AI Avatar Creation

AI Avatar Creation — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

• 

  Pwnie Awards

  The Pwnie Awards are an annual awards ceremony that recognizes both excellence and incompetence in the field of information security, described by SecurityWeek as an event that "recognizes excellence and mocks incompetence in cybersecurity." Winners are selected by a committee of security industry professionals from nominations collected from the information security community. Nominees are announced yearly at Summercon, and the awards themselves are presented at the Black Hat Security Conference. == Origins == The name Pwnie Award is based on the word "pwn", which is hacker slang meaning to "compromise" or "control" based on the previous usage of the word "own" (and it is pronounced similarly). The name "The Pwnie Awards," pronounced as "Pony," is meant to sound like the Tony Awards, an awards ceremony for Broadway theater in New York City. == History == The Pwnie Awards were founded in 2007 by Alexander Sotirov and Dino Dai Zovi following discussions regarding Dino's discovery of a cross-platform QuickTime vulnerability (CVE-2007-2175) and Alexander's discovery of an ANI file processing vulnerability (CVE-2007-0038) in Internet Explorer. == Winners == === 2024 === Most Epic Fail: Crowdstrike for 2024 CrowdStrike incident Best Mobile Bug: Operation Triangulation Lamest Vendor Response: Xiaomi for obstructing Pwn2Own researchers from using their services Best Cryptographic Attack: GoFetch Best Desktop Bug: forcing realtime WebAudio playback in Chrome (CVE-2023-5996) Best Song: Touch Some Grass by UwU Underground Best Privilege Escalation: Windows Streaming Service UAF (CVE-2024-30089) by Valentina Palmiotti (chompie) Best Remote Code Execution: Microsoft Message Queuing (MSMQ) Remote Code Execution Vulnerability (CVE-2024-30080) Most Epic Achievement: Discovery and reverse engineering of the XZ Utils backdoor Most Innovative Research: Let the Cache Cache and Let the WebAssembly Assemble: Knocking’ on Chrome’s Shell by Edouard Bochin, Tao Yan, and Bo Qu Most Underhyped Research: See No Eval: Runtime Dynamic Code Execution in Objective-C === 2023 === Best Desktop Bug: CountExposure! by RyeLv(@b2ahex) Best Cryptographic Attack: Video-based cryptanalysis: Extracting Cryptographic Keys from Video Footage of a Device’s Power LED by Ben Nassi, Etay Iluz, Or Cohen, Ofek Vayner, Dudi Nassi, Boris Zadov, Yuval Elovici Best Song: Clickin’ Most Innovative Research: Inside Apple’s Lightning: Jtagging the iPhone for Fuzzing and Profit Most Under-Hyped Research: Activation Context Cache Poisoning Best Privilege Escalation Bug: URB Excalibur: Slicing Through the Gordian Knot of VMware VM Escapes Best Remote Code Execution Bug: ClamAV RCE Lamest Vendor Response: Three Lessons From Threema: Analysis of a Secure Messenger Most Epic Fail: “Holy fucking bingle, we have the no fly list,” Epic Achievement: Clement Lecigne: 0-days hunter world champion Lifetime Achievement Award: Mudge === 2022 === Lamest Vendor Response: Google's "TAG" response team for "unilaterally shutting down a counterterrorism operation." Epic Achievement: Yuki Chen’s Windows Server-Side RCE Bugs Most Epic Fail: HackerOne Employee Caught Stealing Vulnerability Reports for Personal Gains Best Desktop Bug: Pietro Borrello, Andreas Kogler, Martin Schwarzl, Moritz Lipp, Daniel Gruss, Michael Schwarz for Architecturally Leaking Data from the Microarchitecture Most Innovative Research: Pietro Borrello, Martin Schwarzl, Moritz Lipp, Daniel Gruss, Michael Schwarz for Custom Processing Unit: Tracing and Patching Intel Atom Microcode Best Cryptographic Attack: Hertzbleed: Turning Power Side-Channel Attacks Into Remote Timing Attacks on x86 by Yingchen Wang, Riccardo Paccagnella, Elizabeth Tang He, Hovav Shacham, Christopher Fletcher, David Kohlbrenner Best Remote Code Execution Bug: KunlunLab for Windows RPC Runtime Remote Code Execution (CVE-2022-26809) Best Privilege Escalation Bug: Qidan He of Dawnslab, for Mystique in the House: The Droid Vulnerability Chain That Owns All Your Userspace Best Mobile Bug: FORCEDENTRY Most Under-Hyped Research: Yannay Livneh for Spoofing IP with IPIP Best Song: Dialed Up by Project Mammoth === 2021 === Lamest Vendor Response: Cellebrite, for their response to Moxie, the creator of Signal, reverse-engineering their UFED and accompanying software and reporting a discovered exploit. Epic Achievement: Ilfak Guilfanov, in honor of IDA's 30th Anniversary. Best Privilege Escalation Bug: Baron Samedit of Qualys, for the discovery of a 10-year-old exploit in sudo. Best Song: The Ransomware Song by Forrest Brazeal Best Server-Side Bug: Orange Tsai, for his Microsoft Exchange Server ProxyLogon attack surface discoveries. Best Cryptographic Attack: The NSA for its disclosure of a bug in the verification of signatures in Windows which breaks the certificate trust chain. Most Innovative Research: Enes Göktaş, Kaveh Razavi, Georgios Portokalidis, Herbert Bos, and Cristiano Giuffrida at VUSec for their research on the "BlindSide" Attack. Most Epic Fail: Microsoft, for their failure to fix PrintNightmare. Best Client-Side Bug: Gunnar Alendal's discovery of a buffer overflow on the Samsung Galaxy S20's secure chip. Most Under-Hyped Research: The Qualys Research Team for 21Nails, 21 vulnerabilities in Exim, the Internet's most popular mail server. === 2020 === Best Server-Side Bug: BraveStarr (CVE-2020-10188) – A Fedora 31 netkit telnetd remote exploit (Ronald Huizer') Best Privilege Escalation Bug: checkm8 – A permanent unpatchable USB bootrom exploit for a billion iOS devices. (axi0mX) Epic Achievement: "Remotely Rooting Modern Android Devices" (Guang Gong) Best Cryptographic Attack: Zerologon vulnerability (Tom Tervoort, CVE-2020-1472) Best Client-Side Bug: RCE on Samsung Phones via MMS (CVE-2020-8899 and -16747), a zero click remote execution attack. (Mateusz Jurczyk) Most Under-Hyped Research: Vulnerabilities in System Management Mode (SMM) and Trusted Execution Technology (TXT) (CVE-2019-0151 and -0152) (Gabriel Negreira Barbosa, Rodrigo Rubira Branco, Joe Cihula) Most Innovative Research: TRRespass: When Memory Vendors Tell You Their Chips Are Rowhammer-free, They Are Not. (Pietro Frigo, Emanuele Vannacci, Hasan Hassan, Victor van der Veen, Onur Mutlu, Cristiano Giuffrida, Herbert Bos, Kaveh Razavi) Most Epic Fail: Microsoft; for the implementation of Elliptic-curve signatures which allowed attackers to generate private pairs for public keys of any signer, allowing HTTPS and signed binary spoofing. (CVE-2020-0601) Best Song: Powertrace by Rebekka Aigner, Daniel Gruss, Manuel Weber, Moritz Lipp, Patrick Radkohl, Andreas Kogler, Maria Eichlseder, ElTonno, tunefish, Yuki and Kater Lamest Vendor Response: Daniel J. Bernstein (CVE-2005-1513) === 2019 === Best Server-Side Bug: Orange Tsai and Meh Chang, for their SSL VPN research. Most Innovative Research: Vectorized Emulation Brandon Falk Best Cryptographic Attack: \m/ Dr4g0nbl00d \m/ Mathy Vanhoef, Eyal Ronen Lamest Vendor Response: Bitfi Most Over-hyped Bug: Allegations of Supermicro hardware backdoors, Bloomberg Most Under-hyped Bug: Thrangrycat, (Jatin Kataria, Red Balloon Security) === 2018 === Most Innovative Research: Spectre/Meltdown (Paul Kocher, Jann Horn, Anders Fogh, Daniel Genkin, Daniel Gruss, Werner Haas, Mike Hamburg, Moritz Lipp, Stefan Mangard, Thomas Prescher, Michael Schwarz, Yuval Yarom) Best Privilege Escalation Bug: Spectre/Meltdown (Paul Kocher, Jann Horn, Anders Fogh, Daniel Genkin, Daniel Gruss, Werner Haas, Mike Hamburg, Moritz Lipp, Stefan Mangard, Thomas Prescher, Michael Schwarz, Yuval Yarom) Lifetime Achievement: Michał Zalewski Best Cryptographic Attack: ROBOT - Return Of Bleichenbacher’s Oracle Threat Hanno Böck, Juraj Somorovsky, Craig Young Lamest Vendor Response: Bitfi hardware crypto-wallet, after the "unhackable" device was hacked to extract the keys required to steal coins and rooted to play Doom. === 2017 === Epic Achievement: Federico Bento for Finally getting TIOCSTI ioctl attack fixed Most Innovative Research: ASLR on the line Ben Gras, Kaveh Razavi, Erik Bosman, Herbert Bos, Cristiano Giuffrida Best Privilege Escalation Bug: DRAMMER Victor van der Veen, Yanick Fratantonio, Martina Lindorfer, Daniel Gruss, Clementine Maurice, Giovanni Vigna, Herbert Bos, Kaveh Razavi, Cristiano Giuffrida Best Cryptographic Attack: The first collision for full SHA-1 Marc Stevens, Elie Bursztein, Pierre Karpman, Ange Albertini, Yarik Markov Lamest Vendor Response: Lennart Poettering - for mishandling security vulnerabilities most spectacularly for multiple critical Systemd bugs Best Song: Hello (From the Other Side) - Manuel Weber, Michael Schwarz, Daniel Gruss, Moritz Lipp, Rebekka Aigner === 2016 === Most Innovative Research: Dedup Est Machina: Memory Deduplication as an Advanced Exploitation Vector Erik Bosman, Kaveh Razavi, Herbert Bos, Cristiano Giuffrida Lifetime Achievement: Peiter Zatko aka Mudge Best Cryptographic Attack: DROWN attack Nimrod Aviram et al. Best Song: Cyberlier - Katie Mous

  Read more →
• 

  Global Partnership on Artificial Intelligence

  The Global Partnership on Artificial Intelligence (GPAI, pronounced "gee-pay") is an international initiative established to guide the responsible development and use of artificial intelligence (AI) in a manner that respects human rights and the shared democratic values of its members. The partnership was first proposed by Canada and France at the 2018 44th G7 summit, and officially launched in June 2020. GPAI is hosted by the Organisation for Economic Co-operation and Development (OECD). GPAI seeks to bridge the gap between theory and practice by supporting research and applied activities in areas that are directly relevant to policymakers in the realm of AI. It brings together experts from industry, civil society, governments, and academia to collaborate on the challenges and opportunities presented by artificial intelligence. == History == The Global Partnership on Artificial Intelligence was announced on the margins of the 2018 G7 Summit by Canadian Prime Minister Justin Trudeau and French President Emmanuel Macron. It officially launched on June 15, 2020 with fifteen founding members: Australia, Canada, France, Germany, India, Italy, Japan, Mexico, New Zealand, the Republic of Korea, Singapore, Slovenia, the United Kingdom, the United States, and the European Union. The Organisation for Economic Co-operation and Development (OECD) hosts a dedicated secretariat to support GPAI's governing bodies and activities. UNESCO joined the partnership in December 2020 as an observer. On November 11, 2021, Czechia, Israel and few more EU countries also joined the GPAI, bringing the total membership to 25 countries. Since the November 2022 summit, the list of members stands at 29. Austria, Chile, Finland, Malaysia, Norway, Slovakia and Switzerland were invited. The seven, however, are pending membership approval. == Membership == The following 29 members of the GPAI are: Argentina Australia Belgium Brazil Canada Czech Republic Denmark France Germany India Ireland Israel Italy Japan Mexico Netherlands New Zealand Poland Republic of Korea Senegal Serbia Singapore Slovenia Spain Sweden Turkey United Kingdom United States European Union Invited members: Austria (pending membership approval) Chile (pending membership approval) Finland (pending membership approval) Malaysia (pending membership approval) Norway (pending membership approval) Slovakia (pending membership approval) Switzerland (pending membership approval) == Organization == GPAI's experts collaborate across several Working Groups themes: Responsible AI (including an ad-hoc subgroup on AI and Pandemic Response), Data Governance, Future of Work, and Innovation & Commercialization. GPAI's Working Groups are supported by two Centres of Expertise: one in Montreal that supports the first two Working Groups, and one in Paris that supports the latter two. It also has a Steering Committee, the elected chair of which has also been to date elected chair of the Multi Stakeholder Group (MEG). These chairs have been: Jordan Zed and Baroness Joanna Shields (Shields, MEG chair; 2020-2021), Joanna Shields and Renaud Vedel (Shields, MEG chair; 2021-2022), Yoichi Iida and Inma Martinez (Martinez, MEG chair; 2023-2024) GPAI has a rotating presidency and host (much like the G7). The presidencies to date have been: Canada (2020) France (2021) Japan (2022) India (2023)

  Read more →
• 

  MuZero

  MuZero is a computer program developed by artificial intelligence research company DeepMind, a subsidiary of Google, to master games without knowing their rules and underlying dynamics. Its release in 2019 included benchmarks of its performance in Go, chess, shogi, and a suite of 57 different Atari games. The algorithm uses an approach similar to AlphaZero, where a combination of a tree-based search and a learned model is deployed. It matched AlphaZero's performance in chess and shogi, improved on its performance in Go, and improved on the state of the art in mastering a suite of 57 Atari games (the Arcade Learning Environment), a visually-complex domain. MuZero was trained via self-play, with no access to rules, opening books, or endgame tablebases. The trained algorithm used the same convolutional and residual architecture as AlphaZero, but with 20 percent fewer computation steps per node in the search tree. == History == MuZero really is discovering for itself how to build a model and understand it just from first principles. On November 19, 2019, the DeepMind team released a preprint introducing MuZero. === Derivation from AlphaZero === MuZero (MZ) is a combination of the high-performance planning of the AlphaZero (AZ) algorithm with approaches to model-free reinforcement learning. The combination allows for more efficient training in classical planning regimes, such as Go, while also handling domains with much more complex inputs at each stage, such as visual video games. MuZero was derived directly from AZ code, sharing its rules for setting hyperparameters. Differences between the approaches include: AZ's planning process uses a simulator. The simulator knows the rules of the game. It has to be explicitly programmed. A neural network then predicts the policy and value of a future position. Perfect knowledge of game rules is used in modeling state transitions in the search tree, actions available at each node, and termination of a branch of the tree. MZ does not have access to the rules, and instead learns one with neural networks. AZ has a single model for the game (from board state to predictions); MZ has separate models for representation of the current state (from board state into its internal embedding), dynamics of states (how actions change representations of board states), and prediction of policy and value of a future position (given a state's representation). MZ's hidden model may be complex, and it may turn out it can host computation; exploring the details of the hidden model in a trained instance of MZ is a topic for future exploration. MZ does not expect a two-player game where winners take all. It works with standard reinforcement-learning scenarios, including single-agent environments with continuous intermediate rewards, possibly of arbitrary magnitude and with time discounting. AZ was designed for two-player games that could be won, drawn, or lost. === Comparison with R2D2 === The previous state of the art technique for learning to play the suite of Atari games was R2D2, the Recurrent Replay Distributed DQN. MuZero surpassed both R2D2's mean and median performance across the suite of games, though it did not do better in every game. == Training and results == MuZero used 16 third-generation tensor processing units (TPUs) for training, and 1000 TPUs for selfplay for board games, with 800 simulations per step and 8 TPUs for training and 32 TPUs for selfplay for Atari games, with 50 simulations per step. AlphaZero used 64 second-generation TPUs for training, and 5000 first-generation TPUs for selfplay. As TPU design has improved (third-generation chips are 2x as powerful individually as second-generation chips, with further advances in bandwidth and networking across chips in a pod), these are comparable training setups. R2D2 was trained for 5 days through 2M training steps. === Initial results === MuZero matched AlphaZero's performance in chess and shogi after roughly 1 million training steps. It matched AZ's performance in Go after 500,000 training steps and surpassed it by 1 million steps. It matched R2D2's mean and median performance across the Atari game suite after 500 thousand training steps and surpassed it by 1 million steps, though it never performed well on 6 games in the suite. == Reactions and related work == MuZero was viewed as a significant advancement over AlphaZero, and a generalizable step forward in unsupervised learning techniques. The work was seen as advancing understanding of how to compose systems from smaller components, a systems-level development more than a pure machine-learning development. While only pseudocode was released by the development team, Werner Duvaud produced an open source implementation based on that. MuZero has been used as a reference implementation in other work, for instance as a way to generate model-based behavior. In late 2021, a more efficient variant of MuZero was proposed, named EfficientZero. It "achieves 194.3 percent mean human performance and 109.0 percent median performance on the Atari 100k benchmark with only two hours of real-time game experience". In early 2022, a variant of MuZero was proposed to play stochastic games (for example 2048, backgammon), called Stochastic MuZero, which uses afterstate dynamics and chance codes to account for the stochastic nature of the environment when training the dynamics network.

  Read more →
• 

  Shane Legg

  Shane Legg (born 1973 or 1974) is a machine learning researcher and entrepreneur. With Demis Hassabis and Mustafa Suleyman, he cofounded DeepMind Technologies (later bought by Google and now called Google DeepMind), and works there as the chief AGI scientist. He is also known for his academic work on artificial general intelligence, including his thesis supervised by Marcus Hutter. == Early life and education == Legg attended Rotorua Lakes High School in Rotorua, on New Zealand's North Island. He completed his undergraduate studies at Waikato University in 1996. Also in 1996, he obtained his MSc degree with a thesis entitled "Solomonoff Induction", with Cristian S. Calude at the University of Auckland. == Research interests == In the early 2000s, Legg re-introduced and popularized with Ben Goertzel the term "artificial general intelligence" (AGI), to describe an AI that can do practically any cognitive task a human can do. At that time, talking about AGI "would put you on the lunatic fringe". Legg is known for his concern of existential risk from AI, highlighted in 2011 in an interview on LessWrong and in 2023 he signed the statement on AI risk of extinction. == Career == Before his PhD and before cofounding DeepMind, Shane Legg worked at "a number of software development positions at private companies", including the "big data firm Adaptive Intelligence" and the startup WebMind founded by Ben Goertzel. === Research === Legg later obtained a PhD at the Dalle Molle Institute for Artificial Intelligence Research (IDSIA), a joint research institute of USI Università della Svizzera italiana and SUPSI. He worked on theoretical models of super intelligent machines (AIXI) with Marcus Hutter, and completed in 2008 his doctoral thesis entitled "Machine Super Intelligence". He then went on to complete a postdoctoral fellowship in finance at USI, and began a further fellowship at University College London's Gatsby Computational Neuroscience Unit. === DeepMind === Demis Hassabis and Shane Legg first met in 2009 at University College London, where Legg was a postdoctoral researcher. In 2010, Legg cofounded the start-up DeepMind Technologies along with Demis Hassabis and Mustafa Suleyman. DeepMind Technologies was bought in 2014 by Google. After the merge with Google Brain in 2023, the company is now known as Google DeepMind. According to a 2017 article, a significant part of his job as the chief scientist was to supervise recruitment, to decide where DeepMind should focus its efforts, and to lead DeepMind's AI safety work. As of July 2023, Legg works at Google DeepMind as the Chief AGI Scientist. == Awards and honors == Legg was awarded the $10,000 prize of the Singularity Institute for Artificial Intelligence for his PhD done in 2008. Legg was appointed Commander of the Order of the British Empire (CBE) in the 2019 Birthday Honours for services to the science and technology sector and to investment.

  Read more →
• 

  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 →
• 

  RealSense

  RealSense is an American technology company that develops depth cameras and computer-vision systems used in robotics, access control, industrial automation and healthcare. The company’s stereoscopic 3D cameras and software are marketed as a perception platform for “physical AI”, particularly for humanoid robots and autonomous mobile robots (AMRs). RealSense was incubated for more than a decade inside Intel’s perceptual computing and depth-sensing group before being spun out as an independent company in July 2025 with a US$50 million Series A round backed by a semiconductor-focused private equity firm and strategic investors including Intel Capital and the MediaTek Innovation Fund. Following the spin-out, RealSense announced a strategic collaboration with Nvidia to integrate its AI depth cameras with the Nvidia Jetson Thor robotics platform, the Isaac Sim simulation environment and the Holoscan Sensor Bridge for low-latency sensor fusion. In November 2025, Swiss access-solutions provider dormakaba acquired a minority stake in RealSense and formed a partnership to develop AI-powered biometric access-control and security systems for data centres, airports and other critical infrastructure. == History == === Origins in Intel Perceptual Computing === Intel began developing depth-sensing and perceptual-computing technologies in the early 2010s under the Perceptual Computing brand, with research spanning gesture control, facial recognition and eye-tracking systems. The work led to a series of 3D cameras and developer challenge programmes intended to stimulate software ecosystems for natural-user interfaces. In 2014 Intel rebranded the effort as Intel RealSense, positioning the technology as a family of depth cameras and vision processors for PCs, mobile devices and embedded systems. Early devices such as the F200 and R200 were integrated into laptops and tablets from OEMs including Asus, HP, Dell, Lenovo and Acer, and were also sold as standalone webcams by partners such as Razer and Creative. === Refocus on robotics and near-closure === By the late 2010s Intel had steered RealSense away from mainstream PC peripherals toward robotics, industrial and embedded applications, adding stereo and lidar-based depth cameras to the portfolio. In August 2021, trade publication CRN reported that Intel planned to wind down the RealSense business as part of a broader restructuring, raising questions about the future of the product line. Despite that announcement, Intel continued to invest in new custom silicon for depth cameras, and RealSense remained widely used in mobile robots and automation projects. === Spin-out as RealSense Inc. (2025) === On 11 July 2025, Intel completed the spin-out of its RealSense 3D-camera business into a new privately held company, RealSense Inc., and the new entity announced a US$50 million Series A funding round. The round was led by a semiconductor-focused private equity investor with participation from Intel Capital, MediaTek Innovation Fund and other strategics. Independent coverage described RealSense as serving more than 3,000 active customers and supplying depth cameras to a large share of global AMR and humanoid robot platforms. The company stated that it would continue to support the existing Intel RealSense product roadmap while accelerating development of AI-enabled cameras and perception software. === Strategic partnerships and investments === In October 2025 RealSense and Nvidia announced a strategic collaboration centered on integrating RealSense AI depth cameras with Nvidia’s Jetson Thor robotics compute modules, the Isaac Sim simulation environment and the Holoscan Sensor Bridge for multi-sensor streaming. The collaboration is positioned as enabling “physical AI” workloads such as whole-body humanoid control, real-time mapping and safety-critical human–robot interaction. On 19 November 2025, dormakaba announced that it had acquired a minority stake in RealSense and entered into a partnership to co-develop intelligent access-control solutions, including biometric gates for airports and enterprise facilities. The partnership aims to combine RealSense’s depth and facial-authentication technology with dormakaba’s installed base of sensors, doors and turnstiles. == Products == === Depth-camera families === RealSense’s products are sold as modular components (depth modules, vision processors and complete cameras) and as integrated systems with on-device AI. The company continues to offer and support the Intel RealSense D400 family of active-stereo depth cameras (including the D415, D435 and D455), which are widely used in robotics and automation. These devices combine a RealSense Vision Processor from the D4 family with dual infrared imagers and, on some models, an RGB camera. Earlier generations of Intel RealSense cameras, including the F200, R200, SR300 and the L515 lidar camera, remain in use in niche and legacy applications but are no longer the focus of the independent company’s roadmap. === D555 PoE depth camera === The first new hardware platform announced after the spin-out was the RealSense Depth Camera D555, a ruggedised stereo-depth device aimed at industrial and robotics deployments. The D555 uses the longer-range D450 optical module with a global shutter and integrates RealSense’s Vision SoC V5, a new generation of vision processor optimised for neural-network inference and depth computation. Key features highlighted in technical coverage include: Power over Ethernet (PoE), allowing power and data to be delivered over a single cable and supporting both RJ45 and ruggedised M12 connections; an IP-rated enclosure designed for harsh indoor and outdoor environments; a built-in inertial measurement unit (IMU) to support simultaneous localisation and mapping (SLAM) and motion tracking; native support for ROS 2 and integration with the open-source RealSense SDK. According to independent reporting, the D555 is used in AI-enabled embedded-vision applications in mobile robots and fixed industrial systems, and was among the first RealSense products to be tightly integrated with Nvidia’s Jetson Thor and Holoscan platforms for low-latency sensor fusion. === Software and SDK === RealSense cameras are supported by a cross-platform, open-source software stack historically branded as Intel RealSense SDK 2.0. The SDK provides device drivers, depth and point-cloud processing, tracking and calibration tools, and bindings for languages such as C++, Python and C#. The independent company has continued to maintain and extend the SDK for new hardware, including D555 and other Vision SoC V5-based devices, and publishes reference integrations for ROS 2 and industrial-automation frameworks. === Biometrics and access-control products === In addition to general-purpose depth cameras, RealSense offers facial-authentication hardware and software, commonly referred to as RealSense ID, for biometric access control and identity verification. These products combine an active depth sensor with a dedicated neural-network pipeline running on embedded processors, aimed at applications such as secure doors, turnstiles and kiosks. Use-case material published by partners describes deployments of RealSense-based biometric readers in school lunch programmes, agricultural biosecurity checkpoints and enterprise facilities. The dormakaba partnership announced in 2025 extends this portfolio to integrated biometric gates and sensor-equipped doors in airports and data centres. == Applications == === Robotics and automation === RealSense depth cameras are used in autonomous mobile robots, humanoid robots, drones and industrial automation systems for tasks such as obstacle avoidance, navigation and manipulation. Reuters reported in 2025 that RealSense cameras were embedded in around 60 percent of the world’s AMRs and humanoid robots, citing customers including Unitree Robotics and ANYbotics. Developers and integrators use RealSense systems with platforms such as Nvidia Jetson, ROS and proprietary motion-planning stacks. === Biometrics and security === RealSense technology is also applied in biometric access control and surveillance, where depth and infrared imaging are used to improve anti-spoofing performance for facial recognition. The dormakaba investment and collaboration is aimed at integrating these capabilities into boarding gates, staff entrances and secure facilities, with RealSense providing perception hardware and algorithms and dormakaba providing access-control infrastructure and global distribution. == Reception == Early coverage of Intel RealSense for consumer PCs noted that the technology’s impact would depend on the availability of compelling software and use cases for depth-sensing cameras. Later reporting on the spin-out has characterised the new company as part of a broader wave of investment in robotics and physical AI, with some analysts suggesting that RealSense’s installed base and patent portfolio give it an advantage as dep

  Read more →
• 

  KataGo

  KataGo is a free and open-source computer Go program, capable of defeating top-level human players. First released on 27 February 2019, it is developed by David Wu, who also developed the Arimaa playing program bot_Sharp which defeated three top human players to win the Arimaa AI Challenge in 2015. KataGo's first release was trained by David Wu using resources provided by his employer Jane Street Capital, but it is now trained by a distributed effort. Members of the computer Go community provide computing resources by running the client, which generates self-play games and rating games, and submits them to a server. The self-play games are used to train newer networks and the rating games to evaluate the networks' relative strengths. KataGo supports the Go Text Protocol, with various extensions, thus making it compatible with popular GUIs such as Lizzie. As an alternative, it also implements a custom "analysis engine" protocol, which is used by the KaTrain GUI, among others. KataGo is widely used by strong human go players, including the South Korean national team, for training purposes. KataGo is also used as the default analysis engine in the online Go website AI Sensei, as well as OGS (the Online Go Server). == Technology == Based on techniques used by DeepMind's AlphaGo Zero, KataGo implements Monte Carlo tree search with a convolutional neural network providing position evaluation and policy guidance. Compared to AlphaGo, KataGo introduces many refinements that enable it to learn faster and play more strongly. Notable features of KataGo that are absent in many other Go-playing programs include score estimation; support for small boards, rectangular boards, and large boards; arbitrary values of komi and handicaps; and the ability to use various Go rulesets and adjust its play and evaluation for the small differences between them. === Network === The network used in KataGo are ResNets with pre-activation. While AlphaGo Zero has only game board history as input features (as it was designed as a general architecture for board games, subsequently becoming AlphaZero), the input to the network contains additional features designed by hand specifically for playing Go. These features include liberties, komi parity, pass-alive, and ladders. The trunk is essentially the same as in AlphaGo Zero, but with global pooling layers added to allow the network to be conditioned on global context such as ko fights. This is similar to the Squeeze-and-Excitation Network. The network has two heads: a policy head and a value head. The policy and value heads are mostly the same as in AlphaGo Zero, but both heads have auxiliary subheads to provide auxiliary loss signal for faster training: Policy head: predicts policy for the current player's move this turn, and the opponent player's move in the next turn. A policy Each is a logit array of size 19 × 19 + 1 {\displaystyle 19\times 19+1} , representing the logit of making a move in one of the points, plus the logit of passing. Value head: predicts game outcome, expected score difference, expected board ownership, etc. The network is described in detail in Appendix A of the report. The code base switched from using TensorFlow to PyTorch in version 1.12. === Training === Let its trunk have b {\displaystyle b} residual blocks and c {\displaystyle c} channels. During its first training run, multiple networks were trained with increasing ( b , c ) {\displaystyle (b,c)} . It took 19 days using a maximum of 28 Nvidia V100 GPUs at 4.2 million games. After the first training run, training became a distributed project run by volunteers, with increasing network sizes. As of August 2024, it has reached b28c512 (28 blocks, 512 channels). == Adversarial attacks == In 2022, KataGo was used as the target for adversarial attack research, designed to demonstrate the "surprising failure modes" of AI systems. The researchers were able to trick KataGo into ending the game prematurely. Adversarial training improves defense against adversarial attacks, though not perfectly.

  Read more →
• 

  Sriram Krishnan

  Sriram Krishnan (born 1984) is a tech executive and White House official, currently serving as the Senior White House Policy Advisor on Artificial Intelligence. Krishnan was named a Time Person of the Year in 2025 as an "Architect of Artificial Intelligence." He was described in Time as providing the "wake-up call that we needed" to the other AI builders, leading to "a multiyear, $500 billion initiative dubbed Stargate" to push American-made AI, as well as numerous other AI initiatives. Also in December 2025, President Trump said of Krishnan, "without him, things on AI would not function well" and cited Krishnan as the leading figure behind the American executive order on AI. As the leader of the United States' policy team regarding artificial intelligence, Krishnan plays "a significant role in shaping the administration’s approach to AI and driving measures to advance federal adoption of AI." The role calls for removing barriers to AI adoption within the government, driving vendors toward solutions suitable for federal needs, designing sensible regulation of private-sector AI, and conducting "AI diplomacy". He has stated a policy goal of "reinvigorating US dominance in emerging technologies," including AI. He also represents the United States' interests in AI abroad, such as at the Paris AI Summit. He is one of the authors of the American "AI Action Plan" released in July, 2025, which he contends is necessary to win the "existential race with China" for AI supremacy. Krishnan, a U.S. citizen born in India, is also a venture capitalist, podcaster, product manager and author. Early in his career, he led product teams at Microsoft, Twitter, Yahoo!, Facebook, and Snap. In addition to his work as an investor and technologist, he and his wife, Aarthi Ramamurthy, rose to additional prominence in 2021 as podcast hosts. He served as a general partner at the venture capital firm Andreessen Horowitz and led its London office. In 2022, Krishnan announced that he was working with Elon Musk on the rebuilding of Twitter following Musk's acquisition of the company. On December 22, 2024, US president-elect Donald Trump announced that Krishnan would be Senior White House Policy Advisor on Artificial Intelligence in his incoming administration; in 2026 he joined the National Economic Council. == Early life and education == Krishnan was born in Chennai, India. He earned his Bachelor of Technology in Information Technology from SRM University (2001–2005), moved to the United States in 2007 to join Microsoft, and became a naturalized U.S. citizen in 2016. == Career == === Early career === In 2007, he began working at Microsoft where he served as a program manager for Visual Studio. At Facebook, Krishnan built the Facebook Audience Network, a competitive platform to Google's ad technologies. At Twitter, he led product and core user experience, driving a 20% annual user growth rate and launching a redesigned home page and events experience. === Andreessen Horowitz === Krishnan was appointed a general partner of American venture capital firm Andreessen Horowitz ("a16z") in February 2021. He was anticipated to serve consumer and social markets, however he has also theorized on the impact of "deep tech" on society. In 2023 he was appointed to lead the firm's London office, its first non-US location. The office is expected to serve Web3 investments as well as AI and other fields. Krishnan announced that he would leave the firm at the end of 2024. === Social media and AI === In 2022, various news media reported that Krishnan was assisting Elon Musk in the revamp of Twitter following Musk's takeover of the company. Additional reports named Krishnan as the leading candidate for the role of CEO of the newly private company. Krishnan penned a 2023 New York Times opinion column regarding social media, AI, and related fields. He predicted a rise in the number and diversity of online spaces due to decentralization and platforms like Farcaster, Bluesky and Mastodon. === Public office === In 2024, the Financial Times reported that Krishnan was active in international affairs, reintroducing Boris Johnson to Elon Musk, following Musk's nomination to the proposed Department of Government Efficiency. Krishnan was also reported as potentially leaving a16z at the end of the year to "be jumping into something I've wanted to spend [his] energy on," which was widely reported as being related to Musk's and Vivek Ramaswamy's work at DOGE. Others reported to be involved include Joe Lonsdale, Marc Andreesen, Bill Ackman, and Travis Kalanick. On December 22, 2024, US president-elect Donald Trump announced that he would be Senior White House Policy Advisor on Artificial Intelligence in his incoming administration. On February 6, 2025, Reuters reported that Krishnan would be accompanying Vice President Vance to the Paris AI Summit, a "major artificial intelligence" event later that month. Other members of the White House Office of Science and Technology Policy would also be joining the event with around 100 other countries to "focus on AI's potential." Krishnan joined a U.S. technology policy delegation to the Middle East in advance of President Trump's visit in May 2025. Conducting "AI diplomacy," Krishnan negotiated the spread of U.S. AI technologies with Crown Prince Mohammed bin Salman of Saudi Arabia, as well as other means to strengthen bilateral trade in artificial intelligence technologies. He explained that the goal of the diplomatic mission was that "we want American A.I. to spread." Krishnan, along with David Sacks and Michael Kratsios, were credited as authors of the American AI Action Plan released in July 2025. The plan is "the administration’s most significant policy directive" regarding artificial intelligence; it calls for financing to support the global spread of American AI models and a policy to enforce neutrality in models. The Washington Post referred to the plan as a "bold action to ensure that American AI remains at the cutting edge." The AI Action Plan is a continuation of prior efforts to reduce barriers to U.S. production of AI systems and the removal of rules that were considered to hinder such growth. Later in 2025, at the POLITICO AI & Tech Summit, Krishnan called national AI development "an existential race with China." He suggested that private companies are best positioned to create new models, quipping "let them cook." He further suggested that state-by-state regulation of AI technologies may hinder national AI competitiveness. Also in 2025, at the Axios AI+ Summit, Krishnan stated that the United States and China are in a race for AI supremacy, in which the winner will be judged by market share. Winning the race is a "business strategy" to Krishnan. Krishnan was named in the 2025 Time Person of the Year article as an "AI Architect". === The Aarthi and Sriram Show and other media === In early 2021, Krishnan and his wife, Aarthi Ramamurthy, launched a Clubhouse talk show that "focuses on organic conversations on anything from startups to venture capitalism and cryptocurrencies." An early appearance by Elon Musk on the Good Time Show was described as the first show that "broke Clubhouse" by rapidly exceeding the limit of 5,000 simultaneous users. The desire to interact with a larger community led to a variety of later innovations to allow streaming and replaying of Clubhouse chats. On that episode, Elon Musk grilled Robinhood CEO Vlad Tenev regarding the GameStop trading controversy. As of December 2021, the show had over 187,000 subscribers, plus 735,000 subscribers between Krishnan and Ramamurthy's personal Clubhouse accounts. Other guests have included Facebook CEO Mark Zuckerberg, Diane von Fürstenberg, Tony Hawk, MrBeast, and A.R. Rahman. In 2022, the Good Time Show moved to YouTube. It then evolved to a podcasting format under the name The Aarthi and Sriram Show, with both audio and video content. The Hollywood Reporter reported that the podcast had received more than 1 million downloads by early 2023. == Personal life == Krishnan is married to Aarthi Ramamurthy, co-host of The Aarthi and Sriram Show (formerly the Good Time Show) and a serial entrepreneur. They met in college in 2003 through a Yahoo! chat room related to a coding project and began dating in 2006 and eloped in 2010. == Awards == Time Person of the Year - 2025

  Read more →
• 

  Spotify Kids

  Spotify Kids is a Swedish kid-friendly Music streaming service developed by Spotify. It offers curated content for children, including music, audiobooks, lullabies, and bedtime stories, while providing their parents with parental controls. The service is only available to subscribers to Spotify's Premium Family subscription plan. == Function == Spotify Kids is a Swedish Kid-friendly Music Streaming Service that allows children to browse Spotify with parental controls. Using the app, parents can view their children's listening history, block specific songs, and share playlists with their children. The app also includes sing-along songs, playlists designed for young children, and curated audiobooks, lullabies, and bedtime stories. Access is included in Spotify's Premium Family subscription plan, and is exclusive to subscribers to the plan. Users can configure the app for a specific age group upon first launch. The playlists on Spotify Kids are curated by groups including Discovery Kids, Nickelodeon, Universal Pictures, and The Walt Disney Company. All content on the Spotify Kids app is curated by editors. As of March 2021, there were roughly 8,000 songs available on the platform. The design of the Spotify Kids app is colorful, and user interface varies depending on the age group for which the app is configured. Spotify Kids is designed to comply with consent and data collection regulations for apps used by children. TechCrunch explains that it is "designed on a grand scale to drive subscriptions to Spotify's top-tier $14.99-per-month Premium Family Plan." == Release == After being beta tested in Ireland in October 2019, it was released as a beta across the United Kingdom on February 11, 2020. It was later released in Sweden, Denmark, Australia, New Zealand, Mexico, Argentina, and Brazil. On March 31, 2021, it was made available in France, Canada, and the United States.

  Read more →
• 

  Dendral

  Dendral was a project in artificial intelligence (AI) of the 1960s, and the computer software expert system that it produced. Its primary aim was to study hypothesis formation and discovery in science. For that, a specific task in science was chosen: help organic chemists in identifying unknown organic molecules, by analyzing their mass spectra and using knowledge of chemistry. It was done at Stanford University by Edward Feigenbaum, Bruce G. Buchanan, Joshua Lederberg, and Carl Djerassi, along with a team of highly creative research associates and students. It began in 1964 and spans approximately half the history of AI research. The software program Dendral is considered the first expert system because it automated the decision-making process and problem-solving behavior of organic chemists. The project consisted of research on two main programs Heuristic Dendral and Meta-Dendral, and several sub-programs. It was written in the Lisp programming language, which was considered the language of AI because of its flexibility. Many systems were derived from Dendral, including MYCIN, MOLGEN, PROSPECTOR, XCON, and STEAMER. There are many other programs today for solving the mass spectrometry inverse problem, see List of mass spectrometry software, but they are no longer described as 'artificial intelligence', just as structure searchers. The name Dendral is an acronym of the term "Dendritic Algorithm". == Heuristic Dendral == Heuristic Dendral is a program that uses mass spectra or other experimental data together with a knowledge base of chemistry to produce a set of possible chemical structures that may be responsible for producing the data. A mass spectrum of a compound is produced by a mass spectrometer, and is used to determine its molecular weight, the sum of the masses of its atomic constituents. For example, the compound water (H2O), has a molecular weight of 18 since hydrogen has a mass of 1.01 and oxygen 16.00, and its mass spectrum has a peak at 18 units. Heuristic Dendral would use this input mass and the knowledge of atomic mass numbers and valence rules, to determine the possible combinations of atomic constituents whose mass would add up to 18. As the weight increases and the molecules become more complex, the number of possible compounds increases drastically. Thus, a program that is able to reduce this number of candidate solutions through the process of hypothesis formation is essential. New graph-theoretic algorithms were invented by Lederberg, Harold Brown, and others that generate all graphs with a specified set of nodes and connection-types (chemical atoms and bonds) -- with or without cycles. Moreover, the team was able to prove mathematically that the generator is complete, in that it produces all graphs with the specified nodes and edges, and that it is non-redundant, in that the output contains no equivalent graphs (e.g., mirror images). The CONGEN program, as it became known, was developed largely by computational chemists Ray Carhart, Jim Nourse, and Dennis Smith. It was useful to chemists as a stand-alone program to generate chemical graphs showing a complete list of structures that satisfy the constraints specified by a user. == Meta-Dendral == Meta-Dendral is a machine learning system that receives the set of possible chemical structures and corresponding mass spectra as input, and proposes a set of rules of mass spectrometry that correlate structural features with processes that produce the mass spectrum. These rules would be fed back to Heuristic Dendral (in the planning and testing programs described below) to test their applicability. Thus, "Heuristic Dendral is a performance system and Meta-Dendral is a learning system". The program is based on two important features: the plan-generate-test paradigm and knowledge engineering. === Plan-generate-test paradigm === The plan-generate-test paradigm is the basic organization of the problem-solving method, and is a common paradigm used by both Heuristic Dendral and Meta-Dendral systems. The generator (later named CONGEN) generates potential solutions for a particular problem, which are then expressed as chemical graphs in Dendral. However, this is feasible only when the number of candidate solutions is minimal. When there are large numbers of possible solutions, Dendral has to find a way to put constraints that rules out large sets of candidate solutions. This is the primary aim of Dendral planner, which is a “hypothesis-formation” program that employs “task-specific knowledge to find constraints for the generator”. Last but not least, the tester analyzes each proposed candidate solution and discards those that fail to fulfill certain criteria. This mechanism of plan-generate-test paradigm is what holds Dendral together. === Knowledge Engineering === The primary aim of knowledge engineering is to attain a productive interaction between the available knowledge base and problem solving techniques. This is possible through development of a procedure in which large amounts of task-specific information is encoded into heuristic programs. Thus, the first essential component of knowledge engineering is a large “knowledge base.” Dendral has specific knowledge about the mass spectrometry technique, a large amount of information that forms the basis of chemistry and graph theory, and information that might be helpful in finding the solution of a particular chemical structure elucidation problem. This “knowledge base” is used both to search for possible chemical structures that match the input data, and to learn new “general rules” that help prune searches. The benefit Dendral provides the end user, even a non-expert, is a minimized set of possible solutions to check manually. == Heuristics == A heuristic is a rule of thumb, an algorithm that does not guarantee a solution, but reduces the number of possible solutions by discarding unlikely and irrelevant solutions. The use of heuristics to solve problems is called "heuristics programming", and was used in Dendral to allow it to replicate in machines the process through which human experts induce the solution to problems via rules of thumb and specific information. Heuristics programming was a major approach and a giant step forward in artificial intelligence, as it allowed scientists to finally automate certain traits of human intelligence. It became prominent among scientists in the late 1940s through George Polya’s book, How to Solve It: A New Aspect of Mathematical Method. As Herbert A. Simon said in The Sciences of the Artificial, "if you take a heuristic conclusion as certain, you may be fooled and disappointed; but if you neglect heuristic conclusions altogether you will make no progress at all." == History == During the mid 20th century, the question "can machines think?" became intriguing and popular among scientists, primarily to add humanistic characteristics to machine behavior. John McCarthy, who was one of the prime researchers of this field, termed this concept of machine intelligence as "artificial intelligence" (AI) during the Dartmouth summer in 1956. AI is usually defined as the capacity of a machine to perform operations that are analogous to human cognitive capabilities. Much research to create AI was done during the 20th century. Also around the mid 20th century, science, especially biology, faced a fast-increasing need to develop a "man-computer symbiosis", to aid scientists in solving problems. For example, the structural analysis of myoglobin, hemoglobin, and other proteins relentlessly needed instrumentation development due to its complexity. In the early 1960s, Joshua Lederberg started working with computers and quickly became tremendously interested in creating interactive computers to help him in his exobiology research. Specifically, he was interested in designing computing systems to help him study alien organic compounds. Lederberg had been heading a team designing instruments for the Mars Viking lander to search for precursor molecules of life in samples of the Mars surface, using a mass spectrometer coupled with a minicomputer. As he was not an expert in either chemistry or computer programming, he collaborated with Stanford chemist Carl Djerassi to help him with chemistry, and Edward Feigenbaum with programming, to automate the process of determining chemical structures from raw mass spectrometry data. Feigenbaum was an expert in programming languages and heuristics, and helped Lederberg design a system that replicated the way Djerassi solved structure elucidation problems. They devised a system called Dendritic Algorithm (Dendral) that was able to generate possible chemical structures corresponding to the mass spectrometry data as an output. Dendral then was still very inaccurate in assessing spectra of ketones, alcohols, and isomers of chemical compounds. Thus, Djerassi "taught" general rules to Dendral that could help eliminate most of the "chemically implausible" structures, and p

  Read more →
• 

  Connectionism

  Connectionism is an approach to the study of human mental processes and cognition that utilizes mathematical models known as connectionist networks or artificial neural networks. Connectionism has had many "waves" since its beginnings. The first wave appeared 1943 with Warren Sturgis McCulloch and Walter Pitts both focusing on comprehending neural circuitry through a formal and mathematical approach, and Frank Rosenblatt who published the 1958 paper "The Perceptron: A Probabilistic Model For Information Storage and Organization in the Brain" in Psychological Review, while working at the Cornell Aeronautical Laboratory. The first wave ended with the 1969 book Perceptrons about limitations of the original perceptron idea, written by Marvin Minsky and Seymour Papert, which contributed to discouraging major funding agencies in the US from investing in connectionist research. With a few noteworthy deviations, most connectionist research entered a period of inactivity until the mid-1980s. The term connectionist model was reintroduced in a 1982 paper in the journal Cognitive Science by Jerome Feldman and Dana Ballard. The second wave blossomed in the late 1980s, following a 1987 book Parallel Distributed Processing by James L. McClelland, David E. Rumelhart, et al., which introduced a couple of improvements to the simple perceptron idea, such as intermediate processors (now known as "hidden layers") alongside input and output units, and used a sigmoid activation function instead of the old "all-or-nothing" function. Their work built upon that of John Hopfield, who was a key figure investigating the mathematical characteristics of sigmoid activation functions. From the late 1980s to the mid-1990s, connectionism took on an almost revolutionary tone when Schneider, Terence Horgan and Tienson posed the question of whether connectionism represented a fundamental shift in psychology and so-called "good old-fashioned AI", or GOFAI. Some advantages of the second wave connectionist approach included its applicability to a broad array of functions, structural approximation to biological neurons, low requirements for innate structure, and capacity for graceful degradation. Its disadvantages included the difficulty in deciphering how ANNs process information or account for the compositionality of mental representations, and a resultant difficulty explaining phenomena at a higher level. The current (third) wave has been marked by advances in deep learning, which have made possible the creation of large language models. The success of deep-learning networks in the past decade has greatly increased the popularity of this approach, but the complexity and scale of such networks has brought with them increased interpretability problems. == Basic principle == The central connectionist principle is that mental phenomena can be described by interconnected networks of simple and often uniform units. The form of the connections and the units can vary from model to model. For example, units in the network could represent neurons and the connections could represent synapses, as in the human brain. This principle has been seen as an alternative to GOFAI and the classical theories of mind based on symbolic computation, but the extent to which the two approaches are compatible has been the subject of much debate since their inception. === Activation function === Internal states of any network change over time due to neurons sending a signal to a succeeding layer of neurons in the case of a feedforward network, or to a previous layer in the case of a recurrent network. Discovery of non-linear activation functions has enabled the second wave of connectionism. === Memory and learning === Neural networks follow two basic principles: Any mental state can be described as a n-dimensional vector of numeric activation values over neural units in a network. Memory and learning are created by modifying the 'weights' of the connections between neural units, generally represented as an n×m matrix. The weights are adjusted according to some learning rule or algorithm, such as Hebbian learning. Most of the variety among the models comes from: Interpretation of units: Units can be interpreted as neurons or groups of neurons. Definition of activation: Activation can be defined in a variety of ways. For example, in a Boltzmann machine, the activation is interpreted as the probability of generating an action potential spike, and is determined via a logistic function on the sum of the inputs to a unit. Learning algorithm: Different networks modify their connections differently. In general, any mathematically defined change in connection weights over time is referred to as the "learning algorithm". === Biological realism === Connectionist work in general does not need to be biologically realistic. One area where connectionist models are thought to be biologically implausible is with respect to error-propagation networks that are needed to support learning, but error propagation can explain some of the biologically-generated electrical activity seen at the scalp in event-related potentials such as the N400 and P600, and this provides some biological support for one of the key assumptions of connectionist learning procedures. Many recurrent connectionist models also incorporate dynamical systems theory. Many researchers, such as the connectionist Paul Smolensky, have argued that connectionist models will evolve toward fully continuous, high-dimensional, non-linear, dynamic systems approaches. == Precursors == Precursors of the connectionist principles can be traced to early work in psychology, such as that of William James. Psychological theories based on knowledge about the human brain were fashionable in the late 19th century. As early as 1869, the neurologist John Hughlings Jackson argued for multi-level, distributed systems. Following from this lead, Herbert Spencer's Principles of Psychology, 3rd edition (1872), and Sigmund Freud's Project for a Scientific Psychology (composed 1895) propounded connectionist or proto-connectionist theories. These tended to be speculative theories. But by the early 20th century, Edward Thorndike was writing about human learning that posited a connectionist type network. Hopfield networks had precursors in the Ising model due to Wilhelm Lenz (1920) and Ernst Ising (1925), though the Ising model conceived by them did not involve time. Monte Carlo simulations of Ising model required the advent of computers in the 1950s. == The first wave == The first wave begun in 1943 with Warren Sturgis McCulloch and Walter Pitts both focusing on comprehending neural circuitry through a formal and mathematical approach. McCulloch and Pitts showed how neural systems could implement first-order logic: Their classic paper "A Logical Calculus of Ideas Immanent in Nervous Activity" (1943) is important in this development here. They were influenced by the work of Nicolas Rashevsky in the 1930s and symbolic logic in the style of Principia Mathematica. Hebb contributed greatly to speculations about neural functioning, and proposed a learning principle, Hebbian learning. Lashley argued for distributed representations as a result of his failure to find anything like a localized engram in years of lesion experiments. Friedrich Hayek independently conceived the model, first in a brief unpublished manuscript in 1920, then expanded into a book in 1952. The Perceptron machines were proposed and built by Frank Rosenblatt, who published the 1958 paper “The Perceptron: A Probabilistic Model For Information Storage and Organization in the Brain” in Psychological Review, while working at the Cornell Aeronautical Laboratory. He cited Hebb, Hayek, Uttley, and Ashby as main influences. Another form of connectionist model was the relational network framework developed by the linguist Sydney Lamb in the 1960s. The research group led by Widrow empirically searched for methods to train two-layered ADALINE networks (MADALINE), with limited success. A method to train multilayered perceptrons with arbitrary levels of trainable weights was published by Alexey Grigorevich Ivakhnenko and Valentin Lapa in 1965, called the Group Method of Data Handling. This method employs incremental layer by layer training based on regression analysis, where useless units in hidden layers are pruned with the help of a validation set. The first multilayered perceptrons trained by stochastic gradient descent was published in 1967 by Shun'ichi Amari. In computer experiments conducted by Amari's student Saito, a five layer MLP with two modifiable layers learned useful internal representations to classify non-linearily separable pattern classes. In 1972, Shun'ichi Amari produced an early example of self-organizing network. == The neural network winter == There was some conflict among artificial intelligence researchers as to what neural networks are useful for. Around late 1960s, there was a widespread lull in research a

  Read more →
• 

  Daisy Intelligence

  Daisy Intelligence is a Canadian artificial intelligence (AI) company that provides data analysis services to help retailers, mainly grocers and supermarkets, to determine optimal pricing and promotional mix. The company also helps insurance companies detect fraudulent claims. The company uses a subset of AI known as reinforcement learning. In October 2019, the company moved from the suburban Vaughan, Ontario, to downtown Toronto, joining other AI and technology startups concentrated in the King Street East area. In 2019, the company was ranked No. 39 on The Globe and Mail's annual list of Canada's "top growing companies by three-year revenue growth."

  Read more →
• 

  WHATWG

  The Web Hypertext Application Technology Working Group (WHATWG) was founded by representatives from Apple Inc., the Mozilla Foundation and Opera Software, leading web browser vendors in 2004. WHATWG is responsible for maintaining multiple web-related technical standards, including the specifications for the HyperText Markup Language (HTML) and the Document Object Model (DOM). The central organizational membership and control of WHATWG – its "Steering Group" – consists of Apple, Mozilla, Google, and Microsoft. WHATWG editors of the specifications ensure correct implementation, in consultation with participants, but ultimately in accordance with Steering Group member objectives. == History == The WHATWG was formed in response to the slow development of World Wide Web Consortium (W3C) Web standards and W3C's decision to abandon HTML in favor of XML-based technologies. The WHATWG mailing list was announced on 4 June 2004, two days after the initiatives of a joint Opera–Mozilla position paper had been voted down by the W3C members at the W3C Workshop on Web Applications and Compound Documents. On 10 April 2007, the Mozilla Foundation, Apple, and Opera Software proposed that the new HTML working group of the W3C adopt the WHATWG's HTML5 as the starting point of its work and name its future deliverable as "HTML5" (though the WHATWG specification was later renamed HTML Living Standard). On 9 May 2007, the new HTML working group of the W3C resolved to do that. An Internet Explorer platform architect from Microsoft was invited but did not join, citing the lack of a patent policy to ensure all specifications can be implemented on a royalty-free basis. Since then, the W3C and the WHATWG had been developing HTML independently, at times causing specifications to diverge. In 2017, the WHATWG established an intellectual property rights agreement that includes a patent policy. This spurred a renewed attempt to allow the W3C and the WHATWG to work together on specifications. In 2019, the W3C and WHATWG agreed to a memorandum of understanding where development of HTML and DOM specifications would be done principally in the WHATWG. The editor has significant control over the specification, but the community can influence the decisions of the editor. In one case, editor Ian Hickson proposed replacing the tag with a more generic tag, but the community disagreed and the change was reverted. === Web Hypertext Application Technology Task Force === Initially, the name Web Hypertext Application Technology Task Force was also used, along with variant abbreviations including WHAT Working Group, WHAT Task Force and WHATTF. After some time using both the whattf.org and whatwg.org domain names, the name WHATWG was eventually standardized on. The namespace URI http://whattf.org/datatype-draft remains in use for the HTML validator's data type library. === Transition of HTML Publication to WHATWG === On 28 May 2019, the W3C announced that WHATWG would be the sole publisher of the HTML and DOM standards. The W3C and WHATWG had been publishing competing standards since 2012. While the W3C standard was identical to the WHATWG in 2007 the standards have since progressively diverged due to different design decisions. The WHATWG "Living Standard" had been the de facto web standard for some time. == Specifications == The WHATWG publishes a number of standards that form a substantial portion of the web platform including: The HTML Living Standard (sometimes informally called HTML5). The HTML specification has been a living document without version numbers since 2011. It includes both HTML, the core markup language for the web, and a number of related APIs. The DOM Standard, defines how the Document Object Model on the web is supposed to work and replaces W3C DOM level 3. For example, it replaces mutation events with mutation observers. Fetch Standard, which "defines requests, responses, and the process that binds them: fetching." The fetch standard defines the 'fetch' JavaScript API, and supersedes the HTML5 fetch functionality, CORS and the HTTP Origin header semantics. The Streams Standard provides APIs for creating, composing, and consuming streams of data. These streams are designed to map efficiently to low-level I/O primitives, and allow easy composition with built-in backpressure and queueing. On top of streams, the web platform can build higher-level abstractions, such as filesystem or socket APIs, while at the same time users can use the supplied tools to build their own streams which integrate well with those of the web platform. The Encoding Standard defines how character encodings such as Windows-1252 and UTF-8 are handled in web browsers and is intended to replace the IETF encodings registry. The MIME type sniffing standard defines how MIME types are supposed to be sniffed in web browsers. The URL standard defines how URLs are supposed to be parsed in web browsers. Web IDL used to describe interfaces that are intended to be implemented in web browsers.

  Read more →
• 

  Cellular neural network

  In computer science and machine learning, Cellular Neural Networks (CNN) or Cellular Nonlinear Networks (CNN) are a parallel computing paradigm similar to neural networks, with the difference that communication is allowed between neighbouring units only. Typical applications include image processing, analyzing 3D surfaces, solving partial differential equations, reducing non-visual problems to geometric maps, modelling biological vision and other sensory-motor organs. CNN is not to be confused with convolutional neural networks (also colloquially called CNN). == CNN architecture == Due to their number and variety of architectures, it is difficult to give a precise definition for a CNN processor. From an architecture standpoint, CNN processors are a system of finite, fixed-number, fixed-location, fixed-topology, locally interconnected, multiple-input, single-output, nonlinear processing units. The nonlinear processing units are often referred to as neurons or cells. Mathematically, each cell can be modeled as a dissipative, nonlinear dynamical system where information is encoded via its initial state, inputs and variables used to define its behavior. Dynamics are usually continuous, as in the case of Continuous-Time CNN (CT-CNN) processors, but can be discrete, as in the case of Discrete-Time CNN (DT-CNN) processors. Each cell has one output, by which it communicates its state with both other cells and external devices. Output is typically real-valued, but can be complex or even quaternion, i.e. a Multi-Valued CNN (MV-CNN). Most CNN processors, processing units are identical, but there are applications that require non-identical units, which are called Non-Uniform Processor CNN (NUP-CNN) processors, and consist of different types of cells. === Chua-Yang CNN === In the original Chua-Yang CNN (CY-CNN) processor, the state of the cell was a weighted sum of the inputs and the output was a piecewise linear function. However, like the original perceptron-based neural networks, the functions it could perform were limited: specifically, it was incapable of modeling non-linear functions, such as XOR. More complex functions are realizable via Non-Linear CNN (NL-CNN) processors. Cells are defined in a normed gridded space like two-dimensional Euclidean geometry. However, the cells are not limited to two-dimensional spaces; they can be defined in an arbitrary number of dimensions and can be square, triangle, hexagonal, or any other spatially invariant arrangement. Topologically, cells can be arranged on an infinite plane or on a toroidal space. Cell interconnect is local, meaning that all connections between cells are within a specified radius (with distance measured topologically). Connections can also be time-delayed to allow for processing in the temporal domain. Most CNN architectures have cells with the same relative interconnects, but there are applications that require a spatially variant topology, i.e. Multiple-Neighborhood-Size CNN (MNS-CNN) processors. Also, Multiple-Layer CNN (ML-CNN) processors, where all cells on the same layer are identical, can be used to extend the capability of CNN processors. The definition of a system is a collection of independent, interacting entities forming an integrated whole, whose behavior is distinct and qualitatively greater than its entities. Although connections are local, information exchange can happen globally through diffusion. In this sense, CNN processors are systems because their dynamics are derived from the interaction between the processing units and not within processing units. As a result, they exhibit emergent and collective behavior. Mathematically, the relationship between a cell and its neighbors, located within an area of influence, can be defined by a coupling law, and this is what primarily determines the behavior of the processor. When the coupling laws are modeled by fuzzy logic, it is a fuzzy CNN. When these laws are modeled by computational verb logic, it becomes a computational verb CNN. Both fuzzy and verb CNNs are useful for modelling social networks when the local couplings are achieved by linguistic terms. == History == The idea of CNN processors was introduced by Leon Chua and Lin Yang in 1988. In these articles, Chua and Yang outline the underlying mathematics behind CNN processors. They use this mathematical model to demonstrate, for a specific CNN implementation, that if the inputs are static, the processing units will converge, and can be used to perform useful calculations. They then suggest one of the first applications of CNN processors: image processing and pattern recognition (which is still the largest application to date). Leon Chua is still active in CNN research and publishes many of his articles in the International Journal of Bifurcation and Chaos, of which he is an editor. Both IEEE Transactions on Circuits and Systems and the International Journal of Bifurcation also contain a variety of useful articles on CNN processors authored by other knowledgeable researchers. The former tends to focus on new CNN architectures and the latter more on the dynamical aspects of CNN processors. In 1993, Tamas Roska and Leon Chua introduced the first algorithmically programmable analog CNN processor in the world. The multi-national effort was funded by the Office of Naval Research, the National Science Foundation, and the Hungarian Academy of Sciences, and researched by the Hungarian Academy of Sciences and the University of California. This article proved that CNN processors were producible and provided researchers a physical platform to test their CNN theories. After this article, companies started to invest into larger, more capable processors, based on the same basic architecture as the CNN Universal Processor. Tamas Roska is another key contributor to CNNs. His name is often associated with biologically inspired information processing platforms and algorithms, and he has published numerous key articles and has been involved with companies and research institutions developing CNN technology. === Literature === Two references are considered invaluable since they manage to organize the vast amount of CNN literature into a coherent framework: An overview by Valerio Cimagalli and Marco Balsi. The paper provides a concise intro to definitions, CNN types, dynamics, implementations, and applications. "Cellular Neural Networks and Visual Computing Foundations and Applications", written by Leon Chua and Tamas Roska, which provides examples and exercises. The book covers many different aspects of CNN processors and can serve as a textbook for a Masters or Ph.D. course. Other resources include The proceedings of "The International Workshop on Cellular Neural Networks and Their Applications" provide much CNN literature. The proceedings are available online, via IEEE Xplore, for conferences held in 1990, 1992, 1994, 1996, 1998, 2000, 2002, 2005 and 2006. There was also a workshop held in Santiago de Composetela, Spain. Topics included theory, design, applications, algorithms, physical implementations and programming and training methods. For an understanding of the analog semiconductor based CNN technology, AnaLogic Computers has their product line, in addition to the published articles available on their homepage and their publication list. They also have information on other CNN technologies such as optical computing. Many of the commonly used functions have already been implemented using CNN processors. A good reference point for some of these can be found in image processing libraries for CNN based visual computers such as Analogic’s CNN-based systems. == Related processing architectures == CNN processors could be thought of as a hybrid between artificial neural network (ANN) and Continuous Automata (CA). === Artificial Neural Networks === The processing units of CNN and NN are similar. In both cases, the processor units are multi-input, dynamical systems, and the behavior of the overall systems is driven primarily through the weights of the processing unit’s linear interconnect. However, in CNN processors, connections are made locally, whereas in ANN, connections are global. For example, neurons in one layer are fully connected to another layer in a feed-forward NN and all the neurons are fully interconnected in Hopfield networks. In ANNs, the weights of interconnections contain information on the processing system’s previous state or feedback. But in CNN processors, the weights are used to determine the dynamics of the system. Furthermore, due to the high inter-connectivity of ANNs, they tend not exploit locality in either the data set or the processing and as a result, they usually are highly redundant systems that allow for robust, fault-tolerant behavior without catastrophic errors. A cross between an ANN and a CNN processor is a Ratio Memory CNN (RMCNN). In RMCNN processors, the cell interconnect is local and topologically invariant, but the weights are used to store

  Read more →
• 

  MuZero

  MuZero is a computer program developed by artificial intelligence research company DeepMind, a subsidiary of Google, to master games without knowing their rules and underlying dynamics. Its release in 2019 included benchmarks of its performance in Go, chess, shogi, and a suite of 57 different Atari games. The algorithm uses an approach similar to AlphaZero, where a combination of a tree-based search and a learned model is deployed. It matched AlphaZero's performance in chess and shogi, improved on its performance in Go, and improved on the state of the art in mastering a suite of 57 Atari games (the Arcade Learning Environment), a visually-complex domain. MuZero was trained via self-play, with no access to rules, opening books, or endgame tablebases. The trained algorithm used the same convolutional and residual architecture as AlphaZero, but with 20 percent fewer computation steps per node in the search tree. == History == MuZero really is discovering for itself how to build a model and understand it just from first principles. On November 19, 2019, the DeepMind team released a preprint introducing MuZero. === Derivation from AlphaZero === MuZero (MZ) is a combination of the high-performance planning of the AlphaZero (AZ) algorithm with approaches to model-free reinforcement learning. The combination allows for more efficient training in classical planning regimes, such as Go, while also handling domains with much more complex inputs at each stage, such as visual video games. MuZero was derived directly from AZ code, sharing its rules for setting hyperparameters. Differences between the approaches include: AZ's planning process uses a simulator. The simulator knows the rules of the game. It has to be explicitly programmed. A neural network then predicts the policy and value of a future position. Perfect knowledge of game rules is used in modeling state transitions in the search tree, actions available at each node, and termination of a branch of the tree. MZ does not have access to the rules, and instead learns one with neural networks. AZ has a single model for the game (from board state to predictions); MZ has separate models for representation of the current state (from board state into its internal embedding), dynamics of states (how actions change representations of board states), and prediction of policy and value of a future position (given a state's representation). MZ's hidden model may be complex, and it may turn out it can host computation; exploring the details of the hidden model in a trained instance of MZ is a topic for future exploration. MZ does not expect a two-player game where winners take all. It works with standard reinforcement-learning scenarios, including single-agent environments with continuous intermediate rewards, possibly of arbitrary magnitude and with time discounting. AZ was designed for two-player games that could be won, drawn, or lost. === Comparison with R2D2 === The previous state of the art technique for learning to play the suite of Atari games was R2D2, the Recurrent Replay Distributed DQN. MuZero surpassed both R2D2's mean and median performance across the suite of games, though it did not do better in every game. == Training and results == MuZero used 16 third-generation tensor processing units (TPUs) for training, and 1000 TPUs for selfplay for board games, with 800 simulations per step and 8 TPUs for training and 32 TPUs for selfplay for Atari games, with 50 simulations per step. AlphaZero used 64 second-generation TPUs for training, and 5000 first-generation TPUs for selfplay. As TPU design has improved (third-generation chips are 2x as powerful individually as second-generation chips, with further advances in bandwidth and networking across chips in a pod), these are comparable training setups. R2D2 was trained for 5 days through 2M training steps. === Initial results === MuZero matched AlphaZero's performance in chess and shogi after roughly 1 million training steps. It matched AZ's performance in Go after 500,000 training steps and surpassed it by 1 million steps. It matched R2D2's mean and median performance across the Atari game suite after 500 thousand training steps and surpassed it by 1 million steps, though it never performed well on 6 games in the suite. == Reactions and related work == MuZero was viewed as a significant advancement over AlphaZero, and a generalizable step forward in unsupervised learning techniques. The work was seen as advancing understanding of how to compose systems from smaller components, a systems-level development more than a pure machine-learning development. While only pseudocode was released by the development team, Werner Duvaud produced an open source implementation based on that. MuZero has been used as a reference implementation in other work, for instance as a way to generate model-based behavior. In late 2021, a more efficient variant of MuZero was proposed, named EfficientZero. It "achieves 194.3 percent mean human performance and 109.0 percent median performance on the Atari 100k benchmark with only two hours of real-time game experience". In early 2022, a variant of MuZero was proposed to play stochastic games (for example 2048, backgammon), called Stochastic MuZero, which uses afterstate dynamics and chance codes to account for the stochastic nature of the environment when training the dynamics network.

  Read more →