Point of View: The Federal Reserve Has Over-Eased, with a 65% Chance of Stagflation in the United States

Beating AI Insight News Brief: Large model inference and deployment framework SGLang has added native decision-making APIs, enabling off-the-shelf LLMs and multimodal models like Qwen3.8-27B to directly perform selection, judgment, and scoring without modifying weights or undergoing retraining. When developers input the current context and several candidate answers, the model directly returns the probability of each option, eliminating the need to first generate a text segment for subsequent parsing. This functionality leverages the "next-token probability" that large language models inherently compute. For example, if candidate answers are A, B, and C, a standard chat model would continue generating text, while SGLang directly reads the model’s probabilities for these options to determine its preferred choice. The model itself remains unchanged; only the method of reading outputs is adjusted. SGLang has demonstrated this capability with a *Pokémon FireRed* demo using Qwen3.8-27B. The model decides whether to attack, switch Pokémon, or heal based on real-time game state, with each decision taking less than 100ms, and successfully cleared the Elite Four and Champion in one run. Since Qwen3.8-27B natively supports image input, this approach can also handle multimodal decision-making. Additionally, SGLang has added a new /v1/systemone endpoint, compatible with the TypeSafe SDK used by Jev. Applications already using this SDK can switch their service address to their own SGLang instance to continue making calls. The new feature is now available in the nightly build and scheduled for official release alongside version v0.5.21.

Needham reaffirms SpaceX’s Buy rating and maintains a $250 price target, with rapid growth in AI computing revenue as the core driver. The firm’s report notes that SpaceX’s announced computing transactions are projected to generate an annualized revenue run rate of $54 billion, strongly supporting management’s goal of hitting $100 billion in total annual recurring revenue (ARR) by the end of 2026. Separately, Starship’s 14th flight successfully reached orbit and deployed 26 Starlink V3 satellites, underscoring SpaceX’s sustained technological edge in rocket reusability and satellite constellation expansion.

According to an official announcement, Bitget’s User Protection Fund has been replenished to over $300 million, with its latest on-chain balance standing at 3,705 BTC. Established in 2022, the fund has operated as an additional safeguard for user assets, adopting an on-chain, transparent and traceable management framework. Users can access the relevant wallet addresses via the fund’s page to independently verify the on-chain fund.

According to an official announcement, Concrete (CT) has been listed on Coinbase. Users can trade and transfer the token on the platform.

On-chain detective ZachXBT stated in a post that suspected North Korea-linked attackers in the Bitget hack incident have begun transferring around 2,700 ZEC (valued at approximately $3.8 million) into Zcash’s Ironwood shielded pool. Bitget’s hot wallet was previously hacked for a total of ~18,900 ZEC, worth roughly $28.3 million.

12:30 p.m. ET yesterday, Donald Trump hosted a lunch at the White House with executives from multiple artificial intelligence firms. Attendees included Jensen Huang, Elon Musk, Mark Zuckerberg, Alphabet CEO Sundar Pichai, and Anthropic co-founder and CEO Dario Amodei, among others. After the lunch, Trump and a group of U.S. AI giants took impromptu questions from reporters. When asked about "AI displacing jobs", Trump called on Musk to respond. Musk blurted out "AI" out of habit before quickly correcting to "SI"—a detail that caught Trump’s attention, prompting the president to thank him repeatedly. Musk then stated that robots and superintelligence can bring significant improvements to humanity in fields like healthcare, and while future job markets and work tasks will undergo major changes due to AI, "the overall demand for humans remains very high". Anthropic co-founder Dario Amodei revisited his earlier call for "slowing AI development"—a stance that conflicts with Trump’s views—saying AI has enormous benefits but carries real technological risks, and how to address them is still under discussion. He again emphasized: "We need to work together to ensure we win safely". During the interview, the broader tech community generally viewed Amodei as overly nervous compared to Musk’s calm response. When introducing Amodei to reporters, Trump remarked, "Dario is great, whatever he says is okay". Amodei responded with an exaggerated, forced laugh, then awkwardly dropped his smile, bowed his head and said nothing, creating an awkward moment.