Cover Corp. has expanded outside of Japan, bringing virtual creators known as VTubers to U.S. audiences. The market is getting its test now as to whether VTubers where humans give voice to 3D avatars in concert performances can catch fire on the global stage. And the big talent agency is hoping that the phenomenon that has taken offRead More

Steam breaks concurrent record again, topping 40m simultaneous players for first time | News-in-briefA record-breaking 12.6m of those players were actively in-game at the time, too News by Vikki Blake Contributor Published on March 3, 2025 This is a News-in-brief article, our short format linking to an official source for more information. Read more about this story by following the link below:Steam breaks concurrent record again, topping 40m concurrent players for first time

Sumo Group sells publishing business Secret Mode for undisclosed sumCurrent leadership Blincoe and Schall "have the full backing of Emona and will remain at the helm"Image credit: Secret Mode News by Vikki Blake Contributor Published on March 3, 2025 Sumo Group has sold its publishing business, Secret Mode, for an unknown sum to Emona Capital and Secret Mode's current leadership, Ed Blincoe and James Schall.Whilst Emona Capital, a financial investor with "deep expertise in the gaming industry," already has a portfolio that includes Relic Entertainment and Amber Studio, Secret Mode marks its first investment in game publishing.Secret Mode will reportedly continue to operate as an independent publisher, focused on supporting indie PC and console games. Secret Modes current leadership, Blincoe and Schall, "has the full backing of Emona and will remain at the helm, leading the company as it enters its next phase.""Secret Mode is deeply passionate about the creativity and innovation within the indie space, and so we are thrilled to be partnering with Emona Capital and to fully focus on the independent publishing scene," Blincoe said."I also want to express our gratitude to Sumo not only for believing in Secret Modes original vision and overseeing our startup years, but for granting us this opportunity to become a fully independent entity."Schall added: "By becoming independent, we can now put all our efforts into uncovering the indie gems that gamers never knew they needed in their lives. We are committed to bringing our players the very best indie titles; not just of the future, but also by helping them discover existing indie masterpieces that have undeservedly flown under the radar in years gone by. We are working hard to empower independent developers across the globe, and hope to make some exciting announcements on this front in the coming months."Sumo Group launched its publishing division Secret Mode back in 2021 to handle internally developed games, plus projects from other independent studios. However, last month, Sumo Group announced it would be refocusing its business to cater "exclusively on development services for partners", resulting in a restructure and potential layoffs.

Game designer Zach Barth has a new venture by the name of Coincidence. Instead of being a traditional game developer like his first studio Zachtronics, he says this studio is more of a "flexible business framework" that lets small teams work on solo or group projects, video games or otherwise.Coincidence also revealed Kaizen: A Factory Story, a puzzle automaton game about creating and optimizing factory lines in 1980s Japan. In the announcement, Barth said the company hoped to bring "the open-ended design were known for to a broader audience without compromising on depth or complexity."Kaizen is its first official game project, following the contracted edutainment work Add Astra for an elementary school teacher. The game will be published by Astra Logical later in 2025 on Steam, and "twist your brain into beautiful, efficient knots."Building a game studio is a puzzle all of its ownBarth's game industry career began with the founding of Zachtronics in 2000 and free browser games on his website; one of them, the block-building title Infirmer, inspired Mojang to create Minecraft. Zachtronics' first commercial game was 2011's SpaceChem, which tasked players with using a visual programming language to create chemical modules on assembly lines.While it's tried its hand at wargame and strategy titles, puzzle games have largely been the studio's specialty. Over the years, it became known for Infinifactory, Shenzen I/O, and Last Call BBS. Its final game was 2022's The Zachtronics Solitaire Collection.Speaking of cards, Coincidence also has two physical card games under its belt, The Lucky Seven and Chemistry Set. Barth designed both of them and collaborated with artists and fellow designers Jonathan Stroh, Jay Ackerman, and Drew Messinger-Michaels, and the two games show how serious Coincidence is about not being wed to a particular medium.That freedom may give it more flexibility at a time when developers are closing down or cutting staff on a regular basis, and let it pick and choose when to start up a new team or end an old one, potentially without conducting layoffs.Game Developer spoke with the Zachtronics team numerous times over the years, the last of which was a 2020 interview with designer Matthew Seiji Burns about its then-upcoming visual novel, Eliza.

Various Oculus Rift games from developer Insomniac Games appear to have been delisted from the Meta store, and without any fanfare from the studio.Journalist Dominic Tarason first spotted the delistings, noting the titles were "big, high-budget PC VR games" only available for the virtual reality device. According to him, the impacted titles include 2016's The Unspoken, Edge of Nowhere, and Feral Rites, and 2019's Stormland. When he contacted Meta for an explanation, the customer support told him to "ask Insomniac."At time of writing, the studio itself has not issued a public statement on the matter. In a reply, Tarason believed the delistings happened "some time ago," but their being unavailable to purchase went generally unnoticed. "Others are so thoroughly vanished that not even pirate and torrent sites seem to have the rest anymore," he said. "That's kinda scary, considering that Insomniac are a high-profile triple-A studio."Before it was acquired by PlayStation and making Marvel games, Insomniac put Ratchet & Clank aside to work on several original titles. The most recognizable were Fuse and Sunset Overdrive, but the developer also made phone games (Fruit Fusion, Digit & Dash) and metroidvanias (Song of the Deep). Its handful of VR titles came out during this experimental phase, and it hasn't dabbled with the technology since, not even to port any of its more recent works to either PlayStation VR system.The four VR games were all published by Oculus Studios, which has previously teamed with other prominent non-VR developers like Turtle Rock Studios and Gunfire Games. It's also purchased Sanzaru Games and the late Ready at Dawn, both of which previously worked on spinoffs for PlayStation franchises Ratchet & Clank and God of War.Insomniac's apparent silence on these games' disappearance also underlines the need for a widespread preservation effort for the medium. If studios are unable to speak on their games going away for whatever reason, the titles deserve to be archived at the very least.

At least a dozen probationary staffers at the Federal Trade Commission were terminated last week, The Verge has learned.The terminations took place across the agency, according to two sources familiar with the matter, one of whom said that included both the Bureau of Consumer Protection and Bureau of Competition. The sources did not definitively link the terminations to actions by Elon Musks Department of Government Efficiency (DOGE), but the move followed a familiar DOGE playbook: apparently indiscriminate cuts targeting probationary employees, who may be new to the agency or a specific role. The FTC did not immediately respond to a request for comment.This appears to be one of the first times during the current Trump administration that the FTC has been impacted by staffing cuts of this scope, even as DOGE has driven mass firings across the federal government. Republican FTC Chair Andrew Ferguson has largely aligned himself with Trump, but the cuts indicate that the agency still may not be safe from broader changes to government staffing though its still unclear how far it will go.Are you a current or former US federal government worker? Reach out securely with tips from a non-work device to Lauren Feiner via Signal at laurenfeiner.64.In an email obtained by The Verge, the FTC chapter of the National Treasury Employees Union acknowledged to members last week that it was aware of terminations at the agency, but did not yet know their scope. The email also included information about ongoing legal efforts by the national organization, including the filing of claims at the Office of Special Counsel of the Merit Systems Protection Board, as well as institutional grievances at each agency where probationary employees have been terminated. The FTC union website now hosts a similar notice, with next steps for terminated employees.The Washington Post reported last week that the FTCs internal equal opportunity office had been cut from six to three staffers. Last month, Ferguson had declared that DEI is Over at the FTC, and said he had Terminated the Diversity Council at the agency. The scope of the recent probationary terminations, however, appear distinct from these directives.Also last week, FTC staff in Washington were told by DOGE that theyd soon be moving into the building being vacated by the US Agency for International Development (USAID), according to Bloomberg, which DOGE had effectively dismantled.See More:

SwitchBots inexpensive, adjustable smart roller shades are finally here. First announced in September, the SwitchBot Roller Shades are now available at switch-bot.com, claiming to be the worlds first adjustable smart shades. This means you dont need to worry about getting your measurements right before ordering, as you can adjust the width of both the shades headrail and the fabric to fit your window frame after buying them. Starting at $199, theyre cheap for smart shades.Smart motorized shades are a great convenience. They can be programmed to automatically open and close at set times of day or based on sunrise and sunset, and with an additional sensor, they can adjust based on temperature and light in the room. SwitchBots shades work over Bluetooth and are compatible with Matter, meaning they will work with Apple Home, Amazon Alexa, Google Home, and other smart home ecosystems through a SwitchBot hub.SwitchBots smart shades come in two colors, white and grey, and four sizes: small (22.8 to 31.5 inches) for $199.99), medium (31.5 to 43.3 inches) for $219.99, large (43.3 to 55.1 inches) for $259.99, and extra large (55.1 to 72.8 inches) for $309.99. Pick the size thats closest to your window and do the fine-tuning once you have them in hand.The shade headrail is adjustable to fit inside or outside of a window frame. Image: SwitchBotThat pricing is very competitive. Most motorized shades start around $400 for a small window, and while Ikea used to have an inexpensive option with its Fyrtur smart roller shades, those appear to have been discontinued. Lutron recently launched a lower-cost option to its excellent Serena shades, but those still start at $399.SwitchBots approach is to let you do more of the work at home, providing an adjustable headrail and a specialized cutting tool.Smartwings shades are comparable in price, at around $180, and they offer more fabric and light filtering options than SwitchBot. (SwitchBot says you can swap out the fabric, but it doesnt sell different options on its site, so I guess thats on you if you want a different look.) Plus, Smartwings has the option of Matter over Thread connectivity, so no need for a proprietary hub for smart home control. But you do need to make sure you get your measurements correct, as theyre not adjustable, meaning you could make a fairly expensive mistake if youre not great with a tape measure.The SwitchBot shades come with a small cutting tool that cuts the shade fabric to the correct length. Image: SwitchBotSwitchBots approach is to let you do more of the work at home, providing an adjustable headrail and a specialized cutting tool. Once youve got your headrail set to the correct size, you attach a cutting tool to the end of the shade and roll it until the entire length has been cut. According to Switchbots installation video, it looks pretty straightforward but also fairly time-intensive. Personally, Id rather a pro do that part for me, but if youre really worried about your measuring skills, its nice to have this as an option.The SwitchBot Roller Shade is battery-powered, and the company claims eight months of use per charge. It can be recharged via a USB-C cable or with a separate solar panel that costs $30. The latter also adds a light sensor, allowing for automation based on light levels.The shades can connect to the SwitchBot Hub 2 for Matter compatibility, which is also a temperature sensor, allowing for automating the shades based on the rooms climate. A wired remote control connected by a USB-C cable is included, or you can purchase a $20 SwitchBot wireless remote.See More:

Pi-thon 3.14 Comes Full Circle With New Optimizations 0 like March 3, 2025Share this postAuthor(s): Xuzmonomi Originally published on Towards AI. Two Upgrades that will Make Python a Tad Bit More EfficientThis member-only story is on us. Upgrade to access all of Medium.Python is beloved for its simplicity and readability, but when it comes to performance, it has some well-known bottlenecks.Unlike compiled languages like C++ or Java, Python is an interpreted language. This means the code is executed line by line at runtime, which makes development easier but slows execution speed.Another major limitation is the Global Interpreter Lock (GIL), which prevents multiple threads from running Python bytecode simultaneously. This restriction makes it hard to take full advantage of modern multi-core CPUs. Additionally, Pythons dynamic typing system where variable types are determined at runtime adds further overhead due to frequent type checks and conversions[1].And while automatic memory management through garbage collection makes development smoother, it introduces inefficiencies that impact performance in memory-intensive applications[2].But Python isnt standing still. Version 3.14 is packed with optimizations to make Python faster and more efficient. This article explores two game-changing upgrades: the new annotationlib module and a CPython tail call interpreter which promise to improve execution speed without compromising Pythons ease of use.Photo by Michael Dziedzic on UnsplashPython 3.14 introduces the annotationlib module, which provides tools for inspecting annotations on modules, classes, and functions. Annotations Read the full blog for free on Medium.Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AITowards AI - Medium Share this post

Author(s): Amir Javaheri Originally published on Towards AI. Part 1 of the Agentic Workflow SeriesSource: Image by Pete Linforth from PixabayIntroductionIn the rapidly evolving world of software development, were transitioning from static applications to dynamic, intelligent systems agents. These advanced AI-powered entities are not just reactive but are proactive, capable of perceiving their environment and autonomously acting to fulfill complex objectives. This paradigm shift introduces agentic workflows sophisticated sequences where agents plan, decide, and execute tasks, revolutionizing our technological interactions.To tap into this potential, structured approaches are crucial. Agentic design patterns, akin to traditional software patterns, provide robust frameworks for crafting effective agent behaviors, thus driving the creation of dynamic and responsive systems.Agentic Design Patterns: Structuring IntelligenceUnderstanding foundational agentic patterns is essential before delving into the specifics of the Tool Use Pattern. Each pattern addresses key aspects of agent functionality:The Reflection Pattern: Through iterative self-evaluation, the Reflection Pattern allows agents to learn from experience, driving ongoing performance enhancements. Specifically, it involves the agent analyzing the consequences of its previous actions, identifying errors or areas for improvement, and adjusting its internal model or strategies accordingly.The Planning Pattern: The Planning Pattern empowers agents to proactively chart their course, breaking down complex goals into manageable steps. It involves generating and refining action sequences, anticipating potential obstacles, and adapting to changing circumstances. By creating a roadmap, agents can navigate intricate tasks with greater efficiency and foresight, ensuring they move strategically towards their objectives.The Memory Pattern: Vital for continuity, this pattern allows agents to store and retrieve information, ensuring informed decisions based on context.The Multi-Agent Pattern: Orchestrates the interaction of multiple agents working towards a unified goal, ensuring cohesive operation and enhanced problem-solving capabilities.If youd like to dive deeper into these topics, feel free to follow me on Medium for more insights.The Tool Use Pattern: Empowering AgentsCentral to expanding an agents capabilities is the Tool Use Pattern. This pattern equips agents to integrate and utilize external tools, broadening their operational scope beyond inherent capabilities, which will be explored through detailed workflows and real-world applications.Step-by-Step Breakdown of Workflow1. User Input Reception: Setting the StageThe journey begins with user input, the catalyst for the agents actions. This can range from a simple question to a complex, multi-faceted request. The agent receives this natural language input, preparing it for the intricate process of understanding and action.2. Intent Recognition: Unveiling User GoalsAt the heart of understanding lies intent recognition. The system examines the users input to decode its underlying meaning. This involves:Identifying potential intents, each representing a user goal, with associated confidence scores.Extracting relevant parameters to provide context.For instance, Find climate change info and save it yieldssearch (query: climate change)file_operation (operation: write, path: some_file) intents.A sophisticated intent recognition system handles ambiguity, multiple intents, structured parameter extraction, and accurate confidence assignment.3. Tool Selection: Choosing the Right InstrumentsWith the users intent clarified, the agent selects the appropriate tools. This selection, driven by LLMs, machine learning, or rule-based systems, involves:Mapping intents to available tools.Filtering irrelevant or low-confidence tools.Determining optimal tool execution sequences.Preparing tool-specific parameters.The process considers intent confidence, tool capabilities, tool dependencies, and user preferences.4. Tool Input Generation: Bridging the GapThis crucial step transforms raw parameters into the precise format required by each tool. This ensures seamless tool execution through:Parameter Formatting: Converting extracted parameters into tool-specific formats (e.g., dates, codes).Parameter Validation: Ensuring parameter presence and validity (e.g., valid cities, future dates).Parameter Augmentation: Adding default or derived parameters (e.g., origin city, passenger count).Parameter Transformation: Performing necessary conversions (e.g., city to airport code, timezone adjustments).Context Integration: Incorporating conversation history or system state.Error Prevention: Catching parameter issues before execution.Security Filtering: Sanitizing inputs to prevent vulnerabilities.Example: Booking a FlightLets walk through an example to see how each step is completed before the tool is executed.User input: Book me a flight to Tokyo next Friday returning Sunday, economy class2. Intent Recognition:{ "intent": "book_flight", "confidence": 0.95, "parameters": { "destination": "Tokyo", "departure_date": "next Friday", "return_date": "Sunday", "cabin_class": "economy class" }}3. Tool Selection:{ "selected_tool": "FlightBookingTool", "confidence": 0.98, "intent_mapping": "book_flight FlightBookingTool"}4. Tool Input Generation4.1 Parameter Formatting{ "destination": "Tokyo", "departure_date": "2023-05-19", "return_date": "2023-05-21", "cabin_class": "economy"}4.2 Parameter Validation{ "validation_status": "valid", "checks_passed": [ "destination is a valid city", "departure_date is in the future", "return_date is after departure_date", "cabin_class is a valid option" ], "warnings": []}4.3 Parameter Augmentation{ "destination": "Tokyo", "departure_date": "2023-05-19", "return_date": "2023-05-21", "cabin_class": "economy", "origin": "New York", "passengers": 1, "preferred_airlines": ["ANA", "JAL", "United"], "meal_preference": "vegetarian", "checked_bags": 1}4.4 Parameter Transformation{ "departure_date": "2023-05-19T00:00:00-04:00", "return_date": "2023-05-21T00:00:00-04:00", "cabin_class": "Y", "origin": "JFK", "destination": "HND", "passengers": 1, "preferred_airlines": ["ANA", "JAL", "United"], "meal_preference": "vegetarian", "checked_bags": 1}4.5 Context Integration and 4.6 Error Prevention{ "execution_status": "proceed_with_warning", "parameters": { "departure_date": "2023-05-19T00:00:00-04:00", "return_date": "2023-05-21T00:00:00-04:00", "cabin_class": "Y", "origin": "JFK", "destination": "HND", "passengers": 1, "preferred_airlines": ["United", "ANA", "JAL"], "meal_preference": "vegetarian", "checked_bags": 1, "preferred_area": "Shinjuku, Tokyo", "loyalty_programs": { "United": "UA123456789" }, "trip_purpose": "business" }, "warnings": [ "Your trip is only 2 days. Given the long flight to Tokyo, you might want to consider a longer stay." ]}5. Tool Execution: Putting Tools into ActionThe agent executes the selected tools in sequence, each receiving its designated parameters. This involves:Tool performance of its designated function.Result capture, including success/failure status.Structured output data for later use.Feedback loop.Graceful error handling.Tools range from information retrieval to code execution, each producing a structured result with status, message, data, and error information.6. Response Generation: Crafting the Final OutputAfter tool execution, the agent generates a coherent response, considering:Original user input.Recognized intents.Results from executed tools.The response generator handles successful and failed executions, mixed results, and cases needing a response without tool execution. It prioritizes relevant information, formats tool outputs, provides context, and suggests next steps.Use Cases of the Tool Use Pattern:Code Execution: Automates tasks and solves computations through direct code execution. Example: An agent executes a Python script to calculate ROI or risk exposure.Document Processing: Extracts data, summarizes, translates, and performs OCR on documents. Example: An agent scans PDF invoices, extracts details like invoice number, date, and amount, and logs them into a database.API Integration: Accesses real-time data and services by connecting with APIs. Example: An agent retrieves real-time stock prices from a financial API and alerts users when a stock reaches a target price.Data Retrieval and Analysis: Extracts insights and performs analyses on data. Example: An agent queries a sales database to analyze monthly revenue trends and generate visual reports for decision-makers.Workflow Orchestration: Coordinates complex processes like customer service and supply chain management. Example: An agent automatically manages customer support tickets by routing them to the appropriate department, sending automated responses, and escalating urgent issues.Benefits of the Tool Use Pattern:Enhanced Capabilities: By extending functional reach, agents can perform tasks beyond their basic programming, pushing the boundaries of what automated systems can achieve.Increased Automation and Efficiency: Automation of complex tasks not only conserves resources but also enhances productivity, driving significant operational efficiencies.Adaptive Flexibility: The ability to integrate new tools as needs evolve allows agents to remain effective and relevant in rapidly changing technological landscapes.Seamless Integration: Agents can integrate within diverse systems, creating interconnected and intelligently responsive ecosystems.ConclusionThe Tool Use Pattern is not just a method but a cornerstone of modern agentic design, enabling the creation of agents that are not only intelligent but extraordinarily capable of engaging with the world in dynamic and meaningful ways. As we continue to advance in AI, the role of agentic design patterns becomes increasingly crucial, urging us to explore, innovate, and build the next generation of intelligent systems with the robustness and versatility provided by the Tool Use Pattern.ReferencesJoin thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AI

When Pokemon TCG Pocket first launched, the meta quickly came to be dominated by a very small handful of decks. One of those, centered around Misty and water-type Pokemon, came to be hated early on due to its potential to overpower opponents early on based entirely on how well some coin flips went. Now, three expansions later, you would think other cards would have swooped in to replace or counter Misty decks. But instead, Pokemon TCG Pocket's latest expansion just added a new card that makes Misty decks stronger than ever, and a lot of players are kind of over it.For context, it's not that Misty decks are necessarily the most powerful in the game. It's that the luck-based nature of Misty makes losing to them feel very, very bad. Misty is a Supporter card, with the unique ability to allow the user to pick a water-type Pokemon, and flip a coin until they land on tails. They then attach a water-type energy to that Pokemon for every heads they flipped. This could result in a player attaching 0 water-type energy, effectively wasting a deck spot on the card and a supporter card play. Or they could attach 1, 2, 3, 5, 10 energy, or more, if luck is on their side that day. A well-flipped Misty could allow a player to win on turn one in certain situations, before the opponent even has a chance to make a move. Or, in more common scenarios, the user gets enough bonus energy to bring extremely powerful cards online before their opponent can build a meaningful defense.All that's fine and dandy, but what's made Misty worse are the numerous cards in subsequent expansions that have made it even more powerful. Mythical Island added Vaporean, which has an ability allowing players to essentially move all that bonus energy freely around between their water-types. Space-Time Smackdown brought Manaphy in, which adds even more water energy to the board. Both expansions saw new, powerful water-type Pokemon that can make use of that energy, such as Palkia ex and Gyarados ex. As a result, water decks have been at the top of the meta consistently for multiple expansions in a row.And now, with Triumphant Light, there's another card out there making Misty decks even more hellish: Irida. Irida, like Misty, is a Supporter card, but its effect is that it can heal 40 damage from each Pokemon that has any water-type energy attached. Up to this point, grass-type decks have been the healing experts. But with Irida, water-type decks can make significant comebacks as long as they've collected and spread out enough energy...and between Misty, Manaphy, and Vaporeon, they're very likely to do so.For what it's worth, some Pokemon TCG experts are offering a pretty logical explanation for Irida. In TCG Pocket, decks are limited to 20 cards That means if you want to add something like Irida, you need to take something else out. Water decks are simply swimming in good cards - at a certain point, players must weigh whether or not it's actually worth removing Misty for a card like Irida. Some players are speculating that developer DeNA added Irida in an effort to crowd Misty out and force players to make tough choices about which Supporters to include. But that doesn't mean plenty of wise deckbuilders haven't found ways to include both.As the player above points out, Pokemon TCG Pocket is gearing up for a regularly-scheduled event, where rewards will be doled out for going on win streaks in the game's online competitive mode. The best reward, a gold profile badge, will be given out for winning five matches in a row. This is surprisingly difficult to do, and even moreso when many decks you'll be facing can sweep you away within the first few turns if they flip coins correctly, and have backups like Irida to save them even if the flips go bad for them.Expect to see a lot of water decks during this event and for a long time to come. In fact, at this point, it may just be worth playing one yourself if you're not already.Rebekah Valentine is a senior reporter for IGN. You can find her posting on BlueSky @duckvalentine.bsky.social. Got a story tip? Send it to rvalentine@ign.com.