Spike is a cross-platform email client and AI-powered communication app, available on Windows, MacOS, iOS, Android and the web. It has a chat-like, conversational view for emails with AI-powered inbox management and integrated collaboration features. Depending on the selected plan, it can be used solely as an email application or as a full suite of business communication tools. == History == Founded in 2013 by Erez Pilosof and Dvir Ben-Aroya, Spike is a software application that puts existing e-mails into a multimedia messaging, chat-like interface enhanced with video and voice calls. The application was initially named Hop. In 2019, the developers completed a $5 million funding round including investment from Wix.com and NFX Capital. In 2020, Spike raised $8m in a Series A funding round led by Insight Partners with the participation from previous rounds' investors. In 2021 Spike announced a collaboration with Meta to launch on the Oculus Store and would become one of the first productivity apps to launch in Meta's new virtual world, known as the Metaverse. In June 2023, the company introduced its corporate offering — Teamspace, a corporate communication platform for teams with features such as company-wide channels for broad conversations, private groups for specific topics or projects, direct one-on-one conversations, video meetings, file collaboration, AI-powered email messaging, and custom email domain. It supports file management, search capabilities, and project management. Built on open-protocol technology, Spike Teamspace enables users to send and receive messages from all email providers. Regardless of whether the other party is using Spike. == Company operations == Spike is developed and operated by SpikeNow LTD. Dvir Ben Aroya serves as Spike’s CEO and Erez Pilosof is the CTO. The company is headquartered in Tel Aviv, Israel. == Mode of use == The app enables users to organize email into three types of "conversations,"a traditional inbox/sent format, by subject, or by people. Spike users can also make audio and video calls to each other, and other features include a calendar, contact list, and Groups. Spike is available for Microsoft Windows, MacOS, iOS and Android, and as a web version, and works with Gmail, Outlook, Exchange, iCloud, Yahoo! Mail and IMAP email providers. == Features == Since 2023, the platform features an AI-driven assistant, Magic AI, for customized email creation, document summarization, research, content generation, advanced note-taking, project management, and real-time translation. Since 2023, Spike offers custom email domain management. It supports team collaboration through Channels, uniting members globally with access to historical messages, and combines email with real-time messaging via Conversational Email. The Shared Inbox allows team collaboration on emails, while Groups support private conversations and invitations. It also features integrated video meetings, real-time collaboration on documents and notes, and email hosting with custom domains. Super Search enables retrieval of various content, and the Priority Inbox organizes emails by priority. Collaborative Tasks offer real-time updates and tracking. The platform allows voice message sending from mobile devices and integrates multiple calendar platforms into a unified schedule. File Management optimizes attachment handling, and the Unified Inbox consolidates emails from multiple accounts. Spike ensures data security with AES-256 encryption and private keys. The platform features AI-powered inbox management and communication tools. In May 2025, Spike launched its AI Feed feature, which automatically summarizes unread messages in a unified stream and enables bulk email actions. Additional AI capabilities include email composition assistance, document summarization, content generation, note-taking enhancement, and real-time translation.
Visual temporal attention
Visual temporal attention is a special case of visual attention that involves directing attention to specific instant of time. Similar to its spatial counterpart visual spatial attention, these attention modules have been widely implemented in video analytics in computer vision to provide enhanced performance and human interpretable explanation of deep learning models. As visual spatial attention mechanism allows human and/or computer vision systems to focus more on semantically more substantial regions in space, visual temporal attention modules enable machine learning algorithms to emphasize more on critical video frames in video analytics tasks, such as human action recognition. In convolutional neural network-based systems, the prioritization introduced by the attention mechanism is regularly implemented as a linear weighting layer with parameters determined by labeled training data. == Application in Action Recognition == Recent video segmentation algorithms often exploits both spatial and temporal attention mechanisms. Research in human action recognition has accelerated significantly since the introduction of powerful tools such as Convolutional Neural Networks (CNNs). However, effective methods for incorporation of temporal information into CNNs are still being actively explored. Motivated by the popular recurrent attention models in natural language processing, the Attention-aware Temporal Weighted CNN (ATW CNN) is proposed in videos, which embeds a visual attention model into a temporal weighted multi-stream CNN. This attention model is implemented as temporal weighting and it effectively boosts the recognition performance of video representations. Besides, each stream in the proposed ATW CNN framework is capable of end-to-end training, with both network parameters and temporal weights optimized by stochastic gradient descent (SGD) with back-propagation. Experimental results show that the ATW CNN attention mechanism contributes substantially to the performance gains with the more discriminative snippets by focusing on more relevant video segments. == Literature == Seibold VC, Balke J and Rolke B (2023): Temporal attention. Front. Cognit. 2:1168320. doi: 10.3389/fcogn.2023.1168320.
20Q
20Q is a computerized game of twenty questions that began as a test in artificial intelligence (AI). It was invented by Robin Burgener in 1988. The game was made handheld by Radica in 2003, but was discontinued in 2011 because Techno Source took the license for 20Q handheld devices. The game 20Q is based on the spoken parlor game known as twenty questions, and is both a website and a handheld device. 20Q asks the player to think of something and will then try to guess what they are thinking of with twenty yes-or-no questions. If it fails to guess in 20 questions, it will ask an additional 5 questions. If it fails to guess even with 25 (or 30) questions, the player is declared the winner. Sometimes the first guess of the object can be asked at question 14. == Principle and history == The principle is that the player thinks of something and the 20Q artificial intelligence asks a series of questions before guessing what the player is thinking. This artificial intelligence learns on its own with the information relayed back to the players who interact with it, and is not programmed. The player can answer these questions with: Yes, No, Unknown, and Sometimes. The experiment is based on the classic word game of Twenty Questions, and on the computer game "Animals," popular in the early 1970s, which used a somewhat simpler method to guess an animal. The 20Q AI uses an artificial neural network to pick the questions and to guess. After the player has answered the twenty questions posed (sometimes fewer), 20Q makes a guess. If it is incorrect, it asks more questions, then guesses again. It makes guesses based on what it has learned; it is not programmed with information or what the inventor thinks. Answers to any question are based on players’ interpretations of the questions asked. Newer editions were made for different categories, such as music 20Q which has the player think of a song, and Harry Potter 20Q, which has the player think of something from the world of the Harry Potter series. The 20Q AI can draw its own conclusions on how to interpret the information. It can be described as more of a folk taxonomy than a taxonomy. Its knowledge develops with every game played. In this regard, the online version of the 20Q AI can be inaccurate because it gathers its answers from what people think rather than from what people know. Limitations of taxonomy are often overcome by the AI itself because it can learn and adapt. For example, if the player was thinking of a "Horse" and answered "No" to the question "Is it an animal?," the AI will, nevertheless, guess correctly, despite being told that a horse is not an animal. Patent applications in the US and Europe were submitted in 2005. In August 2014, 20Q.net Inc., with Brashworks Studios, developed and released an iOS iPad version available at the Apple iTunes store. == Game show == On June 13, 2009, GSN began a TV version of the game, hosted by Cat Deeley, with Hal Sparks as the voice of Mr. Q.
ComfyUI
ComfyUI is an open source, node-based program that allows users to generate images from a series of text prompts. It uses free diffusion models such as Stable Diffusion as the base model for its image capabilities combined with other tools such as ControlNet and LCM Low-rank adaptation with each tool being represented by a node in the program. == History == ComfyUI was released on GitHub in January 2023. According to comfyanonymous, the creator, a major goal of the project was to improve on existing software designs in terms of the user interface. The creator had been involved with Stability AI but by 3 June 2024 that involvement had ended and an organization called Comfy Org had been created along with the core developers. In July 2024, Nvidia announced support for ComfyUI within its RTX Remix modding software. In August 2024, support was added for the Flux diffusion model developed by Black Forest Labs, and Comfy Org joined the Open Model Initiative created by the Linux Foundation. As of Sept 2025, the project has 89.2k stars on GitHub. ComfyUI is one of the most popular user interfaces for Stable Diffusion, along with Automatic1111. == Features == ComfyUI's main feature is that it is node based. Each node has a function such as "load a model" or "write a prompt". The nodes are connected to form a control-flow graph called a workflow. When a prompt is queued, a highlighted frame appears around the currently executing node, starting from "load checkpoint" and ending with the final image and its save location. Workflows commonly consist of tens of nodes, forming a complex directed acyclic graph. Node types include loading a model, specifying prompts, samplers, schedulers, VAE decoders, face restoration and upscaling models, LoRAs, embeddings, and ControlNets. Several samplers are supported, such as Euler, Euler_a, dpmpp_2m_sde and dpmpp_3m_sde. Workflows can be saved to a file, allowing users to re-use node workflows and share them with other users. The file format for the workflows is in JSON and can be embedded in the generated images. Users have also created custom extensions to the base system which are exposed as new nodes, such as the extension for AnimateDiff, which aims to create videos. ComfyUI has been described as more complex compared to other diffusion UIs such as Automatic1111. A default node group is also included with the program. As of December 2024, 1,674 nodes were supported. ComfyUI Supports multiple text-to-image models including, Stable Diffusion, Flux and Tencent's Hunyuan-DiT, as well as custom models from Civitai like Pony. == LLMVision extension compromise == In June 2024, a hacker group called "Nullbulge" compromised an extension of ComfyUI to add malicious code to it. The compromised extension, called ComfyUI_LLMVISION, was used for integrating the interface with AI language models GPT-4 and Claude 3, and was hosted on GitHub. Nullbulge hosted a list of hundreds of ComfyUI users' login details across multiple services on its website, while users of the extension reported receiving numerous login notifications. vpnMentor conducted security research on the extension and claimed it could "steal crypto wallets, screenshot the user’s screen, expose device information and IP addresses, and steal files that contain certain keywords or extensions". Nullbulge's website claims they targeted users who committed "one of our sins", which included AI-art generation, art theft, promoting cryptocurrency, and any other kind of theft from artists such as from Patreon. They claimed that they were "a collective of individuals who believe in the importance of protecting artists' rights and ensuring fair compensation for their work" and that they believed that "AI-generated artwork is detrimental to the creative industry and should be discouraged".
Hallin's spheres
Hallin's spheres is a theory of news reporting and its rhetorical framing posited by journalism historian Daniel C. Hallin in his 1986 book The Uncensored War to explain the news coverage of the Vietnam War. Hallin divides the world of political discourse into three concentric spheres: consensus, legitimate controversy, and deviance. In the sphere of consensus, journalists assume everyone agrees. The sphere of legitimate controversy includes the standard political debates, and journalists are expected to remain neutral. The sphere of deviance falls outside the bounds of legitimate debate, and journalists can ignore it. These boundaries shift, as public opinion shifts. Hallin's spheres, which deals with the media, are similar to the Overton window, which deals with public opinion generally, and posits a sliding scale of public opinion on any given issue ranging from conventional wisdom to unacceptable. Hallin used the concept of framing to describe the presentation and reception of issues in public. For example, framing the use of drugs as criminal activity can encourage the public to consider that behavior anti-social. Hallin's work was later referred to in the controversial formulation of the concept of an opinion corridor, in which the range of acceptable public opinion narrows, and opinion outside that corridor moves from legitimate controversy into deviance. == Description == === Sphere of consensus === This sphere contains those topics on which there is widespread agreement, or at least the perception thereof. Within the sphere of consensus, "journalists feel free to invoke a generalized 'we' and to take for granted shared values and shared assumptions". Examples include such things as motherhood and apple pie. For topics in this sphere, journalists feel free to be advocating cheerleaders without having to be neutral or present any opposing view point and be disinterested observers." === Sphere of legitimate controversy === For topics in this sphere rational and informed people hold differing views within limited range. These topics are therefore the most important to cover, and also ones upon which journalists are seemingly obliged to remain disinterested reporters, rather than advocating for or against a particular view. Schudson notes that Hallin, in his influential study of the US media during the Vietnam War, argues that journalism's commitment to objectivity has always been compartmentalized. That is, within a certain sphere—the sphere of legitimate controversy—journalists seek conscientiously to be balanced and objective. The work of Walter Williams professor at the University of Missouri, Rod Petersen, advanced the idea that priming—controlling the narratives that media covers—can be the tool that media use to get deviant news subjects into the legitimate controversial circles of new coverage. === Sphere of deviance === Topics in this sphere are rejected by journalists as being unworthy of general consideration. Such views are perceived as being out of hand, unfounded, taboo, or of such minor consequence that they are not newsworthy. Hallin argues that in the sphere of deviance, "journalists also depart from standard norms of objective reporting and feel authorized to treat as marginal, laughable, dangerous". They either avoid mentioning or ridicule the controversial subject as outside the bounds of acceptable controversy; and they censor the individuals and groups who are associated with it. A simple example: a person claiming that aliens are manipulating college basketball scores might have difficulty finding sports media coverage for such a claim. A more political example: the US media regulator FCC's "Fairness Doctrine" aimed at radio stations, advocated balance between right and left political news and opinions, yet specified that broadcasters did not have to reserve any space or time for Communist viewpoints. == Uses of the terms == Craig Watkins (2001, pp. 92–94) makes use of the Hallin's spheres in a paper examining ABC, CBS, and NBC television network television news coverage of the Million Man March, a demonstration that took place in Washington, D.C., on October 16, 1995. Watkins analyzes the dominant framing practices—problem definition, rhetorical devices, use of sources, and images—employed by journalists to make sense of this particular expression of political protest. He argues that Hallin's three spheres are a way for media framing practices to develop specific reportorial contexts, and each sphere develops its own distinct style of news reporting resources by different rhetorical tropes and discourses. Piers Robinson (2001, p. 536) uses the concept in relation to debates that have emerged over the extent to which the mass media serves elite interests or, alternatively, plays a powerful role in shaping political outcomes. His article reviews Hallin's spheres as an example of media-state relations, that highlights theoretical and empirical shortcomings in the 'manufacturing consent' thesis (Chomsky, McChesney). Robinson argues that a more nuanced and bi-directional understanding is needed of the direction of influence between media and the state that builds upon, rather than rejecting, existing theoretical accounts. Hallin's theory assumed a relatively homogenized media environment, where most producers were trying to reach most consumers. A more fractured media landscape can challenge this assumption because different audiences may place topics in different spheres, a concept related to the filter bubble, which posits that many members of the public choose to limit their media consumption to the areas of consensus and deviance that they personally prefer.
Structural similarity index measure
The structural similarity index measure (SSIM) is a method for predicting the perceived quality of digital television and cinematic pictures, as well as other kinds of digital images and videos. It is also used for measuring the similarity between two images. The SSIM index is a full reference metric; in other words, the measurement or prediction of image quality is based on an initial uncompressed or distortion-free image as reference. SSIM is a perception-based model that considers image degradation as perceived change in structural information, while also incorporating important perceptual phenomena, including both luminance masking and contrast masking terms. This distinguishes from other techniques such as mean squared error (MSE) or peak signal-to-noise ratio (PSNR) that instead estimate absolute errors. Structural information is the idea that the pixels have strong inter-dependencies especially when they are spatially close. These dependencies carry important information about the structure of the objects in the visual scene. Luminance masking is a phenomenon whereby image distortions (in this context) tend to be less visible in bright regions, while contrast masking is a phenomenon whereby distortions become less visible where there is significant activity or "texture" in the image. == History == The predecessor of SSIM was called Universal Quality Index (UQI), or Wang–Bovik index, which was developed by Zhou Wang and Alan Bovik in 2001. This evolved, through their collaboration with Hamid Sheikh and Eero Simoncelli, into the current version of SSIM, which was published in April 2004 in the IEEE Transactions on Image Processing. In addition to defining the SSIM quality index, the paper provides a general context for developing and evaluating perceptual quality measures, including connections to human visual neurobiology and perception, and direct validation of the index against human subject ratings. The basic model was developed in the Laboratory for Image and Video Engineering (LIVE) at The University of Texas at Austin and further developed jointly with the Laboratory for Computational Vision (LCV) at New York University. Further variants of the model have been developed in the Image and Visual Computing Laboratory at University of Waterloo and have been commercially marketed. SSIM subsequently found strong adoption in the image processing community and in the television and social media industries. The 2004 SSIM paper has been cited over 50,000 times according to Google Scholar, making it one of the highest cited papers in the image processing and video engineering fields. It was recognized with the IEEE Signal Processing Society Best Paper Award for 2009. It also received the IEEE Signal Processing Society Sustained Impact Award for 2016, indicative of a paper having an unusually high impact for at least 10 years following its publication. Because of its high adoption by the television industry, the authors of the original SSIM paper were each accorded a Primetime Engineering Emmy Award in 2015 by the Television Academy. == Algorithm == The SSIM index is calculated between two windows of pixel values x {\displaystyle x} and y {\displaystyle y} of common size, from corresponding locations in two images to be compared. These SSIM values can be aggregated across the full images by averaging or other variations. === Special-case formula === In one simple special case, further explained in the next section, the SSIM measure between x {\displaystyle x} and y {\displaystyle y} is: SSIM ( x , y ) = ( 2 μ x μ y + c 1 ) ( 2 σ x y + c 2 ) ( μ x 2 + μ y 2 + c 1 ) ( σ x 2 + σ y 2 + c 2 ) {\displaystyle {\hbox{SSIM}}(x,y)={\frac {(2\mu _{x}\mu _{y}+c_{1})(2\sigma _{xy}+c_{2})}{(\mu _{x}^{2}+\mu _{y}^{2}+c_{1})(\sigma _{x}^{2}+\sigma _{y}^{2}+c_{2})}}} with: μ x {\displaystyle \mu _{x}} the pixel sample mean of x {\displaystyle x} ; μ y {\displaystyle \mu _{y}} the pixel sample mean of y {\displaystyle y} ; σ x 2 {\displaystyle \sigma _{x}^{2}} the sample variance of x {\displaystyle x} ; σ y 2 {\displaystyle \sigma _{y}^{2}} the sample variance of y {\displaystyle y} ; σ x y {\displaystyle \sigma _{xy}} the sample covariance of x {\displaystyle x} and y {\displaystyle y} ; c 1 = ( k 1 L ) 2 {\displaystyle c_{1}=(k_{1}L)^{2}} , c 2 = ( k 2 L ) 2 {\displaystyle c_{2}=(k_{2}L)^{2}} two variables to stabilize the division with weak denominator; L {\displaystyle L} the dynamic range of the pixel-values (typically this is 2 # b i t s p e r p i x e l − 1 {\displaystyle 2^{\#bits\ per\ pixel}-1} ); k 1 = 0.01 {\displaystyle k_{1}=0.01} and k 2 = 0.03 {\displaystyle k_{2}=0.03} by default. === General formula and components === The SSIM formula is based on three comparison measurements between the samples of x {\displaystyle x} and y {\displaystyle y} : luminance ( l {\displaystyle l} ), contrast ( c {\displaystyle c} ), and structure ( s {\displaystyle s} ). The individual comparison functions are: l ( x , y ) = 2 μ x μ y + c 1 μ x 2 + μ y 2 + c 1 {\displaystyle l(x,y)={\frac {2\mu _{x}\mu _{y}+c_{1}}{\mu _{x}^{2}+\mu _{y}^{2}+c_{1}}}} c ( x , y ) = 2 σ x σ y + c 2 σ x 2 + σ y 2 + c 2 {\displaystyle c(x,y)={\frac {2\sigma _{x}\sigma _{y}+c_{2}}{\sigma _{x}^{2}+\sigma _{y}^{2}+c_{2}}}} s ( x , y ) = σ x y + c 3 σ x σ y + c 3 {\displaystyle s(x,y)={\frac {\sigma _{xy}+c_{3}}{\sigma _{x}\sigma _{y}+c_{3}}}} The SSIM for each block is then a weighted combination of those comparative measures: SSIM ( x , y ) = l ( x , y ) α ⋅ c ( x , y ) β ⋅ s ( x , y ) γ {\displaystyle {\text{SSIM}}(x,y)=l(x,y)^{\alpha }\cdot c(x,y)^{\beta }\cdot s(x,y)^{\gamma }} Choosing the third denominator stabilizing constant as: c 3 = c 2 / 2 {\displaystyle c_{3}=c_{2}/2} leads to a simplification when combining the c and s components with equal exponents ( β = γ {\displaystyle \beta =\gamma } ), as the numerator of c is then twice the denominator of s, leading to a cancellation leaving just a 2. Setting the weights (exponents) α , β , γ {\displaystyle \alpha ,\beta ,\gamma } to 1, the formula can then be reduced to the special case shown above. === Mathematical properties === SSIM satisfies the identity of indiscernibles, and symmetry properties, but not the triangle inequality or non-negativity, and thus is not a distance function. However, under certain conditions, SSIM may be converted to a normalized root MSE measure, which is a distance function. The square of such a function is not convex, but is locally convex and quasiconvex, making SSIM a feasible target for optimization. === Application of the formula === In order to evaluate the image quality, this formula is usually applied only on luma, although it may also be applied on color (e.g., RGB) values or chromatic (e.g. YCbCr) values. The resultant SSIM index is a decimal value between -1 and 1, where 1 indicates perfect similarity, 0 indicates no similarity, and -1 indicates perfect anti-correlation. For an image, it is typically calculated using a sliding Gaussian window of size 11×11 or a block window of size 8×8. The window can be displaced pixel-by-pixel on the image to create an SSIM quality map of the image. In the case of video quality assessment, the authors propose to use only a subgroup of the possible windows to reduce the complexity of the calculation. === Variants === ==== Multi-scale SSIM ==== A more advanced form of SSIM, called Multiscale SSIM (MS-SSIM) is conducted over multiple scales through a process of multiple stages of sub-sampling, reminiscent of multiscale processing in the early vision system. It has been shown to perform equally well or better than SSIM on different subjective image and video databases. ==== Multi-component SSIM ==== Three-component SSIM (3-SSIM) is a form of SSIM that takes into account the fact that the human eye can see differences more precisely on textured or edge regions than on smooth regions. The resulting metric is calculated as a weighted average of SSIM for three categories of regions: edges, textures, and smooth regions. The proposed weighting is 0.5 for edges, 0.25 for the textured and smooth regions. The authors mention that a 1/0/0 weighting (ignoring anything but edge distortions) leads to results that are closer to subjective ratings. This suggests that edge regions play a dominant role in image quality perception. The authors of 3-SSIM have also extended the model into four-component SSIM (4-SSIM). The edge types are further subdivided into preserved and changed edges by their distortion status. The proposed weighting is 0.25 for all four components. ==== Structural dissimilarity ==== Structural dissimilarity (DSSIM) may be derived from SSIM, though it does not constitute a distance function as the triangle inequality is not necessarily satisfied. DSSIM ( x , y ) = 1 − SSIM ( x , y ) 2 {\displaystyle {\hbox{DSSIM}}(x,y)={\frac {1-{\hbox{SSIM}}(x,y)}{2}}} ==== Video quality metrics and temporal variants ==== It is worth noting that the original vers
Dr.Fill
Dr.Fill is a computer program that solves American-style crossword puzzles. It was developed by Matt Ginsberg and described by Ginsberg in an article in the Journal of Artificial Intelligence Research. Ginsberg claims in that article that Dr.Fill is among the top fifty crossword solvers in the world. == History == Dr.Fill participated in the 2012 American Crossword Puzzle Tournament, finishing 141st of approximately 650 entrants with a total score of just over 10,000 points. The appearance led to a variety of descriptions of Dr.Fill in the popular press, including The Economist, the San Francisco Chronicle and Gizmodo. A description of Dr.Fill appeared on the front page of the March 17, 2012 New York Times. Dr.Fill's score in 2013 improved to 10,550, which would have earned it 92nd place. Videos of the program solving the problems from the tournament are available on YouTube. The score in 2014 improved further to 10,790, which would have tied for 67th place. A video of the program solving the first six puzzles from that tournament, together with a talk given by Ginsberg describing its performance, can be found on YouTube. Dr.Fill has largely continued to improve since the 2014 event. In 2015, it scored 10,920 points and finished in 55th place. In 2016, it scored 11,205 points and finished in 41st place. In 2017, it scored 11,795 and finished in 11th place. In 2018, it scored 10,740 points, dropping to 78th place. Dr.Fill returned to "form" in 2019, once again scoring 11,795 and finishing in 14th place. The 2020 ACPT was cancelled due to COVID-19, and Dr.Fill participated as a non-competitor in the Boswords tournament instead. The program outperformed the humans, scoring 11,218 points (fast solves with a total of one mistake) while the best scoring human scored 10,994 points (slower solves but no mistakes). The 2021 ACPT was virtual, again due to COVID-19. The Dr.Fill effort was joined by the Berkeley NLP Group, creating a hybrid system named Berkeley Crossword Solver, and Dr.Fill won the main event, scoring 12,825 points with Erik Agard, the highest scoring human, scoring 12,810 points. The tournament was won by Tyler Hinman (12,760 points), who completed the championship puzzle perfectly in three minutes. Dr.Fill also completed that puzzle perfectly, but in 49 seconds. After winning the tournament, Ginsberg announced on August 8, 2021, that both he and Dr.Fill would be retiring from crosswords. == Algorithm == As described by Ginsberg, Dr.Fill works by converting a crossword to a weighted constraint satisfaction problem and then attempting to maximize the probability that the fill is correct. Probabilities for individual words or phrases in the puzzle are computed using relatively simple statistical techniques based on features such as previous appearances of the clue, number of Google hits for the fill, and so on. In doing this, Dr.Fill is attempting to solve a problem similar to that tackled by the Jeopardy!-playing program Watson; Dr.Fill runs on a laptop instead of a supercomputer and Ginsberg remarks that Watson is far more effective than Dr.Fill at solving this portion of the problem. Instead of computational horsepower, Dr.Fill relies on the constraints provided by crossing words to refine its answers. A variety of techniques from artificial intelligence are applied to attempt to find the most likely fill. These include a small amount of lookahead, limited discrepancy search, and postprocessing. Ginsberg remarks that postprocessing was chosen over branch and bound because the two techniques are mutually incompatible and postprocessing was found to be more effective in this domain.