Arbitrum clarifies: Robinhood Chain did not suffer an outage yesterday, only experiencing batch submission delays.

Beating AI News Flash: According to Reuters, citing people familiar with the matter, AI firm Anthropic’s IPO plans are being adjusted to mid-October. Roadshows could kick off as early as mid-October, with the listing targeted to complete before the U.S. midterm elections in November. The company had originally planned to publicly release its IPO prospectus as early as next week, but that timeline is now pushed back to the end of September. This timing shift means Anthropic could aim for a listed valuation of up to roughly $2 trillion. If achieved, the IPO would rank among the largest in history and serve as a key case for capital markets to test the AI industry’s growth potential. Sources also revealed that Anthropic is finalizing a $15 billion revolving credit facility, with participating bank analysts expected to meet with the company ahead of the prospectus filing. Morgan Stanley, Goldman Sachs, JPMorgan Chase, and Citigroup are among the institutions involved in preparing for the IPO. The offering is expected to draw significant market attention and could proceed around the same time as potential listing plans of other AI firms like OpenAI.

Tesla’s stock price plunged 5.92% in Friday’s close, wiping roughly $88 billion (about 591 billion yuan) off its market capitalization in a single day. While Tesla has positioned its Robotaxi as the next growth engine, the commercialization of its Cybercab is facing a “double whammy” of regulatory scrutiny and market expectations. The U.S. National Highway Traffic Safety Administration (NHTSA) has launched an investigation into around 1,000 Cybercab units, focusing on the compliance procedures and technical data Tesla used when it determined certain federal motor vehicle safety standards did not apply. Since the Cybercab lacks a steering wheel, accelerator, brake pedal and traditional side mirrors, the probe could impact its subsequent deployment timeline and expansion of operating regions. Meanwhile, Tesla’s Cybercab commercialization launch event held in Austin was deemed “below expectations” by Wall Street: it lacked specific deployment scale, production targets and timelines, and Elon Musk himself did not attend. Analysts from Evercore ISI, Wells Fargo, Barclays and JPMorgan all noted the event lacked sufficient new details, which could put pressure on Tesla’s stock in the short term. Still, Wall Street has not completely shifted its stance on Cybercab’s long-term potential. JPMorgan forecasts Tesla’s Robotaxi fleet could expand to around 9,000 units by the end of 2027, while the Royal Bank of Canada (RBC) projects the number of Cybercabs in the U.S. market will reach roughly 40,000 by 2030.

According to Lookonchain monitoring, after Shivon Zilis reposted a post containing the SLINK contract address and Elon Musk subsequently replied, SLINK's market cap surged to $80 million at one point. After Shivon deleted the relevant post, SLINK's price plummeted more than 80%. Analysts noted that in this market movement, over 30 suspected insider wallets bought SLINK in advance and sold quickly, reaping a total profit of over $4.7 million, with the highest return exceeding 1,200 times. Meanwhile, numerous traders who followed the trend to buy suffered losses.

According to Lookonchain’s monitoring, a trader executed two consecutive FOMO trades within a single day, logging a total loss of roughly $907,000. The trader first spent $916,500 to purchase 7.28 million MEME tokens at $0.126 apiece, then sold the entire holding for just $353,000, taking a $563,000 loss equivalent to a 61.45% drop. Next, the trader deployed an additional $356,000 to buy 8.93 million SLINK tokens at $0.04 each, only to sell them for approximately $12,000, incurring a $344,000 loss with a 96.62% plunge. BlockBeats reminds users: Most meme coins lack real-world use cases and are subject to extreme price volatility. Please protect your assets and refrain from FOMO trading.

Beating AI Express News: Xiaomi officially launched Xiaomi-TabLDM today, a general tabular data foundational large model. According to introductions, Xiaomi-TabLDM adopts a single pre-trained model with unified default configurations, which can directly adapt to different tabular datasets. It enables classification and regression predictions without requiring retraining, parameter tuning, or post-integration for each individual task. Xiaomi noted that the model is designed to lower the threshold for tabular data modeling, providing general foundational capabilities for structured data analysis and prediction across various scenarios.

Insight Beating AI Flash News: OpenAI has fully rolled out GPT-6 Astra to all Plus, Pro, and Business users. Yesterday, Astra was still in a phased rollout, leaving many Pro users unable to access it; today, both Plus and Pro users can use the model in ChatGPT Work and Codex. Core product lead Thibault Sottiaux noted that the team’s system scaling capability is stronger than expected, enabling them to complete this full rollout quickly. When Astra first launched, it was only available to a small number of enterprise customers. OpenAI originally planned a gradual rollout over several days, but after many paid users waited long for access, Sam Altman apologized for the "messy launch". Now the team says their system scaling is better than anticipated, so they fully opened access to Plus users early, with promised compensations delivered as scheduled. Eligible existing Plus, Pro, and Business users will receive one banked reset each on September 3 and 4. This is equivalent to a Codex full reset voucher that can be used manually later, refreshing both 5-hour and weekly limits. New accounts or plan upgrades made before 11:00 AM Beijing Time on September 5 will also qualify for this reset. While Plus users can access Astra, its limits are not generous. OpenAI currently estimates that Plus users can send roughly 5–45 Astra local messages every 5 hours, compared to 10–100 for GPT-5.6 Sol. Actual consumption depends on task length, inference intensity, and tool calls.