Federal Reserve Governor Jefferson: No Need to Hurry to Further Cut Interest Rates, Strong Economy Makes Caution Appropriate

U.S. sources told media on the 7th that U.S. Chairman of the Joint Chiefs of Staff Dan Caine is seeking ways to disengage from the Iran conflict. Three people familiar with the matter disclosed that over recent weeks, Caine has privately made clear to other senior Trump administration advisors that the U.S. needs a path out of the Iran conflict, as military options on the table could backfire, and air power alone is unlikely to fulfill Trump’s goals. Sources added that Caine has discussed military options for escalating the conflict with other government officials—including Vice President Vance, Secretary of State Rubio, and CIA Director Ratcliffe—while also proposing avenues to end the hostilities. Recently, Caine has also held private meetings with several like-minded Trump advisors to ensure they reach a consensus ahead of their meetings with the president. Additionally, during his recent talks with Trump, Caine voiced concern over the U.S.’s dwindling ammunition reserves, and the two sides also discussed potential options to escalate the conflict.

A developer recently open-sourced the kimi-k3-in-c project, attempting to run the 2.78-trillion-parameter Kimi K3 model on devices with only 8GB of RAM. The project is just 176KB in size, written purely in C99, and does not depend on GPU, CUDA, PyTorch, or BLAS—its model inference can be completed entirely via CPU. This solution leverages the Mixture of Experts (MoE) architecture features of Kimi K3. While the model’s total parameter scale reaches 2.78T, only 16 of the 896 experts per layer are activated. Instead of loading the full ~1.56TB model weights into memory, the developer stores most expert weights on NVMe drives and reads them in real time as needed for inference; some dense trunk layers also adopt layer-by-layer streaming loading. However, the solution still has obvious performance limitations: in 8GB RAM mode, the model takes approximately 32.7 seconds to generate one token, and requires nearly 1.7TB of high-speed storage support. The developer stated that this solution is currently more of an experimental exploration into optimization directions for large model inference infrastructure, and does not have practical production use value. Nevertheless, its approach of "NVMe streaming loading + MoE sparse activation" provides a new idea for the future low-cost operation of ultra-large models.

JPMorgan said the Q2 earnings season is drawing to a close, with roughly 80% to 85% of companies in the U.S. and Europe having reported results so far, and overall earnings performance has come in significantly stronger than expected. Data shows that both the U.S. and Europe have seen widespread EPS (earnings per share) beats, with the proportion of companies exceeding market expectations rising notably from prior periods. Despite already high market valuations and earnings expectations ahead of the earnings season, the combined EPS of S&P 500 constituents has continued to be revised upward, while the share of companies cutting their profit outlooks has fallen to its lowest level since 2021. By region, corporate earnings growth in the U.S. and Europe hit 25% and 23% year-over-year, respectively, both beating consensus market expectations; revenue growth rose 14% and 10% YoY, respectively, indicating robust top-line growth for companies. By sector, JPMorgan noted that the energy sector is one of the key drivers of earnings growth in both regions, fueled primarily by rising energy prices due to geopolitical conflicts. In addition, the financial and technology sectors also made significant contributions to overall profit growth.

X (formerly Twitter) has announced the launch of its "Original Content Reward Program", which aims to reward creators who provide original perspectives, professional analysis, news reporting, creative content, and comments. X stated that it will immediately stop accepting new applications for its Revenue Sharing Program. Existing Revenue Sharing users will continue to receive earnings until September 7, 2026, with their final three payments scheduled around August 14, August 28, and September 11. Starting from September 8, eligible existing creators can apply to join the new Original Content Reward Program. The new program distributes earnings based on "valid impressions" from original content—defined as views from X Premium users’ home feeds where posts are at least 50% visible. The first payment under the new program is expected to be issued on August 28. Eligibility requirements include: being at least 18 years old, having an account in good standing, subscribing to X Premium or Premium+, holding at least 500 verified followers, accumulating at least 500,000 impressions from verified users’ home feeds over the past 90 days, and consistently publishing original content. X emphasized that the new program prioritizes rewarding original writing, threads, reporting, analysis, videos, images, design works, and comments with unique value. Content such as simple reposts, scraped content from other platforms, secondary posts lacking substantial analysis, or content using automated tools to generate engagement will not qualify for rewards. X noted that it will focus future support on creators who bring unique perspectives, professional expertise, and creativity to the platform.

JPMorgan noted that despite the overall strong performance of the U.S. corporate earnings season, stock prices have responded relatively weakly to earnings beats, with even some tech companies seeing share declines after reporting strong financial results. The bank attributes this phenomenon to multiple factors, including the market having already priced in high expectations for corporate earnings, investors being overly concentrated in tech sector positions, and the market’s focus gradually shifting from short-term performance to companies’ future capital expenditure plans. Particularly in the tech sector, investors are reassessing whether companies’ large-scale investments in areas like AI infrastructure will deliver long-term returns, with earnings beats alone no longer sufficient to support further stock price gains.

The XRP Ledger (XRPL)’s latest software version 3.3.0 introduces multiple upgrade proposals, including the “Confidential Transfers” feature designed to deliver enhanced privacy protection for institutional users. The feature encrypts token balances and transfer amounts while retaining visibility into accounts and token types, and is primarily targeted at Multi-Purpose Tokens (MPT) on XRPL, with use cases covering tokenized financial assets such as funds and bonds. Using encryption technologies like zero-knowledge proofs, the network can verify transaction validity without disclosing specific amounts. Per CoinDesk, XRPL currently holds around $1.38 billion in on-chain real-world assets (RWA), including approximately $845.7 million in RLUSD. In addition to RLUSD, there are over $530 million in tokenized assets on XRPL, issued by entities including Ondo, VERT Capital, Archax, and Société Générale. Beyond Confidential Transfers, XRPL version 3.3.0 includes five other proposals: Batch, Sponsor, Permission Delegation, Dynamic MPT, etc., addressing institutional needs such as batch transactions, fee sponsorship, permission management, and dynamic adjustment of token attributes. However, these upgrades have not yet been officially launched; they require sustained support of over 80% from XRPL trusted validator nodes for two weeks before activation. The market is watching whether institutions like Aviva and Ondo, which have already issued assets on XRPL, will adopt this privacy feature.