief scientist. == Awards and honors == In 2025, Schulman received the Mark Bingham Award for Excellence in Achievement by Young Alumni from his alma mater, UC Berkeley.
Read more →
The Ebert test gauges whether a computer-based synthesized voice can tell a joke with sufficient skill to cause people to laugh. It was proposed by film critic Roger Ebert at the 2011 TED conference as a challenge to software developers to have a computerized voice master the inflections, delivery, timing, and intonations of human speech. The test is similar to the Turing test proposed by Alan Turing in 1950 as a way to gauge a computer's ability to exhibit intelligent behavior by generating performance indistinguishable from a human being. If the computer can successfully tell a joke, and do the timing and delivery as well as Henny Youngman, then that's the voice I want. Ebert lost his voice in 2006 after undergoing surgery to treat thyroid cancer. He employed a Scottish company called CereProc, which custom-tailors text-to-speech software for voiceless customers who record their voices at length before losing them, and mined tapes and DVD commentaries featuring Ebert to create a voice that sounded more like his own voice. He first publicly used the voice they devised for him in his March 2, 2010, appearance on The Oprah Winfrey Show. The audience of Ebert's 2011 TED talk about joke delivery by synthesized voices erupted with laughter when a synthesized voice delivered the following joke: "A guy goes into a psychiatrist. The psychiatrist says, 'You’re crazy.' The guy says, 'I want a second opinion.' The psychiatrist says, 'All right, you’re ugly, too.'"
Read more →