Huang Wenhao, head of ByteDance’s Seed Model Evaluation, has reportedly resigned. He previously led the training of the Yi model, and is rumored to be starting a startup next.

IoTeX, a DePIN and AI open ecosystem platform, announced that its Zanzibar hard fork has launched on the mainnet. Under the IIP-59 proposal, voter staking rewards are now directly calculated and distributed on-chain by the protocol for each reward cycle, eliminating the need for off-chain services or intermediaries, with every payment verifiable directly on-chain by anyone. The first on-chain distribution will take place tomorrow. Zanzibar also strengthened its consensus mechanism through delegated key proof-of-possession checks, in preparation for future BLS signature aggregation.

According to monitoring by ai_9684xtpa, three addresses suspected to belong to the same entity or a crypto whale redeemed 18.085 million ENA tokens from Ethena and deposited them into Binance over the past hour, with a total value of roughly $3.64 million and an average deposit price of $0.2013.

Tron founder Justin Sun stated in an offline discussion with Jin Ming that market concerns over the U.S. debt credit crisis may be excessive. He argued that the core issue with U.S. Treasuries is not solvency, but whether the U.S. can maintain its global technological and military dominance to sustain global investors’ ongoing demand for U.S. debt. Sun noted that the U.S. can repay old debt by issuing new debt, with the key being whether newly issued U.S. Treasuries can continue to attract market subscriptions; the U.S.’s global tech and military influence is a critical foundation for investors to hold U.S. debt. As AI emerges as a core area of future development, dominance in chips, storage industry chains and AI technology will grow increasingly important. Regarding whether the continuous rise in U.S. Treasury yields reflects weakening market confidence, Justin Sun said this is also tied to asset allocation competition between AI and U.S. debt: when investors favor equity assets like AI, risk-free assets need higher yields to draw capital. He added that as long as the U.S. retains its dominance in the AI sector, U.S. Treasuries will not lack buyers. Additional quick news roundups: "Justin Sun’s Latest Personal Asset Allocation: BTC Core Holdings Unchanged, ETH Reduced by 80-90% Shifted to U.S. Stocks", "Justin Sun: Quantum Computing Risks Underestimated by the Market, Q-Day May Arrive in 1 to 1.5 Years", "Justin Sun on 'Sun Theory': Essentially Game Theory and Evolutionary Theory, Helping Ordinary Young People Achieve Zero-to-One Growth"

Tron founder Justin Sun recently stated in a conversation with Jin Ming that the "Sun Theory" is not mere sophisticated egoism, but a methodology that combines the era’s context and personal circumstances to select the path of maximizing benefits. Essentially a framework integrating game theory and evolutionary theory, it aims to help ordinary young people achieve progress from scratch. Sun noted that the "Sun Theory" is characterized by low learning cost, accessible to most young people, though those with superior personal conditions who have already developed independent insights in their professional fields do not necessarily need to fully adopt it. He cited his 2016 investment advice as an example: he then advised young people against getting married or buying a house, and to invest in NVIDIA and Tesla, arguing that adapting to era changes can drive significant wealth growth. When asked how to prevent the "Sun Theory" from overemphasizing personal interests and fueling social egoism, Justin Sun explained that its core is not simply pursuing personal gain, but first understanding the era’s context, then choosing a development path suited to one’s own situation. Additionally, Sun said he uses AI to monitor his physical data, hoping to extend his healthy lifespan as much as possible within a reasonable overall lifespan, and face death with dignity and decency. When discussing the meaning of life, he stated that life is a collection of experiences; he will engage with things he is interested in as much as possible, but cannot control how others evaluate him.

Insight: Beating AI News Flash, developer Mihir Gupta spent an afternoon using Claude Code to revamp his laggy TCL smart TV. The device only has 2GB of RAM; the AI helped him disable 25 pre-installed apps and background services, and replace the ad-ridden native home screen with a custom interface he built. The new home screen’s RAM usage dropped from 192MB to 83MB, a roughly 57% reduction. The interface mimics Apple TV’s design, retains shortcuts for frequently used apps, HDMI input, and settings, and automatically switches wallpapers between 12 nature photos. The entire installation package is just 4.7MB, with no requirement for root access or firmware flashing. After enabling the TV’s developer mode, Claude Code connects to the device via Wi-Fi and ADB, checks pre-installed software, backs up configurations, develops the new home screen, and verifies results via screenshots. Gupta has open-sourced all related code, prompts, and recovery scripts. This is not the first AI-powered TV modification: In early September, engineer Mert Cobanov used Claude Code to clean a 4-year-old Android TV. On October 17, developer Gathin open-sourced a similar project named tv-vibehack, and Gupta’s work was inspired by this project.

TRON founder Justin Sun recently discussed Hyperliquid during a 3.5-hour offline conversation with Jin Ming, calling it one of the biggest new breakthroughs to emerge in this cycle. The U.S. stock perpetual contracts it launched have opened up new horizons for the industry. However, as more projects enter the U.S. stock perpetual contract space, Hyperliquid is facing intensifying competition. Sun also noted that the future competition between decentralized exchanges (DEXs) and centralized exchanges (CEXs) hinges on how U.S. regulators define Hyperliquid’s regulatory boundaries. If Hyperliquid can secure regulatory exemptions similar to Uniswap, a large number of CEXs could shift to the DEX model to cut compliance costs and gain broader global operational room, which would likely squeeze CEXs that fail to transition. Conversely, tighter regulatory requirements would also put significant pressure on Hyperliquid. Hyperliquid’s maturity in regulatory affairs and U.S. domestic operations still lags behind onshore compliant firms like Circle. Thus, the long-term competitive landscape between DEXs and CEXs largely depends on regulators’ stance toward new trading models and the compliance boundaries they establish.