Roam has launched an Enterprise eSIM, a customizable management dashboard to empower Web3 teams in cross-border collaboration.

Beating AI News Brief: Andrej Karpathy, who currently works on Anthropic’s pre-training team, has again addressed the public perception gap regarding AI capabilities. He argues this gap is widening, evolving from two distinct user groups talking past each other into an increasingly steep funnel. In his rough estimate, roughly 6 billion people (75% of the global population) have almost no exposure to large models. Around 1 billion to 2 billion people only occasionally use free chatbots, treating AI as a search or writing tool. Moving up the funnel, about 20 million people (0.2%) use cutting-edge models in professional fields like programming and mathematics. They have observed that complex projects that once took weeks or even months can now be handled by AI agents. At the very top are roughly 5,000 people (0.00006%), who have internal access to the most advanced AI systems. Karpathy outlines these frontier capabilities as including thousands of AI agents collaborating continuously for weeks, automatically discovering zero-day vulnerabilities, conducting cyber offense and defense at machine speed, and advancing mathematical research. He notes that human ability to review and understand these AI outputs has begun to lag behind. Karpathy attributes the gap to three factors: differing scales of problems users face; faster AI progress in fields like programming and mathematics where results are easily verifiable; and capability and access gaps between free, paid, and internal AI systems. The general public may still view AI as little more than an error-prone chatbot, while a small group of professionals have already started using it to handle large-scale projects.

The Sui Foundation has announced a $30 million investment in DATA Network as part of a strategic partnership. The AI training data protocol will integrate into the Sui technology stack and connect with on-chain data storage platform Walrus. Per the agreement, the Sui Foundation will become DATA Network’s largest institutional holder, joining existing stakeholders PIP Labs and DATA Foundation. DATA Network will upgrade to a protocol native to Sui, with its independent chain set to be retired. Existing DATA tokens will migrate to Sui at a 1:1 ratio. The original network will remain operational until the transition is finalized, and the foundation will publish a detailed timeline ahead of the upgrade. Walrus will act as the default storage layer for applications operated by the DATA Foundation. Separately, the DATA Foundation shifted its focus from general intellectual property infrastructure to AI training data this June. Its Trace tool, launched in June, tracks the source and usage rights of AI training data and facilitates payments to data contributors. Since its launch, the Trace registry has logged over 250 million contributions, and it currently processes more than 3 million receipts daily.

Beating AI News Flash: Reuters exclusively reports that executives of Moonshot AI are negotiating a settlement with their former company Recurrent AI and its investors, aiming to resolve a nearly two-year arbitration dispute ahead of its Hong Kong IPO. It remains unclear whether the two sides can reach an agreement. The dispute involves Moonshot AI founder Yang Zhilin and CTO Zhang Yutao, who co-founded Recurrent AI in 2016. According to a letter seen by Reuters, Recurrent AI and some of its investors accuse the pair of establishing Moonshot AI without board authorization and providing false information to Recurrent AI’s investors. The letter also alleges that servers holding key large language model data were improperly transferred, with some data deleted. The shareholding arrangement of Moonshot AI President Zhang Yutong has also been called into question. Reuters has not independently verified these allegations. Yang Zhilin has previously stated the startup obtained all necessary approvals, while his legal representative noted the case lacks factual and legal basis. At Moonshot AI’s founding, Recurrent AI was allocated a 9.5% initial stake. Now, Recurrent AI and some of its investors are seeking compensation equivalent to 10.5% of Moonshot AI’s initial share capital. Based on Moonshot AI’s $30 billion valuation from its June funding round, the potential compensation could exceed $700 million, according to people familiar with the matter. Moonshot AI has secretly submitted its Hong Kong IPO application. Sources close to the matter believe the dispute will not directly block the listing for now, but if left unresolved, issues related to equity ownership and disclosure could still act as obstacles.

Keep an eye on this new wallet (0x2f59), which made over $1M in a single day longing $STRK and is now going long on $W! 6 hours ago, it opened a 5x long on 28.88M $W ($643K). $W is already up 25% today. Can it go even higher?

According to Lookonchain monitoring, a mysterious crypto whale is selling Bitcoin to rebalance its holdings into Ethereum. The whale offloaded 410.14 Bitcoin worth $34.16 million, then bought 14,341 Ethereum on decentralized exchanges (DEX) for $36.14 million.

South Korea's Financial Services Commission (FSC) is pushing to expand the tokenized securities market beyond fragmented investments like real estate and artworks to include bonds and funds. Relevant amendments to subordinate laws under the Electronic Securities Act and Capital Markets Act have been open for public comment from the 2nd of this month to the 11th of next month, and are scheduled to take effect on February 4 next year to refine rules for issuance and circulation. The amendments explicitly bring traditional securities such as stocks, bonds, and funds into the tokenization scope. Initially, priority will be given to launching tokenized private money funds, private bonds, and trust-based unlisted stocks for institutional investors, while piloting public fragmented investment securities. Public bonds and funds targeting retail investors will only be expanded after the market stabilizes. Meanwhile, regulators plan to set an annual net purchase cap of 100 million won per platform for retail investors on over-the-counter (OTC) trading platforms, and allow non-financial institutions to directly manage investor accounts upon meeting requirements including 4 billion won in own capital.