Industry News

Tracks company news, strategic changes, funding activities, and personnel adjustments across the blockchain and crypto industries, delivering a full-spectrum industry overview for our users.

Tens of Millions of Errors Per Hour: Investigation Reveals the 'Accuracy Illusion' of Google AI Search

A New York Times investigation, in collaboration with AI startup Oumi, reveals significant accuracy and reliability issues with Google's AI Overviews search feature. Testing over 4,300 queries showed the accuracy rate improved from 85% (Gemini 2) to 91% (Gemini 3). However, given Google's scale of ~5 trillion annual searches, this 9% error rate translates to over 57 million incorrect answers generated hourly. A more critical issue is the prevalence of unsubstantiated citations. For correct answers, the rate of "unfounded citations"—where provided source links do not support the AI's claims—worsened, rising from 37% with Gemini 2 to 56% with Gemini 3. This makes it difficult for users to verify the information. The AI also heavily relies on low-quality sources, with Facebook and Reddit being its second and fourth most cited domains. Furthermore, the system is highly susceptible to manipulation. A BBC journalist successfully "poisoned" it by publishing a fake article; Google's AI began presenting the false information as fact within 24 hours. Google disputed the study's methodology, criticizing the use of the SimpleQA benchmark and an AI model (Oumi's HallOumi) to evaluate its own AI. The company maintains that its internal safeguards and ranking systems improve accuracy beyond the base model's performance.

marsbit04/10 12:30

Tens of Millions of Errors Per Hour: Investigation Reveals the 'Accuracy Illusion' of Google AI Search

marsbit04/10 12:30

TechFlow Intelligence Bureau: Anthropic Mythos Triggers Financial Regulatory Alarms, WLFI Unlocks 27 Billion Tokens with 0 Dividends

TechFlow Intelligence Report: Anthropic's Mythos AI model has triggered urgent meetings between the U.S. Treasury Secretary, Fed Chair, and major bank CEOs due to its ability to autonomously find software vulnerabilities and launch attacks, elevating AI safety to a systemic financial risk. In crypto, Bitcoin surpassed $70k amid significant ETF inflows, while the World Liberty Fi (WLFI) project unlocked 27 billion tokens with a controversial profit model directing 100% of protocol profits to the Trump family, offering stakers only 2% inflationary rewards. Other key developments include Microsoft facing criticism over Copilot's reliability, Google's AI search summaries being caught providing false information, and a major AI supply chain attack affecting projects like Apifox and LiteLLM. A solo Bitcoin miner with minimal computing power defied odds to mine a full block. Regulatory shifts are accelerating with Japan classifying crypto as financial instruments and the SEC frameworks taking shape. In macro, a whale shorted $74M in oil, betting against geopolitical tensions, while Moody's warned of a potential $680B AI investment bubble among tech giants. The overarching theme is the convergence of technology, finance, and geopolitics: AI breakthroughs are now systemic threats, geopolitical events are being priced on-chain in real-time, and supply chain vulnerabilities reveal the fragility of our interconnected digital ecosystem.

marsbit04/10 12:19

TechFlow Intelligence Bureau: Anthropic Mythos Triggers Financial Regulatory Alarms, WLFI Unlocks 27 Billion Tokens with 0 Dividends

marsbit04/10 12:19

Aave Mired in a Crisis of Confidence: Service Providers Exit En Masse, Failures in Technology, Governance, and Risk Control

Aave, a leading DeFi lending protocol, is facing a severe internal crisis marked by the departure of key service providers, raising concerns about its governance, security, and future direction. The crisis began when Chaos Labs, the protocol's long-time risk management provider, terminated its relationship with Aave. The firm cited financial losses, the exit of other major contributors, and fundamental disagreements over the risk architecture of the upcoming Aave V4. Aave Labs declined Chaos Labs' demands for a significant fee increase and exclusive control over key functions like risk management and oracle services. This exit followed the departure of two other critical partners. BGD Labs, the primary technical contributor to Aave V3, accused Aave Labs of forcing an aggressive transition to V4 by limiting V3 development and devaluing its work. Subsequently, the Aave Chan Initiative (ACI), a major governance service provider, announced its planned exit, criticizing Aave Labs for centralizing power and controlling a large portion of voting tokens. The conflict highlights a central paradox within DAOs: the tension between founder-led vision and decentralized governance, and between long-term protocol health and short-term capital interests. Aave Labs is pushing for a more integrated and efficient "Aave Will Win" model with V4, arguing it is necessary for competing at an institutional level. However, critics warn this centralization comes at the cost of the protocol's decentralized credibility and increases systemic risk. The immediate impacts include a potential security downgrade, a loss of institutional knowledge, and damaged community trust. While Aave Labs views this as a painful but necessary transition, the market is watching cautiously as the protocol navigates this period of significant internal turmoil.

marsbit04/10 10:14

Aave Mired in a Crisis of Confidence: Service Providers Exit En Masse, Failures in Technology, Governance, and Risk Control

marsbit04/10 10:14

RWA Weekly: HSBC and Standard Chartered Secure Hong Kong Stablecoin Licenses; US FDIC Releases Draft Guidelines for Institutional Stablecoin Issuance

RWA Weekly: HSBC and Standard Chartered Secure Hong Kong Stablecoin Licenses; US FDIC Issues Draft Guidelines for Institutional Stablecoin Issuance This week’s RWA sector saw significant growth, with the on-chain total market cap rising to $29.06 billion. Stablecoin market capitalization remained high at $300.65 billion, while monthly transfer volume hit a record $10.21 trillion. Active addresses surged 15.24%, indicating strong retail participation recovery. Regulatory milestones were achieved as Hong Kong granted its first stablecoin licenses to HSBC and Standard Chartered, marking the start of a compliant stablecoin era. The U.S. FDIC released draft guidelines for stablecoin issuance, focusing on reserve management, redemptions, and capital requirements. The U.S. Treasury also proposed rules requiring stablecoin issuers to implement anti-money laundering and sanctions compliance systems. South Korea, Dubai, and Russia advanced their stablecoin and RWA regulatory frameworks. Key project developments include six Swiss banks, including UBS, planning to test a digital Swiss franc in 2026. Securitize began tokenizing shares for Nasdaq-listed Currenc, enabling 24/7 trading. SBI Ripple Asia completed development of a token issuance platform on XRP Ledger. Circle launched CPN Managed Payments to expand stablecoin payment services for institutions. Funding highlights: Pharos raised $44 million in Series A funding to develop its RWA-focused blockchain. GSR led an investment in tokenization platform Libeara. Gobi Partners invested in Transak to expand compliant stablecoin and digital asset payment infrastructure in Asia. S&P Global reported that banks remain cautious about stablecoins, with only 7% of small and mid-sized U.S. banks developing related frameworks. Chainalysis projected stablecoin transaction volume could reach $1,500 trillion by 2035, driven by generational wealth transfer and deeper integration into payment systems. Major tech firms like Meta are increasingly adopting stablecoins as a core payment strategy, signaling a shift toward digital asset-based transaction infrastructures.

marsbit04/10 09:48

RWA Weekly: HSBC and Standard Chartered Secure Hong Kong Stablecoin Licenses; US FDIC Releases Draft Guidelines for Institutional Stablecoin Issuance

marsbit04/10 09:48

Domestic AI Booms: Zhipu's Market Cap Surpasses 430 Billion HKD, Mysterious Model Tops Text-to-Video Ranking

China's AI sector is experiencing a significant surge, with Zhipu AI's market capitalization exceeding HK$430 billion and a new model, HappyHorse-1.0, topping the text-to-video generation rankings. On April 9, Hong Kong and A-share AI stocks rallied strongly. Zhipu's shares rose 8.74%, and Xunce Technology surged over 24%. The A-share market saw similar gains, with the China Merchants AI ETF rising over 10%. The rally was fueled by two major catalysts. First, the anonymous model HappyHorse-1.0 topped the Artificial Analysis Video Arena leaderboard, surpassing ByteDance's Seedance 2.0. It generates synchronized video and audio from text in about 38 seconds. Second, Zhipu released its flagship model, GLM-5.1, which can autonomously perform complex software engineering tasks for 8 hours without human intervention. Notably, it was trained entirely on Huawei's Ascend 910B processors, a milestone for China's AI self-sufficiency. Industry experts note the rapid iteration of AI models, with new breakthroughs frequently appearing. While some market hype, the technical capabilities of these models are noteworthy. Zhipu also increased its API prices by 10%, signaling a shift from a growth-at-all-costs model to a focus on sustainable profitability and value creation. The industry is moving from a "technology race" to a "value co-creation" phase, entering an early stage of "order fulfillment and profit release." Paid services for top-tier models are in high demand, indicating the market is moving past the free user acquisition phase.

marsbit04/10 06:25

Domestic AI Booms: Zhipu's Market Cap Surpasses 430 Billion HKD, Mysterious Model Tops Text-to-Video Ranking

marsbit04/10 06:25

MakerDAO Transforms and Receives B- Credit Rating, Same Level as Congo Government Bonds

MakerDAO, the decentralized finance lending protocol, and its subDAO Spark are undertaking measures to improve their credit rating in a bid to attract institutional capital. In August, S&P Global Ratings assigned a B- rating to MakerDAO’s USDS and DAI stablecoins, placing them in the "junk bond" category, equivalent to the government bonds of the Democratic Republic of Congo. To enhance its standing, the protocol plans to repackage debt backing USDS, isolate exposure to high-risk yield sources, and introduce junior capital vaults. These structural changes aim to align with traditional finance expectations and mitigate concerns over riskier assets like Ethena’s USDe, which previously received a high-risk weighting from S&P. Spark intends to launch first-loss capital vaults in Q2, offering higher risk but greater returns, similar to junior tranches in traditional structured finance. The move reflects a broader trend among DeFi protocols seeking credibility through ratings from established agencies like S&P, Moody’s, and Fitch, as well as emerging crypto-native firms like Credora. Improving credit ratings is increasingly critical as DeFi aims to capture capital from traditional financial institutions entering the blockchain space.

marsbit04/10 05:30

MakerDAO Transforms and Receives B- Credit Rating, Same Level as Congo Government Bonds

marsbit04/10 05:30

Treasuries and yield funds – Is this the next phase of tokenization?

Tokenized finance is experiencing significant growth, with BNB Chain's tokenized asset market cap reaching a record $16.6 billion, more than doubling year-over-year. The broader tokenized funds market has expanded to $31.9 billion, driven primarily by yield-bearing products rather than simple stablecoins. While Tether and USDC remain the largest issuers, faster growth is occurring among smaller players like Ondo Finance, Maple Finance, and Centrifuge. The trend highlights a shift beyond cash proxies toward income-generating on-chain assets, signaling a new phase in asset tokenization.

ambcrypto04/10 05:05

Treasuries and yield funds – Is this the next phase of tokenization?

ambcrypto04/10 05:05

Why is Covenant AI Fleeing Bittensor?

Covenant AI announced its exit from the Bittensor network, accusing co-founder Jacob Steeves (Const) of centralized control and governance abuses. The team claimed Steeves unilaterally suspended subnet emissions, revoked community management permissions, deprecated infrastructure without due process, and exerted economic pressure through token sales. Covenant AI highlighted its achievement in developing Covenant-72B—a 72-billion-parameter language model trained collaboratively by over 70 contributors on consumer hardware—as proof that decentralized AI training is viable. However, they argued that Bittensor’s governance is effectively controlled by a small group, with Steeves retaining ultimate authority, contradicting the network’s promised decentralized and permissionless nature. The incident sparked community debate, with some critics pointing to Covenant AI founder Sam Dare’s prior sale of 37,000 $TAO tokens. Meanwhile, $TAO’s price dipped amid the controversy. Covenant AI reaffirmed its commitment to decentralized AI and will continue its mission independently, emphasizing that true decentralization cannot coexist with centralized control.

marsbit04/10 03:44

Why is Covenant AI Fleeing Bittensor?

marsbit04/10 03:44

The TAO Subnet Team Praised by Jensen Huang Has Parted Ways with the Founder Amidst a Fallout

Nvidia CEO Jensen Huang recently praised the decentralized AI project Bittensor (TAO) during a podcast, specifically highlighting a 72-billion-parameter Llama model trained collaboratively by a subnet team called Covenant AI. This endorsement initially boosted TAO's price, but the situation deteriorated rapidly when Covenant AI's founder, Sam Dare, publicly announced the team's departure from the Bittensor network. Covenant AI accused Bittensor and its key figure, Jacob Steeves (known as Const), of centralization and abuse of power, contradicting Bittensor’s decentralized ethos. The team claimed that Const exercised unilateral control by halting subnet emissions, removing administrative rights, discarding infrastructure, and using token sales to pressure the team. They argued that Bittensor’s governance is effectively centralized under Const, despite claims of distributed control. As a result, Covenant AI decided to leave, intending to continue its work on decentralized AI training elsewhere. The exit has sparked significant concern within the Bittensor community, raising doubts about the network’s decentralization narrative, technical future, and token value. TAO’s price fell sharply following the news. Const responded vaguely on social media, suggesting the event would push Bittensor toward more decentralized, “headless” subnets, but has not addressed the specific allegations in detail. The incident has damaged Bittensor’s reputation while raising Covenant AI’s profile.

Odaily星球日报04/10 03:08

The TAO Subnet Team Praised by Jensen Huang Has Parted Ways with the Founder Amidst a Fallout

Odaily星球日报04/10 03:08

5.4 Billion Burned, Sora Dies: Anonymous Chinese Model Kicks Open the Next Door in 38 Seconds

In March-April 2026, two major events reshaped the AI video generation landscape. OpenAI shut down its flagship model Sora, citing unsustainable daily costs of $15 million and low user retention, effectively exiting the consumer video market. Shortly after, an anonymous Chinese model dubbed "HappyHorse-1.0" topped the blind-test leaderboard on Artificial Analysis with a score of 1357 in text-to-video (without audio), outperforming rivals like ByteDance’s Seedance 2.0. HappyHorse-1.4 seconds to generate 1080p video with audio on a single H100 GPU. Its unified Transformer architecture and distilled diffusion techniques significantly improved efficiency compared to Sora’s costly diffusion-based approach. The model is speculated to be developed by Alibaba or based on Sand.ai’s technology, though its anonymous release suggests strategic data collection and legal risk avoidance regarding copyright and deepfake regulations. Meanwhile, commercial leaders like ByteDance impose high barriers—including million-dollar API contracts and strict compliance checks—to mitigate legal risks, focusing on B2B applications rather than consumer use. Key emerging opportunities include automated e-commerce promo videos, AI-assisted short drama production, and localized ad creation for global markets, all driven by plunging generation costs and faster turnaround times. The competition has shifted from pure model performance to cost efficiency, workflow integration, and regulatory compliance.

marsbit04/10 00:19

5.4 Billion Burned, Sora Dies: Anonymous Chinese Model Kicks Open the Next Door in 38 Seconds

marsbit04/10 00:19

活动图片

Популярные категорииright-arrow

Популярные тегиright-arrow

Industry News

Bitcoin