NVIDIA used AI to rewrite Kimi's attention kernel, resulting in a speed nearly three times that of the official version.

Beating AI Express News: X is testing a new unified subscription plan. An early Android test page leaked by the community features an "X Subscription" upgrade card, which states that a single plan will include access to Cursor, Grok, and X’s core benefits, with a Grok Bot icon displayed at the top. Maxime Prades, head of SpaceXAI’s product team, previously confirmed the company is developing a "more unified subscription" that will roll out in "weeks, not months." Currently, several subscription tiers have partial integration but remain charged separately: Cursor Pro, Pro+, and Ultra all include Grok Bot access; SuperGrok, SuperGrok Plus, Heavy, and X Premium+ also grant Grok Bot benefits to Cursor accounts. However, Cursor’s official statement clarifies that this linking does not convert Grok or X subscriptions into Cursor plans, nor does it cancel existing Cursor subscriptions.

According to an official announcement on the website of Japan’s Financial Services Agency (FSA), the regulator announced on September 29, 2026 that it has approved the fourth support project under the FinTech Proof-of-Concept Center’s Payment Advancement Project (PIP). The project was jointly applied for by TradeWaltz, NTT Data, Mizuho Bank, MUFG Bank, Sumitomo Mitsui Banking Corporation, and MUFG Trust Bank. The proof-of-concept experiment will verify the practical application of stablecoins in trade transaction fund settlement and sort out relevant legal arguments, running from September 2026. Upon conclusion, the FSA will publish the experiment results and relevant compliance and regulatory conclusions on its official website.

The Ethereum Foundation announced that the Glamsterdam network upgrade will activate on the Sepolia testnet at 21:53:36 Beijing Time on October 6, 2026 (epoch 353,024, slot 11,296,768). Activation timelines for the Hoodi testnet and Ethereum mainnet have not yet been finalized. The upgrade integrates the Amsterdam execution layer and Gloas consensus layer, with core improvements including: built-in proposer-builder separation (ePBS, EIP-7732), which formally incorporates block proposal and construction into the protocol; block-level access lists (BALs, EIP-7928), enabling parallel state reads and transaction verification; and adjusted gas pricing that more accurately reflects execution and state growth costs (EIP-8037, EIP-8038, etc.). Node operators must update their execution layer and consensus layer clients prior to activation. Corresponding client versions have been rolled out sequentially; detailed compatibility lists are available in the official announcement. Mainnet users and ETH holders do not need to perform any actions at this time.

According to TradingBeats’ monitoring, after Meta (META.US) announced the establishment of its enterprise AI business division and appointed former MongoDB CEO Chirantan “CJ” Desai to lead the unit, its stock closed down roughly 4.79% on the day, ending at $715.62. The enterprise AI platform Meta is launching will integrate products including the Muse AI assistant, Meta Business Agent, Muse API, and Muse Code, aiming to further expand the capabilities of its Llama model to enterprise clients. Desai, who previously served as MongoDB’s CEO and held roles at enterprise tech firms like ServiceNow and Cloudflare, will now drive Meta’s enterprise AI business development. The market has yet to embrace the move. The Information analyst Martin Peers noted that while Meta boasts a large base of enterprise advertising clients that could help promote its AI products, the enterprise AI market is already highly crowded, with competitors including Microsoft, Google, OpenAI, Anthropic, and traditional enterprise software firms all vying for corporate customers. Additionally, Meta’s past forays into enterprise software have not gone smoothly—its workplace collaboration tool Workplace was ultimately shut down. Entering the market via Muse and its new enterprise AI platform means Meta must prove it can extend its advertising ecosystem advantage to the enterprise software business. The market remains cautiously watching whether Meta can translate its AI capabilities into a new source of enterprise revenue.

Beating AI News Flash: SpaceXAI has launched Team Bots for its Grok Bot, transforming the previously personal AI Agent into a shared "AI colleague" for entire teams. Teams can configure profiles, workflows, plugins, and API credentials for the Bot, then share it directly with all members. The feature is currently in public beta for Teams and Enterprise plans. A single Team Bot holds the team’s unified business knowledge, while private chats between individual members and the Bot remain isolated. Team members can assign tasks to it separately, or add it to Slack channels for everyone to ask questions, add context, and view results. It also integrates with tools including Salesforce, Notion, and GitHub, and retains team knowledge learned from work. SpaceXAI is already using Team Bots internally across sales, engineering, marketing, and data analytics. For instance, the engineering Team Bot pulls project progress from Slack, Notion, and Linear, auto-creates tickets, and calls Cursor Cloud Agents to resolve issues. The company noted that during Team Bots’ development, a 5-person team used it to coordinate hundreds of Cloud Agents, submitting over 100 Pull Requests (PRs) daily. Previously, Grok functioned more as a personal long-term Agent capable of operating cloud computers and working continuously. Team Bots fill the team-sharing gap: members no longer need to re-train the AI individually, as the same set of workflows, tools, and business knowledge can be stored long-term in one Bot.

Beating AI Briefing: The Databricks team announced that they integrated GPT-6 Astra and Opus 5 into an automated optimization loop, enabling the models to continuously rewrite, test, and refine GPU kernels, ultimately claiming first place across all four tracks of the NVIDIA SOL-ExecBench benchmark. The benchmark includes 235 GPU kernel tasks covering basic operators, complex fused operators, low-precision computing, and real-world large language model inference. The team stated that the total token cost for the entire process was approximately $70,000. Built on KDA and Humanize, the system allows AI to autonomously write kernels, run correctness and performance tests, and iteratively modify them based on results. The KDA team recently applied a similar method to Kimi Delta Attention, achieving 2.96x the speed of Moonshot AI’s official FlashKDA on the NVIDIA B300. Both results are different applications of the same agent kernel optimization system. The Kimi project proved that a highly complex attention kernel can be rewritten by AI, while Databricks’ work extended this approach to hundreds of distinct kernels.