Amplitude Studios announced Endless Legend 2, a long-awaited next entry in the award-winning fantasy strategy game series.Read More

Don't Nod delays second half of new narrative adventure IP Lost RecordsStudio says this is to "allow for additional optimisation and to give the team the necessary time to refine the experience" News by Vikki Blake Contributor Published on Jan. 23, 2025 Don't Nod has once again delayed the launch of its all-new IP, Lost Records: Bloom & Rage, this time pushing back the release date of the second instalment.Whilst the first part, Tape 1: Bloom, will release as planned on February 18, Tape 2: Rage has been pushed back a month to April 15.In a statement, the studio said this decision was made to "allow for additional optimisation and to give the team the necessary time to refine the experience for players and enhance the storytelling for a truly immersive continuation of the journey."Like Life Is Strange, Lost Records: Bloom & Rage is a narrative adventure that follows a group of teenagers caught up in a mystery involving supernatural elements.Don't Nod also delayed the launch of Lost Records: Bloom & Rage last summer, ostensibly to avoid clashing with the next instalment of Life Is Strange which was originally developed by Don't Nod.

Justin Carter, Contributing EditorJanuary 23, 20251 Min ReadImage via Everplay.At a GlanceThe newly-named Everplay will focus on its back catalog titles and franchises, as well as being a 'global indie powerhouse.'In February, Team17 Group PLC will have changed its name to Everplay Group PLC.The rebrand, according to the announcement, is a "key mechanism" that only affects its brand identity. Team17 as a developer, publisher, and game label will retain its name, as will the simulation-focused Astragon Entertainment (Police Simulator) and the edutainment maker StoryToys.Likewise, the core goal of becoming a "global indie powerhouse" remains the same. So why the change? CEO Steve Bell said it better reflects how the company's business has "evolved greatly" since its IPO in 2018 and "our DNA to never stop playing."As its three divisions focus on their individual specialities, Everplay will prioritize back catalog sales through "prioritizing our portfolio and evergreen franchises." It'll also allow for greater cross-collaboration, both internally and "across complementary sectors within the broader indie market."What's in a game company's name?Recent years have seen game developers and publishers rebrand, either with wholly new names or ones that better reflect what they make. For example, Focus Entertainment changed its name to Pullup Entertainment in January 2024, and like with Everplay, the change only affected what it was called.Conversely, 343 Industries becoming Halo Studios last October came with the studio committing to using Unreal Engine for future Halo projects. Similarly, Gearbox Publishing SF revamped as Arc Games when it remained with Embracer after Gearbox proper was acquired by Take-Two Interactive.Sometimes, a game itself will undergo a full-on rebrand: IO Interactive folded its three Hitman reboot games into Hitman: World of Assassination, which has since gained 75 million global players following its new name.Read more about:Studio announcementAbout the AuthorJustin CarterContributing Editor, GameDeveloper.comA Kansas City, MO native, Justin Carter has written for numerous sites including IGN, Polygon, and SyFy Wire. In addition to Game Developer, his writing can be found at io9 over on Gizmodo. Don't ask him about how much gum he's had, because the answer will be more than he's willing to admit.See more from Justin CarterDaily news, dev blogs, and stories from Game Developer straight to your inboxStay UpdatedYou May Also Like

Celeste creator Extremely OK Games (EXOK) has canceled production of its next title, Earthblade. Founder Maddy Thorson explained she and co-founder Noel Berry chose to scrap the game in December after the team began to "fracture."According to Thorson, a disagreement formed between her and Berry (referred to in the statement as "us") and fellow EXOK co-founder Pedro Medeiros concerning Celeste's IP rights. After resolving that dispute, "both parties agreed in the end that we should go our separate ways." Medeiros' leave wasn't the deciding factor in Earthblade's cancelation, but it made Thorson and Berry consider "whether fighting through to finish was the right path forward."Earthblade was first announced in 2021, and by her admission, development "was not as far along as one would expect," despite starting in 2019. The project "had a lot going for it," and Thorson believes it would have ended up great in the end. But the idea of toughing out development to reach that point was not worth it, and felt indicative of a "deeper problem."Much of what made development so daunting was Celeste's success and the pressure "to deliver something bigger and better," Thorson continued. She said the strife with Medeiros "gave us the clarity to see that we have lost our way, and the opportunity to admit defeat. I feel many ways about it, but one big feeling is undoubtedly relief."Extremely OK will be alright in the endAccording to Thorson, several EXOK staffers have "moved on" following Earthblade's cancelation. The current team has now pivoted to smaller-scale projects in order to "rediscover game development in a manner closer to how we approached it at Celeste's or TowerFall's inception." While she considers building up staff post-Celeste a "failure," she and Berry have made peace with it."We are happy to return to our roots and reclaim some joy in our creative process, and see where that takes us," she concluded.Thorson's full explanation for ending Earthblade and the studio's disagreement can be read here.

Perplexity has turned its AI answer engine into a mobile assistant on Android. The new assistant can answer general questions and perform tasks on your behalf, such as writing an email, setting a reminder, booking dinners, and more.Its also multimodal, meaning you can ask it questions about whats on your screen as well as have it open your camera and see whats in front of you. In an example shared by Perplexity, a user asks the assistant to get me a ride. Once it learns where the user wants to go, the assistant automatically opens Uber with available rides to that destination.I tried it out for myself, and it is kind of neat. When I asked it to open up a good podcast, my phone started playing the latest episode of The Joe Rogan Experience on YouTube. It worked rather quickly, even though its taste may be questionable.Perplexity gave me the rundown on these promotional Pokmon cards. Screenshots: The VergeUsing my phones camera, Perplexitys assistant successfully identified the promotional Pokmon pack I got in a McDonalds Happy Meal (dont judge), which I found impressive since the promotion only started a couple of days ago. It also helped me write and send a text to a family member using the information in my contacts.Alongside Samsungs announcement of the Gemini-equipped Galaxy S25, Google revealed that its AI assistant can now complete tasks across multiple apps, as well as complete multimodal requests.RelatedBut Perplexitys assistant doesnt work across every app and with every feature. Its not able to access Slack or Reddit, for example, and I also couldnt use it to leave a comment on a YouTube video. Right now, the assistant supports Spotify, YouTube, and Uber, along with email, messaging, and clock apps, according to Perplexity spokesperson Sara Platick. Were continuing to add support for more apps and more functionality though, so this is just the starting point, Platnick adds.You can enable the assistant through the Perplexity app, which prompts you to replace your phones default assistant with Perplexity. From there, you can swipe up on the left corner of your screen or hold down your home button to access the assistant.Its currently not available on the iPhone, however. If Apple gives us the right permissions, well make it happen, Platnick says.

Meta and YouTube arent the only platforms looking to benefit from TikTok potentially disappearing Substack wants in on the action, too.The company announced Thursday its launching a $20 million creator accelerator fund, promising content creators they wont lose revenue by jumping ship to Substack. Creators in the program also get strategic and business support from Substack, and early access to new features. We established this fund because weve seen creators who specialize in video, audio, and text expand their audience, revenue, and influence on Substack, where the platforms network effects amplify the quality and impact of the work theyre doing, the company said in a blog post.This pivot on Substacks part has been in the works for a while for months, the company has been marketing itself not as a newsletter delivery service but as a creator platform similar to Patreon. On Substack, [creators] can build their own home on the internet: one where creators, not platform executives or advertisers, own their work and their audience, the blog post reads. The post also cites bans, backlash, and policies that change with the political winds as a reason creators cant depend on traditional social media services.Thats all fine (we at The Verge have been saying this for a while). But creators focusing on Substack are also subject to ebbs and flows depending on what the company is prioritizing: first, it was newsletters, then it was tweet-like micro blogs, followed by full-on websites and livestreaming. For some, Substacks initial stated mission of giving more freedom to independent writers is fading. And TikTok creators looking to move to Substack will need to rebuild their following all over again you obviously cant export your TikTok followers.Substacks business is not immune from political winds, either. Just a year ago the company found itself in hot water over Nazi newsletters that were being monetized on the platform. Substack eventually banned some of the newsletters, but only after sustained and public pressure from high-profile writers. In 2022, when Substacks promise was to be a better version of a newsroom, the company abruptly fired an editor who had worked on a newsletter critical of Substack. The $20 million fund isnt the first time Substack has offered a pool of money meant to entice creators. Under a program called Substack Pro, the company poached top media talent from traditional newsrooms with higher pay, health insurance, and other perks. That program ended in 2022, with Substack cofounder Hamish McKenzie saying the deals werent employment arrangements but seed funding deals to remove the financial risk for a writer in starting their own business. In other words, welcome to Substack. Now that youre here, youre on your own which is more or less the deal other platforms offer.

Evaluating conversational AI systems powered by large language models (LLMs) presents a critical challenge in artificial intelligence. These systems must handle multi-turn dialogues, integrate domain-specific tools, and adhere to complex policy constraintscapabilities that traditional evaluation methods struggle to assess. Existing benchmarks rely on small-scale, manually curated datasets with coarse metrics, failing to capture the dynamic interplay of policies, user interactions, and real-world variability. This gap limits the ability to diagnose weaknesses or optimize agents for deployment in high-stakes environments like healthcare or finance, where reliability is non-negotiable.Current evaluation frameworks, such as -bench or ALMITA, focus on narrow domains like customer support and use static, limited datasets. For example, -bench evaluates airline and retail chatbots but includes only 50115 manually crafted samples per domain. These benchmarks prioritize end-to-end success rates, overlooking granular details like policy violations or dialogue coherence. Other tools, such as those assessing retrieval-augmented generation (RAG) systems, lack support for multi-turn interactions. The reliance on human curation restricts scalability and diversity, leaving conversational AI evaluations incomplete and impractical for real-world demands. To address these limitations, Plurai researchers have introduced IntellAgent, an open-source, multi-agent framework designed to automate the creation of diverse, policy-driven scenarios. Unlike prior methods, IntellAgent combines graph-based policy modeling, synthetic event generation, and interactive simulations to evaluate agents holistically.At its core, IntellAgent employs a policy graph to model the relationships and complexities of domain-specific rules. Nodes in this graph represent individual policies (e.g., refunds must be processed within 57 days), each assigned a complexity score. Edges between nodes denote the likelihood of policies co-occurring in a conversation. For instance, a policy about modifying flight reservations might link to another about refund timelines. The graph is constructed using an LLM, which extracts policies from system prompts, ranks their difficulty, and estimates co-occurrence probabilities. This structure enables IntellAgent to generate synthetic events as shown in Figure 4user requests paired with valid database statesthrough a weighted random walk. Starting with a uniformly sampled initial policy, the system traverses the graph, accumulating policies until the total complexity reaches a predefined threshold. This approach ensures events span a uniform distribution of complexities while maintaining realistic policy combinations.Once events are generated, IntellAgent simulates dialogues between a user agent and the chatbot under testa as shown in Figure 5. The user agent initiates requests based on event details and monitors the chatbots adherence to policies. If the chatbot violates a rule or completes the task, the interaction terminates. A critique component then analyzes the dialogue, identifying which policies were tested and violated. For example, in an airline scenario, the critique might flag failures to verify user identity before modifying a reservation. This step produces fine-grained diagnostics, revealing not just overall performance but specific weaknesses, such as struggles with user consent policiesa category overlooked by -bench.To validate IntellAgent, researchers compared its synthetic benchmarks against -bench using state-of-the-art LLMs like GPT-4o, Claude-3.5, and Gemini-1.5. Despite relying entirely on automated data generation, IntellAgent achieved Pearson correlations of 0.98 (airline) and 0.92 (retail) with -benchs manually curated results. More importantly, it uncovered nuanced insights: all models faltered on user consent policies, and performance declined predictably as complexity increased, though degradation patterns varied between models. For instance, Gemini-1.5-pro outperformed GPT-4o-mini at lower complexity levels but converged with it at higher tiers. These findings highlight IntellAgents ability to guide model selection based on specific operational needs. The frameworks modular design allows seamless integration of new domains, policies, and tools, supported by an open-source implementation built on the LangGraph library.In conclusion, IntellAgent addresses a critical bottleneck in conversational AI development by replacing static, limited evaluations with dynamic, scalable diagnostics. Its policy graph and automated event generation enable comprehensive testing across diverse scenarios, while fine-grained critiques pinpoint actionable improvements. By correlating closely with existing benchmarks and exposing previously undetected weaknesses, the framework bridges the gap between research and real-world deployment. Future enhancements, such as incorporating real user interactions to refine policy graphs, could further elevate its utility, solidifying IntellAgent as a foundational tool for advancing reliable, policy-aware conversational agents.Check out the Paper and GitHub Page. All credit for this research goes to the researchers of this project. Also,dont forget to follow us onTwitter and join ourTelegram Channel andLinkedIn Group. Dont Forget to join our70k+ ML SubReddit. Vineet Kumar+ postsVineet Kumar is a consulting intern at MarktechPost. He is currently pursuing his BS from the Indian Institute of Technology(IIT), Kanpur. He is a Machine Learning enthusiast. He is passionate about research and the latest advancements in Deep Learning, Computer Vision, and related fields. Meet 'Height':The only autonomous project management tool (Sponsored)

#59: The Agentic AI Era, Smolagents, and a Gatekeeper Agent Prototype 0 like January 23, 2025Share this postAuthor(s): Towards AI Editorial Team Originally published on Towards AI. Good morning, AI enthusiasts! As you already know, we recently launched our 8-hour Generative AI Primer course, a programming language-agnostic 1-day LLM Bootcamp designed for developers like you.We also have a special discount for our community members (yes, thats you!). Use code towardsai_8hour to get 15% off on the course. So dont wait, learn to make the most of LLMs before the next big AI update drops. Start here!As always, we have practical tutorials, collaboration opportunities, a fun prototype from the community, and a lot more!P.S. If you are already interested in our LLM developer course, use the code towardsai_8hour on our bundle offering, which includes this new course and our more in-depth companion course, From Beginner to Advanced LLM Developer. Check out the bundle offering here!Whats AI WeeklyThis week in Whats AI, Im diving into the world of APIs what they are, why you might need one, and what deployment options are available. When we talk about building powerful machine learning solutions, like large language models or retrieval-augmented generation, one key element that often flies under the radar is how to connect all the data and models and deploy them in a real product. This is where APIs come in. Read the complete article here or watch the video on YouTube! Louis-Franois Bouchard, Towards AI Co-founder & Head of CommunityLearn AI Together Community section!Featured Community post from the DiscordMalus_aiiola has built an AI voice Gatekeeper agent that handles incoming calls for busy CEOs. In this video, he breaks down how you can benefit from an AI voice Gatekeeper, which will answer for you and record the information of the caller. You can also get the blueprint and prompts to make it yourself. Reach out in the thread if you want to build something like this.AI poll of the week!Is Devin promising? Is it conceptually good, or do you see yourself adopting it? Tell us in the thread!Collaboration OpportunitiesThe Learn AI Together Discord community is flooding with collaboration opportunities. If you are excited to dive into applied AI, want a study partner, or even want to find a partner for your passion project, join the collaboration channel! Keep an eye on this section, too we share cool opportunities every week!1. Mr_oxo is looking for people to collaborate with on Computer Vision projects as accountability partners and problem-solving buddies. If youre passionate about computer vision and want to level up your skills while working on projects, connect in the thread!2. Pulkitplays is building a Next Ball Score Prediction Model and needs a project partner to solve the problem statement. If this sounds interesting, reach out in the thread!Meme of the week!Meme shared by rucha8062TAI Curated sectionArticle of the weekGraph Neural Networks: Unlocking the Power of Relationships in Predictions By Shenggang LiThis article explores Graph Neural Networks (GNNs), focusing on their ability to analyze connected data. It explains how GNNs interpret nodes and edges, using examples like cities connected by roads. The article then delves into three GNN types: Convolutional GNNs for economic forecasting, Graph Attention Networks for feature selection, and Temporal GNNs for stock prediction. Code examples and results demonstrate each GNNs application, including predicting economic indicators, identifying key dataset features, and forecasting stock trends. It highlights GNNs broader potential across diverse fields like healthcare, traffic management, and information retrieval with LLMs.Our must-read articles1. Smolagents + Web Scraper + DeepSeek V3 Python = Powerful AI Research Agent By Gao Dalie ()This article provides a tutorial on creating a multi-agent chatbot using Smolagents, a Python library for building AI agents, combined with web scraping and the DeepSeek V3 language model. The author highlights Smolagents simplicity, requiring minimal code to create agents capable of complex tasks, including Agent-Retrieval-Generation systems. The DeepSeek V3 model was chosen for its cost-effectiveness and performance, which are comparable to GPT-4 and Claude. It concludes by emphasizing Smolagents efficiency and ease of use for developing sophisticated AI agents.2. Building an On-Premise Document Intelligence Stack with Docling, Ollama, Phi-4 | ExtractThinker By Jlio AlmeidaThis article details building an on-premise document intelligence solution using open-source tools. It addresses the data privacy concerns of financial institutions by leveraging local language models (LLMs) like Phi-4 via Ollama, combined with ExtractThinker for orchestrating document processing and Docling/MarkItDown for handling document loading and OCR. It guides readers through choosing appropriate LLMs (text vs. vision-based), document parsing libraries, and local deployment solutions. It also provides strategies for managing limited context windows in local models, including lazy splitting for large documents and pagination for partial responses. Code examples demonstrate integrating these components for a complete document extraction pipeline, offering a secure and compliant solution for sensitive data processing. Finally, it discusses PII masking for cloud-based LLM usage when local deployment isnt feasible.3. Building Multimodal RAG Application #8: Putting it All Together! Building Multimodal RAG Application By Youssef HosniThis article presents a comprehensive guide to building a multimodal Retrieval-Augmented Generation (RAG) application, culminating the eight-part series. It integrates modules for preprocessing multimodal data, retrieving information using a LanceDB vector store, and generating responses with Large Vision Language Models (LVLMs) via a PredictionGuard client. It details setting up the environment, processing data (including image-text embedding), building the retrieval module, implementing LVLM inference, and processing prompts. Finally, it demonstrates combining these modules into a functional multimodal RAG system using LangChain, enabling complex queries against diverse data types like text and video.4. The Agentic AI Era: A Primer By Kaush BThis article provides a primer on the Agentic AI Era, exploring the evolution of AI from automation tools to autonomous agents. It defines AI agents, categorizing them by type and architectural topology, and outlines their characteristics and developmental stages within an enterprise context. It details various agentic workflows, including prompt chaining, routing, parallelization, orchestrator-workers, and evaluator-optimizer models. It also discusses four key agentic design patterns: reflection, tool use, planning, and multi-agent collaboration. Practical implementation guidelines, technical risks, and mitigation strategies are also addressed. It concludes by emphasizing the transformative potential of agentic AI while acknowledging the ethical and societal challenges that require careful consideration for responsible deployment.If you are interested in publishing with Towards AI, check our guidelines and sign up. We will publish your work to our network if it meets our editorial policies and standards.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

During today's Xbox Developer Direct, we got another look at upcoming Xbox RPG Clair Obscur: Expedition 33 along with a release date of April 24, 2025.In a new trailer, we got more info on the game's story, which we're told doesn't focus on good vs. evil. Rather it focuses on a group of complicated, messy people who all are brought together to defeat that which is a threat to all of them. There's a focus on the relationship between the characters, and while it can get introspective, there are also NPCs whose stories and mechanics are somewhat lighter. PlayIGN's Twenty Questions - Guess the game!IGN's Twenty Questions - Guess the game!To start:...try asking a question that can be answered with a "Yes" or "No".000/250As we've seen in past trailers, the UI and battle menus are exceptionally stylish, channeling Persona a bit in their big energy. The world itself is inspired by French art deco style, with art director Nicohlas Maxson-Francombe explaining in a recent Xbox Wire post that he wanted to avoid doing the same kinds of environments he's seen a million times before. "I just wanted to get out of the stuff that weve just seen a million times, science fiction, to space, or zombies, all that kind of stuff. I just wanted to get my head out of there and see what I can try to come up with that was original for me personally."And we also got a look at unique character mechanics and customization via skill tree in the trailer, and what seems to be a massive colletion of possible equippable passive effects. There are far more details in the same Xbox Wire post about the combat system and some of the characters' specific moves.It was also revealed that Clair Obscur will follow in the steps of old-school RPGs and have an actual, navigable world map from which you can access various areas and towns. The world map will also include secrets and even hidden bosses.PlayClair Obscur was first announced last summer at the Xbox Games Showcase from Montpellier-based Sandfall Interactive and came as a complete surprise to most. It takes place in a world where an entity known as the Paintress "repaints" the world each year, during which time she also paints a cursed number on a monolith that causes everyone of that age and older to turn to smoke instantly and disappear. Every year, that number ticks down by one, with younger and younger generations dying each time. With the number "33" up next, a group of individuals led by a man named Gustave sets out to try and stop the Paintress once and for all in a turn-based RPG with incredibly stylish battles and real-time elements.You can catch up on everything shown in today's Xbox Developer Direct right here.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.

During today's Xbox Developer Director, we saw a new trailer for South of Midnight complete with a release date of April 3, 2025 for those who purchase the premium edition, and April 8, 2025 for everyone else.South of Midnight is a third-person action-adventure game set in a fantastical version of the American Deep South. In the trailer, we saw a good look at main character Hazel's powers, which blend the traditional craft of weaving with magic to allow her to fight, conjure things, and overcome environmental obstacles. As she fights, Hazel "unravels" enemies to destroy them in action combat, and has some elements of customization in her fighting style that can help her counter different types of enemies.PlayIGN's Twenty Questions - Guess the game!IGN's Twenty Questions - Guess the game!To start:...try asking a question that can be answered with a "Yes" or "No".000/250Hazel's story starts with a massive hurricane that seemingly kills her mother, but also gives her the ability to see a magic world beyond the real one that gives her hope her mother might still be alive. As she hunts for her, Hazel will visit different different regions ruled by magical creatures, each one with a different Southern Gothic history that she'll need to heal by collecting their fragmented memories. In the trailer, we got a look at several of these creatures, including a real big bird and a massive crocodile covered in spikes with giant trees going out of its back. A giant talking Catfish acts as a guide for Hazel as she awakens her new powers.South of Midnight was first announced at the Xbox Showcase in 2023 as a new game from We Happy Few developer Compulsion Games. It reappeared at the Xbox Showcase last year, and we got a hands-off look at it back in November. In an Xbox Wire post today, it's suggested the game will take between 10-12 hours to finish.You can catch up on everything shown in today's Xbox Developer Direct right here.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.