AI Email Follow Up

AI Email Follow Up — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

  • Pattern theory

    Pattern theory

    Pattern theory, formulated by Ulf Grenander, is a mathematical formalism to describe knowledge of the world as patterns. It differs from other approaches to artificial intelligence in that it does not begin by prescribing algorithms and machinery to recognize and classify patterns; rather, it prescribes a vocabulary to articulate and recast the pattern concepts in precise language. Broad in its mathematical coverage, Pattern Theory spans algebra and statistics, as well as local topological and global entropic properties. In addition to the new algebraic vocabulary, its statistical approach is novel in its aim to: Identify the hidden variables of a data set using real world data rather than artificial stimuli, which was previously commonplace. Formulate prior distributions for hidden variables and models for the observed variables that form the vertices of a Gibbs-like graph. Study the randomness and variability of these graphs. Create the basic classes of stochastic models applied by listing the deformations of the patterns. Synthesize (sample) from the models, not just analyze signals with them. The Brown University Pattern Theory Group was formed in 1972 by Ulf Grenander. Many mathematicians are currently working in this group, noteworthy among them being the Fields Medalist David Mumford. Mumford regards Grenander as his "guru" in Pattern Theory.

    Read more →
  • Conceptual dependency theory

    Conceptual dependency theory

    Conceptual dependency theory is a model of natural language understanding used in artificial intelligence systems. Roger Schank at Stanford University introduced the model in 1969, in the early days of artificial intelligence. This model was extensively used by Schank's students at Yale University such as Robert Wilensky, Wendy Lehnert, and Janet Kolodner. Schank developed the model to represent knowledge for natural language input into computers. Partly influenced by the work of Sydney Lamb, his goal was to make the meaning independent of the words used in the input, i.e. two sentences identical in meaning would have a single representation. The system was also intended to draw logical inferences. The model uses the following basic representational tokens: real world objects, each with some attributes. real world actions, each with attributes times locations A set of conceptual transitions then act on this representation, e.g. an ATRANS is used to represent a transfer such as "give" or "take" while a PTRANS is used to act on locations such as "move" or "go". An MTRANS represents mental acts such as "tell", etc. A sentence such as "John gave a book to Mary" is then represented as the action of an ATRANS on two real world objects, John and Mary.

    Read more →
  • We Appreciate Power

    We Appreciate Power

    "We Appreciate Power" is a song by Canadian musician Grimes, featuring American musician Hana. It was released on November 29, 2018, billed as the lead single from her fifth studio album Miss Anthropocene, however it is only available on the Japanese and deluxe releases. The song was written and produced by Grimes, Poppy (originally), Hana and Chris Greatti. == Background and release == The song was supposed to be one of two collaborations between Grimes and American singer Poppy, for the latter's second studio album Am I a Girl?. In an interview, Poppy mentioned that she wrote two songs with Grimes; one about "destroying things" and another about "power". The other song, "Play Destroy", was featured on the album. Grimes shared a lyric of the song with a photo of her with Poppy on Twitter in May 2018. Following feuds between the two singers, the song was released by Grimes featuring singer Hana instead. On November 26, Grimes announced she would be releasing new music on November 29. Two days later, she revealed that the single is titled "We Appreciate Power" and features Hana, and shared the artwork. The release of the song was accompanied by a lyric video directed by Grimes and her brother Mac Boucher. == Music and lyrics == "We Appreciate Power" is an industrial rock, nu metal, and techno-industrial song. The track is regarded as a further step into Grimes's experimentation with guitars that started on 2015's Art Angels. The track was compared to the works of Nine Inch Nails; Jillian Mapes of Pitchfork described the song as "an immediate onslaught of mutilated noise—distorted metal guitar chug, bloody screams, a guitar loop that conjures fear and demands worship. Flashes of Nine Inch Nails' Pretty Hate Machine reverberate through the drum programming and synths." Brendan Klinkenberg of Rolling Stone placed the song "somewhere between power pop and straightforward industrial (with an extended bridge reminiscent of the most sweeping moments in a Final Fantasy score)" and "a distinctly 2018 take on Nine Inch Nails-esque hard-edged rock." A press release stated that the song was inspired by the North Korean band Moranbong and was written "from the perspective of a Pro-A.I. Girl Group Propaganda machine who use song, dance, sex and fashion to spread goodwill towards Artificial Intelligence." In addition Grimes stated that by simply listening to the song you will be reducing your risk of ending up on any future AI overlord's hit list when it reigns supreme, mirroring the Roko's basilisk theory. Lyrically, the song touches on transhumanist ideas such as the betterment and future of the human race, the possibilities of merging consciousness with machines to extend life indefinitely through mind uploading, and the idea that reality may be simulated. The song's chorus generated a spike in interest in the word "capitulate". == Critical reception == Pitchfork critic Jillian Mapes wrote: "If "Freak on a Leash" isn't a dealbreaker, then the supervillain allure of "We Appreciate Power" might pull you in (it legitimately slaps), but it just as well may leave you weighed down by Grimes' commitment to the absolute darkest timeline." Billboard's Gil Kaufman described the song as "a dystopian, aggressive dive into a more rock-leaning sound." Similarly, Brendan Klinkenberg of Rolling Stone called it "the most aggressive single Grimes has released to date" Noisey called the song "an absolute motherfucker of a single" and opined it sounds "like a K-pop band covering nu-metal". Justin Kamp of Paste described the track as a "glitchy empowerment anthem that chugs along on screeching synths and Grimes' repeated exultations of power." == Personnel == Credits adapted from Tidal. Grimes – vocals, guitar, production, engineering Hana – vocals, guitar, additional production Chris Greatti – guitar, keyboards, production, engineering Zakk Cervini – mixing == Track listing == == Charts ==

    Read more →
  • Emma Hart (computer scientist)

    Emma Hart (computer scientist)

    Professor Emma Hart, FRSE (born 1967) is an English computer scientist known for her work in artificial immune systems (AIS), evolutionary computation and optimisation. She is a professor of computational intelligence at Edinburgh Napier University, editor-in-chief of the Journal of Evolutionary Computation (MIT Press), and D. Coordinator of the Future & Emerging Technologies (FET) Proactive Initiative, Fundamentals of Collective Adaptive Systems. == Early life and education == Hart was born in Middlesbrough, England in 1967. In 1990 she graduated from the University of Oxford with a first class BA(Hons) in Chemistry. She then continued her studies at the University of Edinburgh, graduating with an MSc in Artificial Intelligence in 1994, followed by a PhD that explored the use of immunology as an inspiration for computing, examining a range of techniques applied to optimization and data classification problems. Her dissertation was titled Immunology as a metaphor for computational information processing: Fact or fiction?, and her doctoral advisor was Peter Ross. == Career == In 2000 Hart took a position as a lecturer at Edinburgh Napier University, and was promoted to a Reader, Professor, and in 2008 Chair in Natural Computation. She is now director of the Centre of Algorithms, Visualisation and Evolving Systems (CAVES) group in the School of Computing. She continues to research in the area of developing novel bio-inspired techniques for solving a range of real-world optimisation and classification problems, as well as exploring the fundamental properties of immune-inspired computing through modelling and simulation. She is also involved in editorial activity and currently occupies the position of Editor-in-Chief of the Journal of Evolutionary Computation (MIT Press). Her interests lie in the area of bio-inspired computing, in particular artificial immune systems (AIS). She also undertakes research in three main areas: optimisation, self-organising/self-adaptive systems, and artificial intelligence. Hart is D. Coordinator of Fundamentals of Collective Adaptive Systems (FoCAS), a Future and Emerging Technologies Proactive Initiative funded by the European Commission under FP7. == Selected works == === Conference talks === Hart, Emma. "Lifelong learning in optimization (video)". 28th European Conference on Operational Research. The Association of European Operational Research Societies. Hart, Emma (December 2021). "Self-assembling robots and the potential of artificial evolution". TED talk 2021. === Journal articles === "An immune system approach to scheduling in changing environments". E.Hart, P.Ross. 1999. Proceedings of the 1st Annual Conference on Genetic and Evolutionary Computation (2), 1559–1566. "Exploiting the analogy between immunology and sparse distributed memories: A system for clustering non-stationary data". E.Hart, P.Ross. 2002. 1st International Conference on Artificial Immune Systems. "Evolutionary scheduling: A review". E Hart, P Ross, D Corne. 2005. Genetic Programming and Evolvable Machines 6(2), 191–220. DOI: https://doi.org/10.1007/s10710-005-7580-7 "Application areas of AIS: The past, the present and the future". E.Hart, J.Timmis. 2008. Applied soft computing 8(1), 191–201. DOI: https://doi.org/10.1016/j.asoc.2006.12.004 "Structure versus function: a topological perspective on immune networks". E.Hart, H.Bersini, F.Santos. 2010. Natural computing 9(3), 603–624. DOI: https://doi.org/10.1007/s11047-009-9138-8 "On the life-long learning capabilities of a nelli: A hyper-heuristic optimisation system". E.Hart, K.Sim. 2014. International Conference on Parallel Problem Solving from Nature, 282–291. DOI: https://doi.org/10.1007/978-3-319-10762-2_28 "A hyper-heuristic ensemble method for static job-shop scheduling". E.Hart, K.Sim. 2016. Evolutionary computation 24(4), 609-635. DOI: https://dx.doi.org/10.1162/EVCO_a_00183 == Awards and recognition == 2016, Featured article on Lifelong Learning in Optimisation, IFORS newsletter 2016, "A Combined Generative and Selective Hyper-heuristic for the Vehicle Routing Problem" presented at GECCO 2016 (Denver, USA), ACM 2016, "A Hybrid Parameter Control Approach Applied to a Diversity-based Multi-objective Memetic Algorithm for Frequency Assignment Problems" presented at WCCI 2016 (Vancouver, Canada), IEEE 2017, Keynote Speaker, 2017 International Joint Conference on Computational Intelligence 2018, Bronze Award in International Human-Competitive Awards (Humies), International Conference on Genetic and Evolutionary Computation, Kyoto Japan 2018, Nomination for best paper award, GECCO 18, Kyoto, Japan 2022, Elected Fellow of the Royal Society of Edinburgh

    Read more →
  • SCADA Strangelove

    SCADA Strangelove

    SCADA Strangelove is an independent group of information security researchers founded in 2012, focused on security assessment of industrial control systems (ICS) and SCADA. == Activities == Main fields of research include: Discovery of 0-day vulnerabilities in cyber physical systems and coordinated vulnerability disclosure; Security assessment of ICS protocols and development suites; Identification of publicly Internet-connected ICS components and secure it with help of proper authorities; Development of security hardening guides for ICS software; Mapping cybersecurity on to functional safety; Awareness control and delivery of information regarding the actual security state of ICS systems. SCADA Strangelove's interests expand further than classic ICS components and covers various embedded systems, however, and encompass smart home components, solar panels, wind turbines, SmartGrid as well as other areas. == Projects == Group members have and continue to develop and publish numerous open source tools for scanning, fingerprinting, security evaluation and password bruteforcing for ICS devices. These devices work over industrial protocols such as modbus, Siemens S7, MMS, ISO EC 60870, ProfiNet. In 2014 Shodan used some of the published tools for building a map of ICS devices which is publicly available on the Internet. Open source security assessment frameworks, such as THC Hydra, Metasploit, and DigitalBond Redpoint have used Shodan-developed tools and techniques. The group has published security-hardening guidelines for industrial solutions based on Siemens SIMATIC WinCC and WinCC Flexible. The guidelines contain detailed security configuration walk-throughs, descriptions of internal security features and appropriate best practices. Among the group’s more noticeable projects is Choo Choo PWN (CCP) also named the Critical Infrastructure Attack (CIA). This is an interactive laboratory built upon ICS software and hardware used in real world. Every system is connected to a toy city infrastructure, which includes factories, railroads and other facilities. The laboratory has been demonstrated at various conferences including PHDays, Power of Community, and 30C3. Primarily the laboratory is used for the discovery of new vulnerabilities and for evaluation of security mechanisms, however it is also used for workshops and other educational activities. At Positive Hack Days IV, contestants found several 0-day vulnerabilities in Indusoft Web Studio 7.1 by Schneider Electric, and in specific ICS hardware RTU PET-7000 during the ICS vulnerability discovery challenge. The group supports Secure Open SmartGrid (SCADASOS) project to find and fix vulnerabilities in intellectual power grid components such as photovoltaic power station, wind turbine, power inverter. More than 80 000 industrial devices were discovered and isolated from the Internet in 2015. == Appearances == Group members are frequently seen presenting at conferences like CCC, SCADA Security Scientific Symposium, Positive Hack Days. Most notable talks are: === 29C3 === An overview of vulnerabilities discovered in the widely distributed Siemens SIMATIC WinCC software and tools that are implemented for searching ICS on the Internet. === PHDays === This talk consisted of an overview of vulnerabilities discovered in various systems produced by ABB, Emerson, Honeywell and Siemens and was presented at PHDays III and PHDays IV. === Confidence 2014 === Implications of security research aimed at realization of various industrial network protocols Profinet, Modbus, DNP3, IEC 61850-8-1 (MMS), IEC (International Electrotechnical Commission) 61870-5-101/104, FTE (Fault Tolerant Ethernet), Siemens S7. === PacSec 2014 === Presentations of security research showing the impact of radio and 3G/4G networks on the security of mobile devices as well as on industrial equipment. === 31C3 === Analysis of security architecture and implementation of the most wide spread platforms for wind and solar energy generation which produce many gigawatts of it. === 32C3 === Cybersecurity assessment of railway signaling systems such as Automatic Train Control (ATC), Computer-based interlocking (CBI) and European Train Control System (ETCS). === China Internet Security Conference 2016 === In "Greater China Cyber Threat Landscape" keynote by Sergey Gordeychik an overview of vulnerabilities, attacks and cyber-security incidents in Greater China region was presented. === Recon 2017 === In talk "Hopeless: Relay Protection for Substation Automation" by Kirill Nesterov and Alexander Tlyapov security analysis results of key Digital Substation component - Relay Protection Terminals was presented. Vulnerabilities, including remote code execution in Siemens SIPROTEC, General Electric Line Distance Relay, NARI and ABB protective relays was presented. == Philosophy == All names, catchwords and graphical elements refer to Stanley Kubrick’s film, Dr. Strangelove. In their talks, group members often refer to Cold War events such as the Caribbean Crisis, and draw parallels between nuclear arms race and the current escalation of cyberwar. Group members follow the approach of “responsible disclosure” and “ready to wait for years, while vendor is patching the vulnerability”. Public exploits for discovered vulnerabilities are not published. This is on account of the longevity of ICS and by implication the long process of patching ICS. However, conflicts still happen, notably in 2012 when the talk at DEF CON was called off due to a dispute of persistent weaknesses in Siemens industrial software.

    Read more →
  • DialogOS

    DialogOS

    DialogOS is a graphical programming environment to design computer system which can converse through voice with the user. Dialogs are clicked together in a Flowchart. DialogOS includes bindings to control Lego Mindstorms robots by voice and has bindings to SQL databases, as well as a generic plugin architecture to integrate with other types of backends. DialogOS is used in computer science courses in schools and universities to teach programming and to introduce beginners in the basic principles of human/computer interaction and dialog design. It has also been used in research systems. DialogOS was initially developed commercially by CLT Sprachtechnologie GmbH until its liquidation in 2017. The rights were then acquired by Saarland University and the software was released as open-source. == Bindings to Lego Mindstorms NXT == DialogOS can control the LEGO Mindstorms NXT Series. It uses sensor-nodes to obtain values for the following sensors: noise sensor ultrasonic sensor touch sensor luminosity sensor

    Read more →
  • AI art

    AI art

    Artificial intelligence visual art, or AI art, is visual artwork generated or enhanced through the implementation of artificial intelligence (AI) programs, most commonly using text-to-image models. The process of automated art-making has existed since antiquity. The field of artificial intelligence was founded in the 1950s, and artists began to create art with artificial intelligence shortly after the discipline's founding. A select number of these creations have been showcased in museums and have been recognized with awards. Throughout its history, AI has raised many philosophical questions related to the human mind, artificial beings, and the nature of art in human–AI collaboration. During the AI boom of the 2020s, text-to-image models such as Midjourney, DALL-E and Stable Diffusion became widely available to the public, allowing users to quickly generate imagery with little effort. Commentary about AI art in the 2020s has often focused on issues related to copyright, deception, defamation, and its impact on more traditional artists, including technological unemployment. In August 2023, the US Supreme Court ruled that AI art is ineligible for copyright due to failure to meet human authorship. In March 2026, it declined to hear a case over whether AI-generated art can be subject to copyright. == History == === Early history === Automated art dates back at least to the automata of ancient Greek civilization, when inventors such as Daedalus and Hero of Alexandria were described as designing machines capable of writing text, generating sounds, and playing music. Creative automatons have flourished throughout history, such as Maillardet's automaton, created around 1800 and capable of creating multiple drawings and poems. Also in the 19th century, Ada Lovelace, wrote that "computing operations" could potentially be used to generate music and poems. In 1950, Alan Turing's paper "Computing Machinery and Intelligence" focused on whether machines can mimic human behavior convincingly. Shortly after, the academic discipline of artificial intelligence was founded at a research workshop at Dartmouth College in 1956. Since its founding, AI researchers have explored philosophical questions about the nature of the human mind and the consequences of creating artificial beings with human-like intelligence; these issues have previously been explored by myth, fiction, and philosophy since antiquity. === Artistic history === Since the founding of AI in the 1950s, artists have used artificial intelligence to create artistic works. These works were sometimes referred to as algorithmic art, computer art, digital art, or new media art. One of the first significant AI art systems is AARON, developed by Harold Cohen beginning in the late 1960s at the University of California at San Diego. AARON uses a symbolic rule-based approach to generate technical images in the era of GOFAI programming, and it was developed by Cohen with the goal of being able to code the act of drawing. AARON was exhibited in 1972 at the Los Angeles County Museum of Art. From 1973 to 1975, Cohen refined AARON during a residency at the Artificial Intelligence Laboratory at Stanford University. In 2024, the Whitney Museum of American Art exhibited AI art from throughout Cohen's career, including re-created versions of his early robotic drawing machines. Karl Sims has exhibited art created with artificial life since the 1980s. He received an M.S. in computer graphics from the MIT Media Lab in 1987 and was artist-in-residence from 1990 to 1996 at the supercomputer manufacturer and artificial intelligence company Thinking Machines. In both 1991 and 1992, Sims won the Golden Nica award at Prix Ars Electronica for his videos using artificial evolution. In 1997, Sims created the interactive artificial evolution installation Galápagos for the NTT InterCommunication Center in Tokyo. Sims received an Emmy Award in 2019 for outstanding achievement in engineering development. In 1999, Scott Draves and a team of several engineers created and released Electric Sheep as a free software screensaver. Electric Sheep is a volunteer computing project for animating and evolving fractal flames, which are distributed to networked computers that display them as a screensaver. The screensaver used AI to create an infinite animation by learning from its audience. In 2001, Draves won the Fundacion Telefónica Life 4.0 prize for Electric Sheep. In 2014, Stephanie Dinkins began working on Conversations with Bina48. For the series, Dinkins recorded her conversations with BINA48, a social robot that resembles a middle-aged black woman. In 2019, Dinkins won the Creative Capital award for her creation of an evolving artificial intelligence based on the "interests and culture(s) of people of color." In 2015, Sougwen Chung began Mimicry (Drawing Operations Unit: Generation 1), an ongoing collaboration between the artist and a robotic arm. In 2019, Chung won the Lumen Prize for her continued performances with a robotic arm that uses AI to attempt to draw in a manner similar to Chung. In 2018, an auction sale of artificial intelligence art was held at Christie's in New York where the AI artwork Edmond de Belamy sold for US$432,500, which was almost 45 times higher than its estimate of US$7,000–10,000. The artwork was created by Obvious, a Paris-based collective. In 2024, Japanese film generAIdoscope was released. The film was co-directed by Hirotaka Adachi, Takeshi Sone, and Hiroki Yamaguchi. All video, audio, and music in the film were created with artificial intelligence. In 2025, the Japanese anime television series Twins Hinahima was released. The anime was produced and animated with AI assistance during the process of cutting and conversion of photographs into anime illustrations and later retouched by art staff. Most of the remaining parts such as characters and logos were hand-drawn with various software. === Technical history === Deep learning, characterized by its multi-layer structure that attempts to mimic the human brain, first came about in the 2010s, causing a significant shift in the world of AI art. During the deep learning era, there are mainly these types of designs for generative art: autoregressive models, diffusion models, GANs, normalizing flows. In 2014, Ian Goodfellow and colleagues at Université de Montréal developed the generative adversarial network (GAN), a type of deep neural network capable of learning to mimic the statistical distribution of input data such as images. The GAN uses a "generator" to create new images and a "discriminator" to decide which created images are considered successful. Unlike previous algorithmic art that followed hand-coded rules, generative adversarial networks could learn a specific aesthetic by analyzing a dataset of example images. In 2015, a team at Google released DeepDream, a program that uses a convolutional neural network to find and enhance patterns in images via algorithmic pareidolia. The process creates deliberately over-processed images with a dream-like appearance reminiscent of a psychedelic experience. Later, in 2017, a conditional GAN learned to generate 1000 image classes of ImageNet, a large visual database designed for use in visual object recognition software research. By conditioning the GAN on both random noise and a specific class label, this approach enhanced the quality of image synthesis for class-conditional models. Autoregressive models were used for image generation, such as PixelRNN (2016), which autoregressively generates one pixel after another with a recurrent neural network. Immediately after the Transformer architecture was proposed in Attention Is All You Need (2018), it was used for autoregressive generation of images, but without text conditioning. The website Artbreeder, launched in 2018, uses the models StyleGAN and BigGAN to allow users to generate and modify images such as faces, landscapes, and paintings. In the 2020s, text-to-image models, which generate images based on prompts, became widely used, marking yet another shift in the creation of AI-generated artworks. In 2021, using the influential large language generative pre-trained transformer models that are used in GPT-2 and GPT-3, OpenAI released a series of images created with the text-to-image AI model DALL-E 1. It is an autoregressive generative model with essentially the same architecture as GPT-3. Along with this, later in 2021, EleutherAI released the open source VQGAN-CLIP based on OpenAI's CLIP model. Diffusion models, generative models used to create synthetic data based on existing data, were first proposed in 2015, but they only became better than GANs in early 2021. Latent diffusion model was published in December 2021 and became the basis for the later Stable Diffusion (August 2022), developed through a collaboration between Stability AI, CompVis Group at LMU Munich, and Runway. In 2022, Midjourney was released, followed by Google Brain's Imagen and Pa

    Read more →
  • Monkey and banana problem

    Monkey and banana problem

    The monkey and banana problem is a famous toy problem in artificial intelligence, particularly in logic programming and planning. It has been framed as: A monkey is in a room containing a box and a bunch of bananas. The bananas are hanging from the ceiling out of reach of the monkey. How can the monkey obtain the bananas? The situation is used as a toy problem for computer science and can be solved with an expert system such as CLIPS. The example set of rules that CLIPS provides is somewhat fragile, in that, naive changes to the rulebase that might seem to a human of average intelligence to make common sense can cause the engine to fail to get the monkey to reach the banana. Other examples exist using Rules Based System (RBS), including a project implemented in Python.

    Read more →
  • Hint (app)

    Hint (app)

    Hint (hint.app) is an American software platform that provides astrological content, personality assessments, and relationship compatibility tools. The application was launched in 2018 and is based in Claymont, Delaware. The platform has been described in media coverage as part of a broader trend of astrology-based and self-reflection applications, particularly among younger users. As of 2026, the company reports that it has reached more than 25 million users worldwide. == History == Hint was founded in 2018 and is headquartered in Claymont, Delaware. The platform was developed to address a growing demand among Millennials and Gen Z for structured self-reflection tools that deviate from traditional religious or clinical psychological frameworks. The app has become a prominent figure in the "emotional technology" sector, reaching over 25 million global users by 2026. The platform is frequently cited by sociologists and media outlets as a primary driver of the Open-source intelligence trend, where individuals use digital tools to vet and analyze personal relationships in the dating economy. Media coverage has described the platform as part of a broader trend in which digital tools incorporate astrology and symbolic frameworks into wellness and relationship advice. == Reception == Coverage of Hint has appeared alongside reporting on changing attitudes toward dating and relationships, particularly among younger adults. Surveys reported by media outlets have described shifts in dating behavior, including reduced interest in casual relationships and increased reliance on digital tools for emotional reflection and compatibility assessment. Additional reporting has linked the use of astrology apps to broader trends in emotional fatigue and changing relationship expectations. Lifestyle and culture publications have described Hint, as an example of applications that integrate astrology into digital self-reflection and relationship analysis.

    Read more →
  • Transhuman Space

    Transhuman Space

    Transhuman Space (THS) is a role-playing game by David Pulver, published by Steve Jackson Games as part of the "Powered by GURPS" (Generic Universal Role-Playing System) line. Set in the year 2100, humanity has begun to colonize the Solar System. The pursuit of transhumanism is now in full swing, as more and more people reach fully posthuman states. In 2002, the Transhuman Space adventure "Orbital Decay" received an Origins Award nomination for Best Role-Playing Game Adventure. Transhuman Space won the 2003 Grog d'Or Award for Best Role-playing Game, Game Line or RPG Setting. == Setting == The game assumes that no cataclysm — natural or human-induced — swept Earth in the 21st century. Instead, constant developments in information technology, genetic engineering, nanotechnology and nuclear physics generally improved condition of the average human life. Plagues of the 20th century (like cancer or AIDS) have been suppressed, the ozone layer is being restored and Earth's ecosystems are recovering (although thermal emission by fusion power plants poses an environmental threat—albeit a much lesser one than previous sources of energy). Thanks to modern medicine humans live biblical timespans surrounded by various artificially intelligent helper applications and robots (cybershells), sensory experience broadcasts (future TV) and cyberspace telepresence. Thanks to cheap and clean fusion energy humanity has power to fuel all these wonders, restore and transform its home planet and finally settle on other heavenly bodies. Human genetic engineering has advanced to the point that anyone—single individuals, same-sex couples, groups of three or more—can reproduce. The embryos can be allowed to be developed naturally, or they can undergo three levels of tinkering: 1. Genefixing, which corrects defects; 2. Upgrades, which boost natural abilities (Ishtar Upgrades are slightly more attractive than usual, Metanoia Upgrades are more intelligent, etc.); and... 3. Full transition to parahuman status (Nyx Parahumans only need a few hours of sleep per week, Aquamorphs can live underwater, etc.) Another type of human genetic engineering, far more controversial, is the creation of bioroids, fully sentient slave races. People can "upload" by recording the simulation of their brains on computer disks. The emulated individual then becomes a ghost, an infomorph very easily confused with "sapient artificial intelligence". However, this technology has several problems as the solely available "brainpeeling" technique is fatal to the original biological lifeform being simulated, has a significant failure rate and the philosophical questions regarding personal identity remain equivocal. Any infomorph, regardless of its origin, can be plugged into a "cybershell" (robotic or cybernetic body), or a biological body, or "bioshell". Or, the individual can illegally make multiple "xoxes", or copies of themselves, and scatter them throughout the system, exponentially increasing the odds that at least one of them will live for centuries more, if not forever. This is also a time of space colonization. First, humanity (specifically China, followed by the United States and others) colonized Mars in a fashion resembling that outlined in the Mars Direct project. The Moon, Lagrangian points, inner planets and asteroids soon followed. In the late 21st century even some of Saturn's moons have been settled as a base for that planet's Helium-3 scooping operations. Transhuman Space's setting is neither utopia nor dystopia, however: several problems have arisen from these otherwise beneficial developments. The generation gap has become a chasm as lifespans increase. No longer do the elite fear death, and no longer can the young hope to replace them. While it seemed that outworld colonies would offer accommodation and work for those young ones, they are being replaced by genetically tailored bioroids and AI-powered cybershells. The concept of humanity is no longer clear in a world where even some animals speak of their rights and the dead haunt both cyberspace and reality (in form of infomorph-controlled bioshells or cybershells). And the wonders of high science are not universally shared — some countries merely struggle with informatization while others suffer from nanoplagues, defective drugs, implants and software tested on their populace. In some poor countries high-tech tyrants oppress their backward people. And in outer space all sort of modern crime thrives, barely suppressed by military forces. == Publication history == After the initial set of GURPS books that were published using the GURPS Lite, later publications such as Transhuman Space by David Pulver were labelled simply "Powered by GURPS" without using the name "GURPS" in the book title. Transhuman Space received a significant amount of supporting publications, and was the largest original background setting that Steve Jackson Games produced in 15 years. Shannon Appelcline noted that by its inclusion of posthuman characters, the book began to show the limits of the GURPS system as it was, which is something that Pulver would address soon thereafter. Steve Jackson Games has not updated the core book (GURPS Transhuman Space) to 4th edition, although the supplement Transhuman Space: Changing Times provides a path for migrating to 4th edition. It has produced several 4th edition supplements for the setting: Transhuman Space: Bioroid Bazaar, Transhuman Space: Cities on the Edge, Transhuman Space: Martial Arts 2100, Transhuman Space: Personnel Files 2-5, Transhuman Space: Shell-Tech, GURPS Spaceships 8: Transhuman Spacecraft, Transhuman Space: Transhuman Mysteries, and Transhuman Space: Wings of the Rising Sun. == Reception == In a review of Transhuman Space in Black Gate, William Stoddard said "Transhuman Space was a richly detailed setting; if it had imperfections, it had enough depth to make up for them. I think it has the potential to become a classic in its field. Perhaps a campaign set in its default start year of 2100 could leave the early twenty-first century blurry enough to avoid obvious incongruities." == Reviews == Review in Vol. 20, No. 1 of Prometheus, the journal of the Libertarian Futurist Society.

    Read more →
  • Estimation of distribution algorithm

    Estimation of distribution algorithm

    Estimation of distribution algorithms (EDAs), sometimes called probabilistic model-building genetic algorithms (PMBGAs), are stochastic optimization methods that guide the search for the optimum by building and sampling explicit probabilistic models of promising candidate solutions. Optimization is viewed as a series of incremental updates of a probabilistic model, starting with the model encoding an uninformative prior over admissible solutions and ending with the model that generates only the global optima. EDAs belong to the class of evolutionary algorithms. The main difference between EDAs and most conventional evolutionary algorithms is that evolutionary algorithms generate new candidate solutions using an implicit distribution defined by one or more variation operators, whereas EDAs use an explicit probability distribution encoded by a Bayesian network, a multivariate normal distribution, or another model class. Similarly as other evolutionary algorithms, EDAs can be used to solve optimization problems defined over a number of representations from vectors to LISP style S expressions, and the quality of candidate solutions is often evaluated using one or more objective functions. The general procedure of an EDA is outlined in the following: t := 0 initialize model M(0) to represent uniform distribution over admissible solutions while (termination criteria not met) do P := generate N>0 candidate solutions by sampling M(t) F := evaluate all candidate solutions in P M(t + 1) := adjust_model(P, F, M(t)) t := t + 1 Using explicit probabilistic models in optimization allowed EDAs to feasibly solve optimization problems that were notoriously difficult for most conventional evolutionary algorithms and traditional optimization techniques, such as problems with high levels of epistasis. Nonetheless, the advantage of EDAs is also that these algorithms provide an optimization practitioner with a series of probabilistic models that reveal a lot of information about the problem being solved. This information can in turn be used to design problem-specific neighborhood operators for local search, to bias future runs of EDAs on a similar problem, or to create an efficient computational model of the problem. For example, if the population is represented by bit strings of length 4, the EDA can represent the population of promising solution using a single vector of four probabilities (p1, p2, p3, p4) where each component of p defines the probability of that position being a 1. Using this probability vector it is possible to create an arbitrary number of candidate solutions. == Estimation of distribution algorithms (EDAs) == This section describes the models built by some well known EDAs of different levels of complexity. It is always assumed a population P ( t ) {\displaystyle P(t)} at the generation t {\displaystyle t} , a selection operator S {\displaystyle S} , a model-building operator α {\displaystyle \alpha } and a sampling operator β {\displaystyle \beta } . == Univariate factorizations == The most simple EDAs assume that decision variables are independent, i.e. p ( X 1 , X 2 ) = p ( X 1 ) ⋅ p ( X 2 ) {\displaystyle p(X_{1},X_{2})=p(X_{1})\cdot p(X_{2})} . Therefore, univariate EDAs rely only on univariate statistics and multivariate distributions must be factorized as the product of N {\displaystyle N} univariate probability distributions, D Univariate := p ( X 1 , … , X N ) = ∏ i = 1 N p ( X i ) . {\displaystyle D_{\text{Univariate}}:=p(X_{1},\dots ,X_{N})=\prod _{i=1}^{N}p(X_{i}).} Such factorizations are used in many different EDAs, next we describe some of them. === Univariate marginal distribution algorithm (UMDA) === The UMDA is a simple EDA that uses an operator α U M D A {\displaystyle \alpha _{UMDA}} to estimate marginal probabilities from a selected population S ( P ( t ) ) {\displaystyle S(P(t))} . By assuming S ( P ( t ) ) {\displaystyle S(P(t))} contain λ {\displaystyle \lambda } elements, α U M D A {\displaystyle \alpha _{UMDA}} produces probabilities: p t + 1 ( X i ) = 1 λ ∑ x ∈ S ( P ( t ) ) x i , ∀ i ∈ 1 , 2 , … , N . {\displaystyle p_{t+1}(X_{i})={\dfrac {1}{\lambda }}\sum _{x\in S(P(t))}x_{i},~\forall i\in 1,2,\dots ,N.} Every UMDA step can be described as follows D ( t + 1 ) = α UMDA ∘ S ∘ β λ ( D ( t ) ) . {\displaystyle D(t+1)=\alpha _{\text{UMDA}}\circ S\circ \beta _{\lambda }(D(t)).} === Population-based incremental learning (PBIL) === The PBIL, represents the population implicitly by its model, from which it samples new solutions and updates the model. At each generation, μ {\displaystyle \mu } individuals are sampled and λ ≤ μ {\displaystyle \lambda \leq \mu } are selected. Such individuals are then used to update the model as follows p t + 1 ( X i ) = ( 1 − γ ) p t ( X i ) + ( γ / λ ) ∑ x ∈ S ( P ( t ) ) x i , ∀ i ∈ 1 , 2 , … , N , {\displaystyle p_{t+1}(X_{i})=(1-\gamma )p_{t}(X_{i})+(\gamma /\lambda )\sum _{x\in S(P(t))}x_{i},~\forall i\in 1,2,\dots ,N,} where γ ∈ ( 0 , 1 ] {\displaystyle \gamma \in (0,1]} is a parameter defining the learning rate, a small value determines that the previous model p t ( X i ) {\displaystyle p_{t}(X_{i})} should be only slightly modified by the new solutions sampled. PBIL can be described as D ( t + 1 ) = α PIBIL ∘ S ∘ β μ ( D ( t ) ) {\displaystyle D(t+1)=\alpha _{\text{PIBIL}}\circ S\circ \beta _{\mu }(D(t))} === Compact genetic algorithm (cGA) === The CGA, also relies on the implicit populations defined by univariate distributions. At each generation t {\displaystyle t} , two individuals x , y {\displaystyle x,y} are sampled, P ( t ) = β 2 ( D ( t ) ) {\displaystyle P(t)=\beta _{2}(D(t))} . The population P ( t ) {\displaystyle P(t)} is then sorted in decreasing order of fitness, S Sort ( f ) ( P ( t ) ) {\displaystyle S_{{\text{Sort}}(f)}(P(t))} , with u {\displaystyle u} being the best and v {\displaystyle v} being the worst solution. The CGA estimates univariate probabilities as follows p t + 1 ( X i ) = p t ( X i ) + γ ( u i − v i ) , ∀ i ∈ 1 , 2 , … , N , {\displaystyle p_{t+1}(X_{i})=p_{t}(X_{i})+\gamma (u_{i}-v_{i}),\quad \forall i\in 1,2,\dots ,N,} where, γ ∈ ( 0 , 1 ] {\displaystyle \gamma \in (0,1]} is a constant defining the learning rate, usually set to γ = 1 / N {\displaystyle \gamma =1/N} . The CGA can be defined as D ( t + 1 ) = α CGA ∘ S Sort ( f ) ∘ β 2 ( D ( t ) ) {\displaystyle D(t+1)=\alpha _{\text{CGA}}\circ S_{{\text{Sort}}(f)}\circ \beta _{2}(D(t))} == Bivariate factorizations == Although univariate models can be computed efficiently, in many cases they are not representative enough to provide better performance than GAs. In order to overcome such a drawback, the use of bivariate factorizations was proposed in the EDA community, in which dependencies between pairs of variables could be modeled. A bivariate factorization can be defined as follows, where π i {\displaystyle \pi _{i}} contains a possible variable dependent to X i {\displaystyle X_{i}} , i.e. | π i | = 1 {\displaystyle |\pi _{i}|=1} . D Bivariate := p ( X 1 , … , X N ) = ∏ i = 1 N p ( X i | π i ) . {\displaystyle D_{\text{Bivariate}}:=p(X_{1},\dots ,X_{N})=\prod _{i=1}^{N}p(X_{i}|\pi _{i}).} Bivariate and multivariate distributions are usually represented as probabilistic graphical models (graphs), in which edges denote statistical dependencies (or conditional probabilities) and vertices denote variables. To learn the structure of a PGM from data linkage-learning is employed. === Mutual information maximizing input clustering (MIMIC) === The MIMIC factorizes the joint probability distribution in a chain-like model representing successive dependencies between variables. It finds a permutation of the decision variables, r : i ↦ j {\displaystyle r:i\mapsto j} , such that x r ( 1 ) x r ( 2 ) , … , x r ( N ) {\displaystyle x_{r(1)}x_{r(2)},\dots ,x_{r(N)}} minimizes the Kullback–Leibler divergence in relation to the true probability distribution, i.e. π r ( i + 1 ) = { X r ( i ) } {\displaystyle \pi _{r(i+1)}=\{X_{r(i)}\}} . MIMIC models a distribution p t + 1 ( X 1 , … , X N ) = p t ( X r ( N ) ) ∏ i = 1 N − 1 p t ( X r ( i ) | X r ( i + 1 ) ) . {\displaystyle p_{t+1}(X_{1},\dots ,X_{N})=p_{t}(X_{r(N)})\prod _{i=1}^{N-1}p_{t}(X_{r(i)}|X_{r(i+1)}).} New solutions are sampled from the leftmost to the rightmost variable, the first is generated independently and the others according to conditional probabilities. Since the estimated distribution must be recomputed each generation, MIMIC uses concrete populations in the following way P ( t + 1 ) = β μ ∘ α MIMIC ∘ S ( P ( t ) ) . {\displaystyle P(t+1)=\beta _{\mu }\circ \alpha _{\text{MIMIC}}\circ S(P(t)).} === Bivariate marginal distribution algorithm (BMDA) === The BMDA factorizes the joint probability distribution in bivariate distributions. First, a randomly chosen variable is added as a node in a graph, the most dependent variable to one of those in the graph is chosen among those not yet in the graph, this procedure is repeated until no remain

    Read more →
  • Generative adversarial network

    Generative adversarial network

    A generative adversarial network (GAN) is a class of machine learning frameworks and a prominent framework for approaching generative artificial intelligence. The concept was initially developed by Ian Goodfellow and his colleagues in June 2014. In a GAN, two neural networks compete with each other in the form of a zero-sum game, where one agent's gain is another agent's loss. Given a training set, this technique learns to generate new data with the same statistics as the training set. For example, a GAN trained on photographs can generate new photographs that look at least superficially authentic to human observers, having many realistic characteristics. Though originally proposed as a form of generative model for unsupervised learning, GANs have also proved useful for semi-supervised learning, fully supervised learning, and reinforcement learning. The core idea of a GAN is based on the "indirect" training through the discriminator, another neural network that can tell how "realistic" the input seems, which itself is also being updated dynamically. This means that the generator is not trained to minimize the distance to a specific image, but rather to fool the discriminator. This enables the model to learn in an unsupervised manner. GANs are similar to mimicry in evolutionary biology, with an evolutionary arms race between both networks. == Definition == === Mathematical === The original GAN is defined as the following game: Each probability space ( Ω , μ ref ) {\displaystyle (\Omega ,\mu _{\text{ref}})} defines a GAN game. There are 2 players: generator and discriminator. The generator's strategy set is P ( Ω ) {\displaystyle {\mathcal {P}}(\Omega )} , the set of all probability measures μ G {\displaystyle \mu _{G}} on Ω {\displaystyle \Omega } . The discriminator's strategy set is the set of Markov kernels μ D : Ω → P [ 0 , 1 ] {\displaystyle \mu _{D}:\Omega \to {\mathcal {P}}[0,1]} , where P [ 0 , 1 ] {\displaystyle {\mathcal {P}}[0,1]} is the set of probability measures on [ 0 , 1 ] {\displaystyle [0,1]} . The GAN game is a zero-sum game, with objective function L ( μ G , μ D ) := E x ∼ μ ref , y ∼ μ D ( x ) ⁡ [ ln ⁡ y ] + E x ∼ μ G , y ∼ μ D ( x ) ⁡ [ ln ⁡ ( 1 − y ) ] . {\displaystyle L(\mu _{G},\mu _{D}):=\operatorname {E} _{x\sim \mu _{\text{ref}},y\sim \mu _{D}(x)}[\ln y]+\operatorname {E} _{x\sim \mu _{G},y\sim \mu _{D}(x)}[\ln(1-y)].} The generator aims to minimize the objective, and the discriminator aims to maximize the objective. The generator's task is to approach μ G ≈ μ ref {\displaystyle \mu _{G}\approx \mu _{\text{ref}}} , that is, to match its own output distribution as closely as possible to the reference distribution. The discriminator's task is to output a value close to 1 when the input appears to be from the reference distribution, and to output a value close to 0 when the input looks like it came from the generator distribution. === In practice === The generative network generates candidates while the discriminative network evaluates them. This creates a contest based on data distributions, where the generator learns to map from a latent space to the true data distribution, aiming to produce candidates that the discriminator cannot distinguish from real data. The discriminator's goal is to correctly identify these candidates, but as the generator improves, its task becomes more challenging, increasing the discriminator's error rate. A known dataset serves as the initial training data for the discriminator. Training involves presenting it with samples from the training dataset until it achieves acceptable accuracy. The generator is trained based on whether it succeeds in fooling the discriminator. Typically, the generator is seeded with randomized input that is sampled from a predefined latent space (e.g. a multivariate normal distribution). Thereafter, candidates synthesized by the generator are evaluated by the discriminator. Independent backpropagation procedures are applied to both networks so that the generator produces better samples, while the discriminator becomes more skilled at flagging synthetic samples. When used for image generation, the generator is typically a deconvolutional neural network, and the discriminator is a convolutional neural network. === Relation to other statistical machine learning methods === GANs are implicit generative models, which means that they do not explicitly model the likelihood function nor provide a means for finding the latent variable corresponding to a given sample, unlike alternatives such as flow-based generative model. Compared to fully visible belief networks such as WaveNet and PixelRNN and autoregressive models in general, GANs can generate one complete sample in one pass, rather than multiple passes through the network. Compared to Boltzmann machines and linear ICA, there is no restriction on the type of function used by the network. Since neural networks are universal approximators, GANs are asymptotically consistent. Variational autoencoders might be universal approximators, but it is not proven as of 2017. == Mathematical properties == === Measure-theoretic considerations === This section provides some of the mathematical theory behind these methods. In modern probability theory based on measure theory, a probability space also needs to be equipped with a σ-algebra. As a result, a more rigorous definition of the GAN game would make the following changes:Each probability space ( Ω , B , μ ref ) {\displaystyle (\Omega ,{\mathcal {B}},\mu _{\text{ref}})} defines a GAN game. The generator's strategy set is P ( Ω , B ) {\displaystyle {\mathcal {P}}(\Omega ,{\mathcal {B}})} , the set of all probability measures μ G {\displaystyle \mu _{G}} on the measure-space ( Ω , B ) {\displaystyle (\Omega ,{\mathcal {B}})} . The discriminator's strategy set is the set of Markov kernels μ D : ( Ω , B ) → P ( [ 0 , 1 ] , B ( [ 0 , 1 ] ) ) {\displaystyle \mu _{D}:(\Omega ,{\mathcal {B}})\to {\mathcal {P}}([0,1],{\mathcal {B}}([0,1]))} , where B ( [ 0 , 1 ] ) {\displaystyle {\mathcal {B}}([0,1])} is the Borel σ-algebra on [ 0 , 1 ] {\displaystyle [0,1]} .Since issues of measurability never arise in practice, these will not concern us further. === Choice of the strategy set === In the most generic version of the GAN game described above, the strategy set for the discriminator contains all Markov kernels μ D : Ω → P [ 0 , 1 ] {\displaystyle \mu _{D}:\Omega \to {\mathcal {P}}[0,1]} , and the strategy set for the generator contains arbitrary probability distributions μ G {\displaystyle \mu _{G}} on Ω {\displaystyle \Omega } . However, as shown below, the optimal discriminator strategy against any μ G {\displaystyle \mu _{G}} is deterministic, so there is no loss of generality in restricting the discriminator's strategies to deterministic functions D : Ω → [ 0 , 1 ] {\displaystyle D:\Omega \to [0,1]} . In most applications, D {\displaystyle D} is a deep neural network function. As for the generator, while μ G {\displaystyle \mu _{G}} could theoretically be any computable probability distribution, in practice, it is usually implemented as a pushforward: μ G = μ Z ∘ G − 1 {\displaystyle \mu _{G}=\mu _{Z}\circ G^{-1}} . That is, start with a random variable z ∼ μ Z {\displaystyle z\sim \mu _{Z}} , where μ Z {\displaystyle \mu _{Z}} is a probability distribution that is easy to compute (such as the uniform distribution, or the Gaussian distribution), then define a function G : Ω Z → Ω {\displaystyle G:\Omega _{Z}\to \Omega } . Then the distribution μ G {\displaystyle \mu _{G}} is the distribution of G ( z ) {\displaystyle G(z)} . Consequently, the generator's strategy is usually defined as just G {\displaystyle G} , leaving z ∼ μ Z {\displaystyle z\sim \mu _{Z}} implicit. In this formalism, the GAN game objective is L ( G , D ) := E x ∼ μ ref ⁡ [ ln ⁡ D ( x ) ] + E z ∼ μ Z ⁡ [ ln ⁡ ( 1 − D ( G ( z ) ) ) ] . {\displaystyle L(G,D):=\operatorname {E} _{x\sim \mu _{\text{ref}}}[\ln D(x)]+\operatorname {E} _{z\sim \mu _{Z}}[\ln(1-D(G(z)))].} === Generative reparametrization === The GAN architecture has two main components. One is casting optimization into a game, of form min G max D L ( G , D ) {\displaystyle \min _{G}\max _{D}L(G,D)} , which is different from the usual kind of optimization, of form min θ L ( θ ) {\displaystyle \min _{\theta }L(\theta )} . The other is the decomposition of μ G {\displaystyle \mu _{G}} into μ Z ∘ G − 1 {\displaystyle \mu _{Z}\circ G^{-1}} , which can be understood as a reparametrization trick. To see its significance, one must compare GAN with previous methods for learning generative models, which were plagued with "intractable probabilistic computations that arise in maximum likelihood estimation and related strategies". At the same time, Kingma and Welling and Rezende et al. developed the same idea of reparametrization into a general stochastic backpropagation method. Among its first applications was the variational autoencoder. === Move order and st

    Read more →
  • Inception (deep learning architecture)

    Inception (deep learning architecture)

    Inception is a family of convolutional neural network (CNN) for computer vision, introduced by researchers at Google in 2014 as GoogLeNet (later renamed Inception v1). The series was historically important as an early CNN that separates the stem (data ingest), body (data processing), and head (prediction), an architectural design that persists in all modern CNN. == Version history == === Inception v1 === In 2014, a team at Google developed the GoogLeNet architecture, an instance of which won the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The name came from the LeNet of 1998, since both LeNet and GoogLeNet are CNNs. They also called it "Inception" after a "we need to go deeper" internet meme, a phrase from Inception (2010) the film. Because later, more versions were released, the original Inception architecture was renamed again as "Inception v1". The models and the code were released under Apache 2.0 license on GitHub. The Inception v1 architecture is a deep CNN composed of 22 layers. Most of these layers were "Inception modules". The original paper stated that Inception modules are a "logical culmination" of Network in Network and (Arora et al, 2014). Since Inception v1 is deep, it suffered from the vanishing gradient problem. The team solved it by using two "auxiliary classifiers", which are linear-softmax classifiers inserted at 1/3-deep and 2/3-deep within the network, and the loss function is a weighted sum of all three: L = 0.3 L a u x , 1 + 0.3 L a u x , 2 + L r e a l {\displaystyle L=0.3L_{aux,1}+0.3L_{aux,2}+L_{real}} These were removed after training was complete. This was later solved by the ResNet architecture. The architecture consists of three parts stacked on top of one another: The stem (data ingestion): The first few convolutional layers perform data preprocessing to downscale images to a smaller size. The body (data processing): The next many Inception modules perform the bulk of data processing. The head (prediction): The final fully-connected layer and softmax produces a probability distribution for image classification. This structure is used in most modern CNN architectures. === Inception v2 === Inception v2 was released in 2015, in a paper that is more famous for proposing batch normalization. It had 13.6 million parameters. It improves on Inception v1 by adding batch normalization, and removing dropout and local response normalization which they found became unnecessary when batch normalization is used. === Inception v3 === Inception v3 was released in 2016. It improves on Inception v2 by using factorized convolutions. As an example, a single 5×5 convolution can be factored into 3×3 stacked on top of another 3×3. Both has a receptive field of size 5×5. The 5×5 convolution kernel has 25 parameters, compared to just 18 in the factorized version. Thus, the 5×5 convolution is strictly more powerful than the factorized version. However, this power is not necessarily needed. Empirically, the research team found that factorized convolutions help. It also uses a form of dimension-reduction by concatenating the output from a convolutional layer and a pooling layer. As an example, a tensor of size 35 × 35 × 320 {\displaystyle 35\times 35\times 320} can be downscaled by a convolution with stride 2 to 17 × 17 × 320 {\displaystyle 17\times 17\times 320} , and by maxpooling with pool size 2 × 2 {\displaystyle 2\times 2} to 17 × 17 × 320 {\displaystyle 17\times 17\times 320} . These are then concatenated to 17 × 17 × 640 {\displaystyle 17\times 17\times 640} . Other than this, it also removed the lowest auxiliary classifier during training. They found that the auxiliary head worked as a form of regularization. They also proposed label-smoothing regularization in classification. For an image with label c {\displaystyle c} , instead of making the model to predict the probability distribution δ c = ( 0 , 0 , … , 0 , 1 ⏟ c -th entry , 0 , … , 0 ) {\displaystyle \delta _{c}=(0,0,\dots ,0,\underbrace {1} _{c{\text{-th entry}}},0,\dots ,0)} , they made the model predict the smoothed distribution ( 1 − ϵ ) δ c + ϵ / K {\displaystyle (1-\epsilon )\delta _{c}+\epsilon /K} where K {\displaystyle K} is the total number of classes. === Inception v4 === In 2017, the team released Inception v4, Inception ResNet v1, and Inception ResNet v2. Inception v4 is an incremental update with even more factorized convolutions, and other complications that were empirically found to improve benchmarks. Inception ResNet v1 and v2 are both modifications of Inception v4, where residual connections are added to each Inception module, inspired by the ResNet architecture. === Xception === Xception ("Extreme Inception") was published in 2017. It is a linear stack of depthwise separable convolution layers with residual connections. The design was proposed on the hypothesis that in a CNN, the cross-channels correlations and spatial correlations in the feature maps can be entirely decoupled. Training each network took 3 days on 60 K80 GPUs, or approximately 0.5 petaFLOP-days.

    Read more →
  • Veo (text-to-video model)

    Veo (text-to-video model)

    Veo, or Google Veo, is a text-to-video model developed by Google DeepMind and announced in May 2024. As a generative AI model, it creates videos based on user prompts. Veo 3, released in May 2025, can also generate accompanying audio. == Development == In May 2024, a multimodal video generation model called Veo was announced at Google I/O 2024. Google claimed that it could generate 1080p videos over a minute long. In December 2024, Google released Veo 2, available via VideoFX. It supports 4K resolution video generation and has an improved understanding of physics. In April 2025, Google announced that Veo 2 became available for advanced users on the Gemini app. In May 2025, Google released Veo 3, which not only generates videos but also creates synchronized audio — including dialogue, sound effects, and ambient noise — to match the visuals. Google also announced Flow, a video-creation tool powered by Veo and Imagen. Google DeepMind CEO Demis Hassabis described the release as the moment when AI video generation left the era of the silent film. This was rebranded as Google Flow at the 2026 Google I/O keynote, along with the announcement of Google Flow Music. == Capabilities == Google Veo can be purchased at multiple subscription tiers and through Google "AI credits". The software itself can be run by two different consoles, Google Gemini and Google Flow. Gemini being geared towards shorter, quicker, and faster projects, using the Gemini AI chat model, with Google Flow, which is essentially a movie editor allowing users to create longer projects with continuity, using the same characters and actors. Users can create a maximum of eight seconds per clip. According to Gizmodo Veo 3 users were directing the model to generate low-quality content, such as man on the street interviews or haul videos of people unboxing products. 404 Media reported that the tool tended to repeat the same joke in response to different prompts. Commentators speculated that Google had trained the service on YouTube videos or Reddit posts. Google itself had not stated the source of its training content. In July 2025, Media Matters for America reported that racist and antisemitic videos generated using Veo 3 were being uploaded to TikTok. Ryan Whitwam of Ars Technica commented, "In a perfect world, Veo 3 would refuse to create these videos, but vagueness in the prompt and the AI's inability to understand the subtleties of racist tropes (i.e., the use of monkeys instead of humans in some videos) make it easy to skirt the rules."

    Read more →
  • Thompson sampling

    Thompson sampling

    Thompson sampling, named after William R. Thompson, is a heuristic for choosing actions that address the exploration–exploitation dilemma in the multi-armed bandit problem. It consists of choosing the action that maximizes the expected reward with respect to a randomly drawn belief. == Description == Consider a set of contexts X {\displaystyle {\mathcal {X}}} , a set of actions A {\displaystyle {\mathcal {A}}} , and rewards in R {\displaystyle \mathbb {R} } . The aim of the player is to play actions under the various contexts, such as to maximize the cumulative rewards. Specifically, in each round, the player obtains a context x ∈ X {\displaystyle x\in {\mathcal {X}}} , plays an action a ∈ A {\displaystyle a\in {\mathcal {A}}} and receives a reward r ∈ R {\displaystyle r\in \mathbb {R} } following a distribution that depends on the context and the issued action. The elements of Thompson sampling are as follows: a likelihood function P ( r | θ , a , x ) {\displaystyle P(r|\theta ,a,x)} ; a set Θ {\displaystyle \Theta } of parameters θ {\displaystyle \theta } of the distribution of r {\displaystyle r} ; a prior distribution P ( θ ) {\displaystyle P(\theta )} on these parameters; past observations triplets D = { ( x ; a ; r ) } {\displaystyle {\mathcal {D}}=\{(x;a;r)\}} ; a posterior distribution P ( θ | D ) ∝ P ( D | θ ) P ( θ ) {\displaystyle P(\theta |{\mathcal {D}})\propto P({\mathcal {D}}|\theta )P(\theta )} , where P ( D | θ ) {\displaystyle P({\mathcal {D}}|\theta )} is the likelihood function. Thompson sampling consists of playing the action a ∗ ∈ A {\displaystyle a^{\ast }\in {\mathcal {A}}} according to the probability that it maximizes the expected reward; action a ∗ {\displaystyle a^{\ast }} is chosen with probability ∫ I [ E ( r | a ∗ , x , θ ) = max a ′ E ( r | a ′ , x , θ ) ] P ( θ | D ) d θ , {\displaystyle \int \mathbb {I} \left[\mathbb {E} (r|a^{\ast },x,\theta )=\max _{a'}\mathbb {E} (r|a',x,\theta )\right]P(\theta |{\mathcal {D}})d\theta ,} where I {\displaystyle \mathbb {I} } is the indicator function. In practice, the rule is implemented by sampling. In each round, parameters θ ∗ {\displaystyle \theta ^{\ast }} are sampled from the posterior P ( θ | D ) {\displaystyle P(\theta |{\mathcal {D}})} , and an action a ∗ {\displaystyle a^{\ast }} chosen that maximizes E [ r | θ ∗ , a ∗ , x ] {\displaystyle \mathbb {E} [r|\theta ^{\ast },a^{\ast },x]} , i.e. the expected reward given the sampled parameters, the action, and the current context. Conceptually, this means that the player instantiates their beliefs randomly in each round according to the posterior distribution, and then acts optimally according to them. In most practical applications, it is computationally onerous to maintain and sample from a posterior distribution over models. As such, Thompson sampling is often used in conjunction with approximate sampling techniques. == History == Thompson sampling was originally described by Thompson in 1933. It was subsequently rediscovered numerous times independently in the context of multi-armed bandit problems. A first proof of convergence for the bandit case has been shown in 1997. The first application to Markov decision processes was in 2000. A related approach (see Bayesian control rule) was published in 2010. In 2010 it was also shown that Thompson sampling is instantaneously self-correcting. Asymptotic convergence results for contextual bandits were published in 2011. Thompson Sampling has been widely used in many online learning problems including A/B testing in website design and online advertising, and accelerated learning in decentralized decision making. A Double Thompson Sampling (D-TS) algorithm has been proposed for dueling bandits, a variant of traditional MAB, where feedback comes in the form of pairwise comparison. == Relationship to other approaches == === Probability matching === Probability matching is a decision strategy in which predictions of class membership are proportional to the class base rates. Thus, if in the training set positive examples are observed 60% of the time, and negative examples are observed 40% of the time, the observer using a probability-matching strategy will predict (for unlabeled examples) a class label of "positive" on 60% of instances, and a class label of "negative" on 40% of instances. === Bayesian control rule === A generalization of Thompson sampling to arbitrary dynamical environments and causal structures, known as Bayesian control rule, has been shown to be the optimal solution to the adaptive coding problem with actions and observations. In this formulation, an agent is conceptualized as a mixture over a set of behaviours. As the agent interacts with its environment, it learns the causal properties and adopts the behaviour that minimizes the relative entropy to the behaviour with the best prediction of the environment's behaviour. If these behaviours have been chosen according to the maximum expected utility principle, then the asymptotic behaviour of the Bayesian control rule matches the asymptotic behaviour of the perfectly rational agent. The setup is as follows. Let a 1 , a 2 , … , a T {\displaystyle a_{1},a_{2},\ldots ,a_{T}} be the actions issued by an agent up to time T {\displaystyle T} , and let o 1 , o 2 , … , o T {\displaystyle o_{1},o_{2},\ldots ,o_{T}} be the observations gathered by the agent up to time T {\displaystyle T} . Then, the agent issues the action a T + 1 {\displaystyle a_{T+1}} with probability: P ( a T + 1 | a ^ 1 : T , o 1 : T ) , {\displaystyle P(a_{T+1}|{\hat {a}}_{1:T},o_{1:T}),} where the "hat"-notation a ^ t {\displaystyle {\hat {a}}_{t}} denotes the fact that a t {\displaystyle a_{t}} is a causal intervention (see Causality), and not an ordinary observation. If the agent holds beliefs θ ∈ Θ {\displaystyle \theta \in \Theta } over its behaviors, then the Bayesian control rule becomes P ( a T + 1 | a ^ 1 : T , o 1 : T ) = ∫ Θ P ( a T + 1 | θ , a ^ 1 : T , o 1 : T ) P ( θ | a ^ 1 : T , o 1 : T ) d θ {\displaystyle P(a_{T+1}|{\hat {a}}_{1:T},o_{1:T})=\int _{\Theta }P(a_{T+1}|\theta ,{\hat {a}}_{1:T},o_{1:T})P(\theta |{\hat {a}}_{1:T},o_{1:T})\,d\theta } , where P ( θ | a ^ 1 : T , o 1 : T ) {\displaystyle P(\theta |{\hat {a}}_{1:T},o_{1:T})} is the posterior distribution over the parameter θ {\displaystyle \theta } given actions a 1 : T {\displaystyle a_{1:T}} and observations o 1 : T {\displaystyle o_{1:T}} . In practice, the Bayesian control amounts to sampling, at each time step, a parameter θ ∗ {\displaystyle \theta ^{\ast }} from the posterior distribution P ( θ | a ^ 1 : T , o 1 : T ) {\displaystyle P(\theta |{\hat {a}}_{1:T},o_{1:T})} , where the posterior distribution is computed using Bayes' rule by only considering the (causal) likelihoods of the observations o 1 , o 2 , … , o T {\displaystyle o_{1},o_{2},\ldots ,o_{T}} and ignoring the (causal) likelihoods of the actions a 1 , a 2 , … , a T {\displaystyle a_{1},a_{2},\ldots ,a_{T}} , and then by sampling the action a T + 1 ∗ {\displaystyle a_{T+1}^{\ast }} from the action distribution P ( a T + 1 | θ ∗ , a ^ 1 : T , o 1 : T ) {\displaystyle P(a_{T+1}|\theta ^{\ast },{\hat {a}}_{1:T},o_{1:T})} . === Upper-confidence-bound (UCB) algorithms === Thompson sampling and upper-confidence bound algorithms share a fundamental property that underlies many of their theoretical guarantees. Roughly speaking, both algorithms allocate exploratory effort to actions that might be optimal and are in this sense "optimistic". Leveraging this property, one can translate regret bounds established for UCB algorithms to Bayesian regret bounds for Thompson sampling or unify regret analysis across both these algorithms and many classes of problems.

    Read more →