Researchers from Meta’s FAIR team and The Hebrew University of Jerusalem have discovered that forcing large language models to “think” less actually improves their performance on complex reasoning tasks. The study found that shorter reasoning processes in AI systems lead to more accurate results while significantly reducing computational costs. The researchers discovered that within the same reasoning task, “shorter reasoning chains are significantly more likely to yield correct answers — up to 34.5% more accurate than the longest chain sampled for the same question.” This finding held true across multiple leading AI models and benchmarks. Based on these findings, the team developed a novel approach called “short-m@k,” which executes multiple reasoning attempts in parallel but halts computation once the first few processes complete. The final answer is then selected through majority voting among these shorter chains. The researchers found their method could reduce computational resources by up to 40% while maintaining the same level of performance as standard approaches. “Our findings suggest rethinking current methods of test-time compute in reasoning LLMs, emphasizing that longer ‘thinking’ does not necessarily translate to improved performance and can, counter-intuitively, lead to degraded results,” the researchers conclude. The study points toward potential cost savings and performance improvements by optimizing for efficiency rather than raw computing power.
Integration challenges requiring asynchronous transaction processing flow with multiple steps for additional layer of authentication is a key factor behind low adoption of EMV 3DS
While EMV 3DS has many benefits, adoption may be slow in the regions where EMV 3DS is not mandatory. Reasons may include the following: Data inconsistency. The quality of merchant data provided in EMV 3DS plays a critical role in issuer fraud detection. Merchants may be reluctant to share more data and may decide to provide a minimum set of data elements excluding optional data elements. There are cases when the data provided is not accurate, causing issues in fraud engines. Approval rates and cardholder friction. Shopping cart abandonment has been one of the major reasons that EMV 3DS adoption is low. Many enhancements have been added to the protocol from EMV 3DS 1.0 to EMV 3DS 2.x to challenge the cardholder only when needed. Complexity of integration. EMV 3DS integration is complex and adds an additional layer of authentication flow before authorization, resulting in higher implementation costs. Most systems are built with synchronous authorization request and response; EMV 3DS is a major change since it requires an asynchronous transaction processing flow with multiple steps. Liability shift. EMV 3DS is designed to help with fraud. However, determining if, how and when a liability shift occurs for merchants is not a simple answer and depends on several factors. Payment network and local regulatory requirements should be checked for specific use cases to assess any applicable liability shifts. Some factors are: Region and payment network. It is important to be familiar with payment network rules for EMV 3DS usage. Merchant category code (MCC). Not all MCCs are allowed for a liability shift.
AI chatbots prioritize relevance and authority over traditional SEO metrics such as traffic or backlinks, requiring brands to offer “citable evidence” of expertise, such as real data, FAQs, testimonials and contextual content
Local Falcon’s research showed that AI chatbots often bypass traditional SEO metrics such as link volume or page ranking, relying instead on relevance, prominence and authority. That means brands that surface in AI responses are those that offer “citable evidence” of expertise, such as real data, FAQs, testimonials and contextual content that AI can easily access, the study said. AI chatbots don’t prioritize traffic or backlinks when they choose which brands to list in their responses to consumers. Instead, they look for provable expertise. The good news is that smaller brands that are expert in what they do can leapfrog over well-trafficked retail sites in search listings. For example, stores shouldn’t just say they are the leading retailer in a niche product but actually prove it by explaining why they’re the best. It also helps for the brand to be mentioned in places like social media. For smaller businesses, it’s an opportunity to compete with larger players — if they provide genuine value and clear proof of expertise. “It gives smaller stores a shot at being found if they play it right and they’re really good at what they do,” David Hunter, CEO of Local Falcon said. Hunter thinks that Google transitioning from traditional search to AI-powered search is the cause of this surprise finding. Ultimately, location will continue to play an important part in search even for an AI chatbot. Location-based marketing and messaging is a powerful tool that drives higher engagement, foot traffic and revenue, according to Radar CEO Nick Patrick. Location is important not just for retailers but also for financial services — to do things such as find bank branches, real-time fraud detection and geo-triggered cash back promotions through retail partners. But instead of location, the AI appears to favor authoritative results from sources including social media, Reddit, and other community forums, which were previously not prioritized by traditional algorithms.
Odyssey’s AI world model uses a 360-degree, backpack-mounted camera system to capture real-world landscapes that lets users “interact” with streaming video and explore areas within a video, similar to a 3D-rendered video game
Odyssey is taking a different approach than many AI labs in the world modeling space. Odyssey has developed an AI model that lets users “interact” with streaming video. It designed a 360-degree, backpack-mounted camera system to capture real-world landscapes, which Odyssey thinks can serve as a basis for higher-quality models than models trained solely on publicly available data. The model generates and streams video frames every 40 milliseconds. Via basic controls, viewers can explore areas within a video, similar to a 3D-rendered video game. Powering this is a new world model, demonstrating capabilities like generating pixels that feel realistic, maintaining spatial consistency, learning actions from video, and outputting coherent video streams for 5 minutes or more.” World models could one day be used to create interactive media, such as games and movies, and run realistic simulations like training environments for robots. “Interactive video … opens the door to entirely new forms of entertainment, where stories can be generated and explored on demand, free from the constraints and costs of traditional production,” the company says. “Over time, we believe everything that is video today — entertainment, ads, education, training, travel, and more — will evolve into interactive video, all powered by Odyssey.” The model can currently stream video at up to 30 frames per second from clusters of Nvidia H100 GPUs at the cost of $1 to $2 per “user-hour.”
Improving stablecoin UX and trust requires merchants to verify crypto wallet addresses, look for security indicators like hardware wallet support and ENS (Ethereum Name Service) domain names and comply with local KYC/AML laws for higher-value transactions
Stablecoins aim to solve the crypto payments bottleneck, but bridging this gap requires resolving significant infrastructure and user experience challenges. Despite their simplicity on the surface, stablecoin payments involve a complex back end. Despite the technological advances, stablecoins are not immune to threats, and their rise has attracted new classes of digital fraud. For consumers, one of the biggest risks is fake apps or phishing links that mimic real wallets or payment interfaces. Merchants need to educate their customers — and themselves — on how to verify wallet addresses and look for security indicators like hardware wallet support and ENS (Ethereum Name Service) domain names. Consumer friendly also doesn’t mean compliance light. Merchants accepting stablecoins must ensure they comply with local know your customer/anti-money laundering (KYC/AML) laws, especially for higher-value transactions. “Even if stablecoins are the preferred medium for a lot of criminal activity, creating a regulated environment where these companies can operate in conjunction with law enforcement is probably a positive,” Dan Boyle, partner at Boies Schiller Flexner, told. The bottom line is that, for consumers and merchants, the potential shift stablecoins represent isn’t just technological — it’s psychological. Trust, familiarity and education are critical. And while today’s stablecoin experience is faster and cheaper than legacy systems, the true breakthrough will come when it becomes invisible.
Kurrent’s MCP Server features self-correction when prototyping, allowing developers to test projection, eliminating the traditional learning curve for database interactions
Event-native data platform innovator Kurrent has launched the open source KurrentDB MCP Server, which makes it possible for developers to use AI agents to interact with data inside KurrentDB without writing code. Users can now not only read and write events to the database but also create, test and debug projections or extend the available prompts to prototype applications using natural language. This is a capability exclusive to Kurrent that eliminates the traditional learning curve for database interactions. Kurrent MCP Server features self-correction when prototyping, allowing developers to test projection logic and help debug issues through conversational commands. Compatible with all frontier AI models and released under MIT license for community contribution and enterprise adoption without licensing restriction, Kurrent’s MCP Server makes development tasks that previously required hours of coding and corrections achievable in minutes. “Our new MCP Server makes it possible to use the main features of the KurrentDB database, like reading and writing events to streams and using projections, in a way that’s as simple as having a conversation,” said Kirk Dunn, CEO at Kurrent. “The system’s ability to test and fix itself reduces the need for debugging and increases reliability. Copilots and AI assistants become productive database partners rather than just code generators, seamlessly interfacing with KurrentDB.” The Kurrent MCP Server is compatible with every frontier model including Claude, GPT-4 and Gemini, and can be used with Cursor, Windsurf and other MCP-supported IDEs, clients and agent frameworks. It offers eight core capabilities that respond to natural language prompts, eliminating traditional database interaction friction points. These features make visual and analytical exploration conversational and intuitive. Kurrent MCP Server’s unique self-correction capability also sets it apart, making prototyping projections much easier and faster.
Peer, an AI-native platform reinvents the internet as a persistent, explorable universe, where users show up as avatars, connected by location, and building relationships in a living, spatial network
Peer, the AI-native platform reinventing the internet as a persistent, explorable universe, launched its Global Simulation—a real-time digital Earth where users show up as avatars, connect by location, and build relationships in a living, spatial network. Peer is designed as a native environment for agentic AI—an operating system where AI can move, perceive, and act within a shared world. By giving AI form and placing it in a common, persistent, and spatial environment, Peer creates a natural interface for human-AI interaction. It can be a Web3 or Web2 experience, said Tony Tran, CEO of Peer. Every user is paired with their own AI agent—an intelligent companion that learns, evolves, and operates alongside them. This makes AI feel personal, useful, and embodied—an essential step toward making AI adoption universal and seamlessly woven into daily life. At the core of Peer is a dynamic 3D map of the world—a new substrate where digital experiences are anchored to real places. This isn’t a game world or escapist simulation—it’s a living layer that blends the physical and digital into one continuous environment. Users appear as avatars tied to real-world and virtual locations, with full control over their visibility and presence. Powered by GPS and VPS (Virtual Positioning System), Peer transforms the map into an interactive space where people can explore, drop content, discover others, and build communities. Unlike traditional apps or games, this is a digital reality that starts from the real world—making it feel authentic, grounded, and alive. By combining a simulation of Earth with generative AI, avatars, and real-time spatial interaction, Peer creates a new operating system for human-AI coexistence. Peer is currently rolling out its first phase and will expand into AI agents, immersive worldbuilding, and custom hardware.
BaaS Griffin is creating an agentic bank enabling an MCP server to have agents that can open accounts, make payments, and analyze historic events
Banking-as-a-Service bank Griffin is opening up access to a Model Context Protocol (MCP) server, providing a way for AI agents to autonomously perform tasks on behalf of customers. Griffin, which secured a full banking licence in March last year, says the initiative is the beginning of massive technological platform shift, which will see people delegating more and more of their work to AI. “We think there is much further to go…but to get there, the financial system has to be fundamentally rewired to accommodate a world in which agents can freely transact — while still retaining appropriate safeguards.” Potential use cases cited include end-to-end wealth management, payment admin and transactional capabilities. “This is early for us – we’re in beta – but it shows the power of what’s possible,” says the bank. “You can use the Griffin MCP server to have an agent open accounts, make payments, and analyse historic events. You can also use it to build complete prototypes of your own fintech applications on top of the Griffin API – which we’re already seeing customers doing in real time.”
Fintech Chime is said to plan IPO launch as soon as Monday weighing a valuation of about $11 billion
Chime Financial Inc. is planning to launch its initial public offering as soon as Monday, as US listings bounce back from a disappointing April. The no-fee banking services company is weighing a valuation of about $11 billion. Chime was valued at $25 billion after raising $750 million in a funding round in 2021, according to an announcement at the time. The firm which filed for a listing with the US Securities and Exchange Commission in May, is set to disclose how much it will seek to raise in the IPO. No final decisions have been made and details of the offering including valuation and timing could change. Chime disclosed net income of $12.9 million on revenue of $518.7 million for the first three months of 2025, according to its latest filing. That compares with net income of $15.9 million on revenue of $392 million a year earlier. Chime’s largest investors include affiliates of DST Global, Crosslink Capital, Len Blavatnik’s Access Industries, General Atlantic, Menlo Ventures, Sino French (Innovation) Fund and Iconiq Strategic Partners, the filing shows. The offering is being led by Morgan Stanley, Goldman Sachs Group Inc. and JPMorgan Chase & Co. The company plans for its shares to trade on the Nasdaq Global Select Market under the symbol CHYM.
Stripe co-founder says banks are ‘very interested’ in how to integrate stablecoins into their product offerings
Payments firm Stripe has reportedly held initial talks with banks about using stablecoins. These discussions are happening as Stripe is rolling out a series of stablecoin-related products. “In the conversations we have with them, they’re very interested,” Stripe President and Co-Founder John Collison said. “This is not something that banks are just kind of brushing away or treating as a fad. Banks are very interested in how they should be integrated with stablecoins into their product offerings as well.” Stripe is among several companies in the FinTech space — including PayPal, FIS, and Fiserv — that are using stablecoins as a method of payment, and not just something used in crypto trading. “Regulated bank-issued stablecoins offer faster, more efficient and globally accessible payment options,” said Julia Demidova, head of digital currencies product and strategy at FIS. “With proper regulation, banks will become central players in digital assets, driving innovation while ensuring consumer protection.”