Jump Capital has raised a $350 million fund for AI investments.

According to GMGN data, the on-chain Meme coin PIPEDOG (Pipedog) on Robinhood’s blockchain hit a market cap of over $74.5 million and a trading volume of $34.8 million within four hours of its launch. BlockBeats reminds users that most Meme coins lack practical use cases, exhibit high price volatility, and advises investors to exercise caution.

According to Blockaid's monitoring, the Pro token under Crypto DAO was hacked. As of 00:23 this morning, the attacker and associated profit-linked addresses currently hold a total of approximately 8.2 million USDT.

Kim Yong-beom, Director of the Policy Office at the South Korean President’s Office, provided a systematic response to the causes of the recent sharp plunge in the KOSPI index. He noted that the market is currently undergoing internal debate over AI investments: on one hand, concerns over the commercial sustainability of large-scale capital expenditures by major tech firms, with "semiconductor supply continuing to rise, while there is a lack of confidence that future demand can sustain this growth"; on the other, accelerating catch-up by China’s semiconductor industry has brought new competitive pressure. News such as the listing of ChangXin Memory Technologies and breakthroughs in China’s photolithography equipment manufacturing technology has led the market to re-examine whether South Korea’s competitive edge in memory chips remains solid. However, Kim clearly emphasized that AI demand is not temporary, stating "demand will remain steady and sustained." He argued that South Korea should seize this opportunity to accelerate factory construction, make timely investments, and strengthen R&D spending. Addressing the causes of amplified volatility in South Korea’s domestic market, Kim clearly stated that leveraged ETFs are not the sole driver. Factors including South Korea’s market structure where semiconductor and AI-related stocks account for 40% to 50%, high weights of derivatives and ETFs, and investor composition have combined to make volatility significantly higher than in overseas markets. While leveraged ETFs may cause temporary market impacts during end-of-day rebalancing, volatility sometimes emerges as early as the morning. This temporal distribution alone rules out leveraged ETFs as the sole driver. Kim recommended that South Korea’s Financial Services Commission conduct a comprehensive review of the overall capital market structure, including leveraged ETFs.

Per Bitget market data, South Korean stocks opened higher then moved lower, with the KOSPI index accelerating its slump to a 1% daily drop after earlier rising over 3%. SK Hynix’s decline widened to 4.5%, while Samsung Electronics’ gain narrowed to under 1%.

Uniswap founder Hayden Adams has responded to the controversy over Uniswap v4’s fee switch, clarifying that the claim “liquidity provider (LP) fees are being reduced” is a misunderstanding. Protocol fees are additive rather than deducted, and LPs still earn 30 basis points per trade. Addressing the assertion that “the protocol takes 25% of LP profits,” he explained that for pools with a 30 basis point fee, the protocol fee is 5 basis points—accounting for roughly 14% of total transaction fees—while LPs’ own earned fees remain unchanged. Adams also noted that centralized exchanges (CEXs) charge 100 to 200 basis points per trade, making Uniswap’s 5 basis point fee at the 30 basis point tier 20 to 40 times cheaper.

Coursera (NYSE: COUR) announced a $100 million strategic equity investment in LearnVector, the new AI-native learning startup founded by Coursera co-founder Andrew Ng. The investment grants Coursera an approximately one-third stake in LearnVector, valuing the startup at around $300 million. Founded by Ng, who serves as CEO, LearnVector is an AI education startup targeting white-collar workers, with a core goal of leveraging agent AI to deliver one-on-one personalized learning experiences, shifting from traditional "one-to-many" instruction to "one-on-one" dedicated tutoring. The AI agents will map personalized learning paths for each learner, adapt to their learning styles, and provide ongoing support until they fully master relevant skills. The first batch of courses is expected to launch in early 2027, and the deal has been unanimously approved by Coursera’s special committee of independent directors.