An X user created a Minecraft clone in 20 minutes using Anthropic's AI model, Claude Fable 5. The game features multiple biomes, day-night cycles, a variety of ores, and intricate caves.
Google has released benchmark results for evaluating AI models in Android coding, revealing that the Gemini 3.5 Flash is the most resource-intensive model but ranks sixth overall. The benchmarks indicate that Gemini 3.5 Flash has higher latency and a 9% performance gap compared to its predecessor, Gemini 3.1 Pro Preview, despite being marketed as a faster alternative. In terms of cost, Gemini 3.5 Flash averages 355.9 tokens per benchmark run at approximately 7.1, while Gemini 3.1 Pro Preview uses only 73.3 tokens at about a third of that cost. The top-ranked models include GPT 5.5, GPT 5.4, and Gemini 3.1 Pro Preview, while Claude Opus 4.7 ranks fourth. The rankings feature both open-weight and closed-weight models, with the list remaining consistent since the last release, except for the removal of GPT 5.3 Codex.
On June 9, Anthropic launched the Claude Fable 5 “Mythos-class” model, which has gained attention for its capabilities in game development. The model can clone popular games like Minecraft and Pokémon efficiently, producing a Minecraft clone in 20 minutes and a Pokémon clone in one hour. Users have reported impressive results, including the recreation of all 151 Gen-1 Pokémon with real sprites and game mechanics. Claude Fable 5 is part of Anthropic's premium offerings, with a pay-as-you-go pricing model that is more expensive than previous models. Benchmarks show that both Mythos 5 and Fable 5 excel in various domains, including coding and cybersecurity.
Microsoft's latest Patch Tuesday addressed 198 security vulnerabilities, the most extensive update in recent memory. Among these, 32 flaws are classified as critical, and three are zero-day vulnerabilities. The updates are detailed in KB articles: KB5094126 for Windows 11 versions 24H2 and 25H2, KB5093998 for version 23H2, and KB5094127 for Windows 10. The updates will automatically download and install, but users must verify their installation status and reboot their computers for changes to take effect.
The vulnerabilities addressed this month are attributed to advancements in artificial intelligence, with companies like Microsoft leveraging AI models to expedite the identification and resolution of security flaws. The three zero-day vulnerabilities include one that allows an attacker to gain Windows System privileges through a flaw in file link resolution, another that could facilitate a denial-of-service attack via an HTTP vulnerability, and a third related to a flaw in Windows BitLocker that could enable data capture from an unpatched PC.
Additionally, the update introduces new features to Windows 11, including new Secure Boot certificates, a Low Latency Profile for enhanced performance, support for shared audio devices for multiple Bluetooth connections, webcam functionality across multiple applications, and the ability to assign a custom name to the user folder during setup.
Nvidia is developing the RTX Spark, a powerful chip that integrates AI agents into consumer PCs, in collaboration with Microsoft. This chip features a CPU and an RTX 5070-class GPU, providing 1 petaflop of computing power and up to 128GB of unified memory, enabling sophisticated AI agents to operate directly on devices without the need for dedicated AI PCs or cloud services. Windows will introduce new security frameworks to facilitate the use of AI agents, addressing previous limitations that hindered their effectiveness on PCs. Users will be able to delegate complex tasks to AI agents, which can learn from interactions and adapt to individual workflows, enhancing the computing experience. However, challenges such as market demand, compatibility issues, and competition from other companies may impact the success of the RTX Spark.
Microsoft is advancing in the field of artificial intelligence with the introduction of agentic AI, which aims to perform tasks on behalf of users rather than just responding to queries. A key feature is 'autoagents,' starting with Scout, which will integrate into Microsoft 365 applications and autonomously manage tasks like monitoring communications and summarizing documents. Scout is currently being rolled out to select users and is designed with safety measures, including Execution Containers to control resource access and mitigate risks. Additionally, Microsoft has launched Discovery, a tool for enhancing scientific research, and Rayfin, which simplifies application development for individual creators.
Microsoft is introducing Scout, its first Autopilot agent designed to enhance productivity within the Microsoft 365 ecosystem by autonomously managing tasks and providing timely notifications. Scout will integrate with applications like Teams, Outlook, OneDrive, and SharePoint, utilizing OpenClaw's open-source technology to manage emails, summarize documents, and allocate time for overdue tasks. To ensure safety, Microsoft has implemented Execution Containers, which define access parameters for agents and integrate with existing security tools like Defender and Intune. Additionally, Microsoft Discovery is now available to all researchers, allowing them to use AI in scientific research with a user-friendly chatbot interface. Microsoft also unveiled Rayfin, a tool that simplifies app development by enabling users to define backend systems through code and deploy them directly to Microsoft Fabric.
Google has initiated a confidential program to acquire app source code from select developers on the Play Store to enhance its AI-powered coding tools. The company is offering financial incentives to developers who share their app source code, allowing them to retain ownership and intellectual property rights. Google is specifically targeting real-world Android applications, including older projects, to improve its Gemini models and coding functionalities. The initiative aims to elevate the performance of Google's AI offerings, particularly the upcoming Gemini 3.5 and Antigravity 2.0 coding agent, and to position the company more competitively against rivals like Microsoft and Anthropic.
Google has reached out to select Android app developers with a proposal to monetize their codebases through a "confidential content offer pilot." Developers are invited to share their active production code and archived projects, allowing them to generate additional revenue while retaining full intellectual property rights. The licensing agreement is non-exclusive. The initiative aims to enhance Google's AI coding tools by utilizing real-world code from developers, including unused prototypes and side projects. This move comes as Google faces competition in the AI coding space from companies like Anthropic and Microsoft.
At the recent Google I/O event, Google introduced a suite of AI tools for Android and Chrome applications aimed at enhancing marketing and management for developers. In the Play Store, AI will assist in app discovery through Google’s Gemini and provide a Q&A interface for user inquiries. New AI-driven management tools for developers include creating app listings based on keyword insights, automating catalog management, analyzing payment issues, and offering retention incentives for users. Google also reduced Play Store charges for developers to a commission of 10% to 20%. For Android programmers, Google’s AI Studio offers coding assistance for specific app categories, while additional enhancements include a command-line interface guided by programming-assistant AIs and new marketing options. In Chrome, Google introduced a trial version of WebMCP for AI agents to connect with site APIs and expanded built-in AI tools for writing and debugging. New web-coding options are being tested to improve app-like interfaces and compatibility metrics. The “Immediate UI Mode” feature allows for a unified login interface to streamline passkey authentication.
