Technology TrendsNews

Explores the latest innovations, protocol upgrades, cross-chain solutions, and security mechanisms in the blockchain space. It provides a developer-focused perspective to analyze emerging technological trends and potential breakthroughs.

Codex Goal Mode Usage Guide: How to Make AI Continuously Pursue a Specific Objective

"Codex Goal Mode: How to Make AI Work Continuously Toward a Specific Goal" OpenAI's Codex "goal mode" (/goal) transforms the AI from a reactive code assistant into a proactive execution agent capable of working autonomously for hours or even days to achieve a defined objective. To maximize its effectiveness, follow these key principles: 1. **Define Clear, Verifiable Exit Criteria:** The goal prompt should be a concise, measurable success condition, not a lengthy specification. Use quantifiable metrics like "reduce build time by 30%" or "achieve 100% test parity." 2. **Provide Initial Guidance and Tools:** Direct Codex toward likely problem areas and specify available tools (e.g., browsers, testing environments) to prevent it from exploring unproductive paths. 3. **Enable Progress Measurement:** Equip Codex with ways to track advancement, such as creating comparison tools for visual tasks or evaluation sets, ensuring it can gauge its own progress. 4. **Use a Realistic Execution Environment:** For tasks like performance optimization, provide access to environments that closely mimic production (e.g., similar configs, databases) to yield valid results. 5. **Be Cautious with Visual Goals:** Avoid vague "pixel-perfect" instructions. Instead, supplement visual references with functional checklists or design system specifications to prevent Codex from obsessing over minor details. 6. **Implement Progress Tracking:** For long-running tasks, have Codex commit code to draft PRs, update progress documents, or send Slack updates to maintain visibility into its work. 7. **Review and Consolidate Results:** Once the goal is met, instruct Codex to review its work, clean up ineffective experimental code, and reflect on what strategies succeeded or failed. Ultimately, using goal mode shifts the developer's role from writing prompts to managing a persistent engineering agent—defining objectives, establishing metrics, configuring environments, and conducting final reviews.

marsbit06/06 08:11

Codex Goal Mode Usage Guide: How to Make AI Continuously Pursue a Specific Objective

marsbit06/06 08:11

From Ethereum to AI's 'CROPS': What Exactly is This Set of 'Slow Variables' That Vitalik Repeatedly Emphasizes?

In recent discussions, Vitalik Buterin has frequently emphasized the concept of "CROPS," a framework defining core values for Ethereum's development. CROPS stands for Censorship Resistance, Capture Resistance, Open Source, Privacy, and Security. Initially outlined in the Ethereum Foundation's "EF Mandate," it represents a commitment to user sovereignty, ensuring that the network resists external control, remains open, protects privacy, and prioritizes security. The relevance of CROPS extends beyond Ethereum's foundational principles, becoming crucial in the context of AI integration. As AI agents begin handling wallet operations and automated transactions, the risk increases that users may cede control over their digital assets, privacy, and intentions to centralized AI service providers. A "CROPS AI" would therefore emphasize local execution where possible, privacy-preserving remote model calls (e.g., using zero-knowledge proofs), and transparent, verifiable processes to maintain user agency. Vitalik highlights a significant convergence between "CROPS Ethereum access layer" and "CROPS AI." Both address the same fundamental challenge: how users can access powerful services—be it blockchain data via RPCs or AI models—without exposing sensitive information or relinquishing ultimate control. This intersection points toward a future digital entry point that is more private, secure, and user-controlled. Ultimately, CROPS is not merely an abstract ideal but a practical guidepost. It steers development—from protocol resilience and wallet design to AI agent safety—towards a future where users retain self-sovereignty even as digital systems grow more complex and powerful. In an era of accelerating AI adoption, these "slow variables" of censorship resistance, openness, privacy, and security may define Ethereum's enduring value.

marsbit06/05 12:40

From Ethereum to AI's 'CROPS': What Exactly is This Set of 'Slow Variables' That Vitalik Repeatedly Emphasizes?

marsbit06/05 12:40

Silicon Valley 'Startup Guru' Steve Hoffman: Web3 + AI Could Be a Trap

Silicon Valley investor and "Godfather of Startups" Steve Hoffman warns that combining Web3 with AI is likely a trap, not a promising venture. In an interview, Hoffman argues that while AI is a foundational technology touching all industries, Web3 adds complexity, friction, and regulatory risk without solving mainstream consumer or business needs. He advises founders to focus on deep, specialized applications where startups can out-iterate giants, rather than on generic features easily replicated by large tech companies. Hoffman observes that Silicon Valley will lead foundational AI research, while China excels at rapid, large-scale application and commercialization, particularly in robotics. He stresses that AI-driven autonomous agents capable of collaborative, multi-step tasks are 2-4 years away, which will cause significant job displacement. The solution is not to slow AI but to redesign business models around human-AI collaboration and reform social systems like education and retraining. For startups, Hoffman recommends focusing on vertical, expertise-heavy domains to build defensibility. He sees major opportunities in AI fraud detection and cybersecurity. Key founder mindsets include systemic thinking over feature-focus, relentless customer centricity, building adaptive teams, and deeply understanding AI's capabilities and limits. Hoffman is also leading a non-profit initiative to establish university centers aimed at training future leaders in responsible, human-value-aligned AI innovation.

marsbit06/05 11:18

Silicon Valley 'Startup Guru' Steve Hoffman: Web3 + AI Could Be a Trap

marsbit06/05 11:18

Crossing the 'Memory Wall': The Wafer-Level Revolution and Computing Power Routes in the AI Inference Era

In 2026, a historic shift occurred in AI as major cloud providers' inference spending surpassed training spending for the first time, signaling a move from "building large models" to "using large models." This shifts the core challenge from computing power to the "memory wall"—the bottleneck of data movement (model weights, activations, KV Cache) between external DRAM and processors, where energy and latency from data transfer far exceed computation itself. Companies like Nvidia face GPU idle time due to bandwidth limits. In contrast, Cerebras Systems adopts a radical "wafer-scale" approach with its Wafer-Scale Engine (WSE). Instead of cutting a silicon wafer into many chips, Cerebras uses almost the entire wafer as one massive chip (WSE-3). This design provides 44GB of on-chip SRAM, delivering memory bandwidth thousands of times higher than traditional HBM (e.g., 21 PB/s vs. Nvidia B200). For LLM inference, weights are streamed layer-by-layer from external MemoryX storage to the chip, avoiding HBM bottlenecks. This results in token generation speeds 1.5–5 times faster than Nvidia's B200 in some models and significant advantages in first-token latency and long-context tasks. Additionally, Cerebras's architecture offers much lower interconnect power consumption (0.15 pJ/bit vs. GPU's ~10 pJ/bit). However, Cerebras faces challenges: SRAM scaling has slowed with advanced nodes, limiting future capacity gains; the chip requires specialized liquid cooling and custom software stacks; and its external I/O bandwidth (150 GB/s) is low compared to NVLink, hindering multi-system scaling for very large models. Competition is intensifying. Major players are pursuing three paths: 1) Developing proprietary inference ASICs (e.g., Google TPU, Microsoft Maia), 2) Leveraging advanced packaging (e.g., TSMC's SoW) to democratize wafer-scale-like integration, potentially eroding Cerebras's process advantage within a few years, and 3) Exploring optical interconnects for ultimate bandwidth. Commercially, Cerebras is transitioning from a hardware vendor to a service provider, facing the immense challenge of building high-power, specialized data centers to meet large contracts (e.g., 250MW/year from 2026–2028). In conclusion, the AI inference era presents a fundamental architectural trade-off. Cerebras opts for extreme physical optimization for low-latency, single-task performance, while Nvidia prioritizes versatility and massive cluster throughput. The path forward remains uncertain, with technology and business models still evolving in the race toward advanced AI.

marsbit06/05 11:07

Crossing the 'Memory Wall': The Wafer-Level Revolution and Computing Power Routes in the AI Inference Era

marsbit06/05 11:07

Anthropic Cries Wolf: Is the AGI Threat Real, or Just an IPO Story?

Anthropic has published an article titled "When AI builds itself," discussing the emerging concept of "recursive self-improvement," where AI begins to actively participate in designing, training, testing, and optimizing its own subsequent versions. The company presents internal data showing that by May 2026, over 80% of code merged into its codebase was written by Claude, its AI model. Claude's capabilities have expanded to handling complex, open-ended engineering tasks, achieving a 76% success rate in such areas, and even contributing to research processes, such as optimizing code performance and conducting AI safety experiments. Anthropic outlines an evolution from human-driven development to AI-assisted workflows, culminating in the current stage where AI agents can autonomously write, run, and delegate code. The company cautions that the path toward a "closed loop," where AI continuously improves itself, is becoming visible. It calls for coordinated global mechanisms to potentially slow or pause frontier AI development to allow safety research and societal structures to catch up. However, the timing of this warning coincides with Anthropic's preparations for an IPO, framing the narrative not just as a safety concern but also as a demonstration of Claude's advanced capabilities and its integral role in accelerating Anthropic's own R&D—creating a potential "flywheel" effect for competitive advantage. This contrasts with OpenAI's recent, more policy-oriented discussion of the same risks, highlighting the competitive dynamics in the AI industry as companies position themselves in both the technological and regulatory landscape.

marsbit06/05 07:06

Anthropic Cries Wolf: Is the AGI Threat Real, or Just an IPO Story?

marsbit06/05 07:06

Breaking the Curse of DeFi Cascading Liquidations, Vitalik Proposes a New Solution

**Vitalik Buterin Proposes New DeFi Design to Eliminate Forced Liquidations** Ethereum co-founder Vitalik Buterin has published a proposal for a new decentralized finance (DeFi) architecture aimed at removing the automatic liquidation mechanisms prevalent in current lending protocols. The core idea involves creating synthetic assets using options as building blocks, fundamentally avoiding the抵押借贷结构 that triggers forced sell-offs. The proposal responds to a recurring flaw in DeFi: during sharp market downturns, mass自动清算 of under-collateralized positions can exacerbate price declines, creating systemic selling pressure and market instability, as evidenced by recent crypto market volatility. Buterin's model would split an asset like 1 ETH into two option-like derivatives, P and N, pegged to a price index with a set strike price and expiration. At expiry, an oracle determines the settlement price to allocate the underlying ETH between P and N holders. This design eliminates the "cliff" of instant liquidation. Instead, a position's value would gradually drift from its target peg if not actively rebalanced by the user, transferring the rebalancing decision from the protocol to the user or automated tools. A key advantage is the reduced reliance on high-frequency, real-time oracle price feeds, which are vulnerable to manipulation and errors in current systems. The delayed settlement in the options model allows for more robust, fault-tolerant oracle designs. However, significant challenges remain for practical adoption. High transaction costs (slippage) from frequent rebalancing on automated market makers (AMMs) could erode user funds. The model may not be suitable for stablecoins requiring a strict 1:1 dollar peg, as it inherently allows for value drift. Success would depend on developing new liquidity provisioning models and deep markets for these synthetic assets. The proposal represents a fundamental rethinking of DeFi risk management, challenging the industry to explore alternatives to被动集中平仓 rather than merely optimizing existing liquidation processes. It remains a theoretical framework awaiting implementation and testing by development teams.

foresightnews_api06/05 04:31

Breaking the Curse of DeFi Cascading Liquidations, Vitalik Proposes a New Solution

foresightnews_api06/05 04:31

Ethereum Foundation Researcher: Quantum Day Is Approaching, Plans to Complete Quantum-Resistant Migration by 2029

Ethereum Foundation researcher Justin Drake discusses the implications of a recent quantum computing breakthrough by Google’s quantum AI team, which demonstrated a 10x efficiency improvement in Shor’s algorithm against the secp256k1 elliptic curve used in Bitcoin and Ethereum. Notably, Google kept key algorithmic details confidential, using zero-knowledge proofs to verify the result without disclosure—a first in academia. Shortly after, the core optimization was independently reproduced, and an open-source competition (ecdsa.fail) emerged, further improving the algorithm by 8.4%. Meanwhile, startup Oratomic published research suggesting that neutral-atom quantum architectures could break secp256k1 with only 10,000 physical qubits, accelerating the timeline for "Q-Day"—the day quantum computers can break widely used cryptography. Drake estimates a 50% probability of Q-Day by 2032 and a 10% chance by 2030, contrasting with the U.S. government’s more conservative 2035 forecast. He warns against panic but stresses timely migration to post-quantum cryptography. Ethereum plans to complete its migration by 2029, covering consensus, data, and execution layers with hash-based systems. The Foundation is also developing leanVM, a formally verifiable zkVM, and has launched two $1 million initiatives to advance SNARK-friendly cryptography.

foresightnews_api06/05 04:07

Ethereum Foundation Researcher: Quantum Day Is Approaching, Plans to Complete Quantum-Resistant Migration by 2029

foresightnews_api06/05 04:07

Breaking the DeFi Cascading Liquidation Curse: Vitalik Proposes a New Solution

Vitalik Buterin has proposed a new DeFi design to eliminate the automatic liquidation mechanism that causes market instability during sharp downturns. The current system, used by protocols like Aave, triggers forced sales when collateral value falls below a threshold, often exacerbating price drops and creating systemic selling pressure. Buterin's alternative model is based on splitting an asset like ETH into two synthetic option-like tokens, P and N, pegged to a price index. Their combined value always equals one ETH. Instead of sudden liquidation, a position's value gradually drifts from its target peg if the market moves. Users must proactively rebalance their holdings to maintain their desired exposure, transferring the management burden from the protocol to the user or automated tools. A key advantage is the reduced reliance on real-time oracles. Pricing decisions are deferred until contract expiry, allowing for more robust, fault-tolerant oracle designs. This removes a clear liquidation threshold that speculators can target for manipulation or MEV extraction. However, significant challenges remain. Frequent rebalancing could incur high slippage and transaction costs, necessitating new liquidity provider models. The design is better suited for hedging instruments than for stablecoins requiring a rigid 1:1 peg. While not an immediate replacement for existing systems, the proposal challenges the foundational assumption that instantaneous forced liquidation is an unavoidable necessity in DeFi, opening the door for fundamentally different risk management architectures.

marsbit06/05 01:47

Breaking the DeFi Cascading Liquidation Curse: Vitalik Proposes a New Solution

marsbit06/05 01:47

Breaking News! Anthropic Calls for a Universal Pause in AI Research

Anthropic warns of AI self-evolution, reporting that over 80% of its internal code is now written by its AI, Claude. Productivity has surged, with engineers merging 8x more code than in 2024. Claude's performance on complex, open-ended tasks jumped from 26% to 76% success in six months, nearing human parity. The company introduces a new metric: "AI task duration." In 2024, AI handled 4-minute tasks; by 2026, it manages 16-hour tasks, with capability doubling every 4 months. Claude also reviews code, catching bugs that previously caused outages, and significantly outperforms humans in research tasks like optimizing code (52x speedup) and conducting AI safety experiments. Anthropic outlines three potential futures: 1) Progress plateaus, 2) AI accelerates but humans remain in control, or 3) AI achieves full recursive self-improvement (RSI), designing its own successors. This final path could revolutionize fields like medicine but also risks catastrophic alignment failure if control is lost. The call echoes similar concerns from OpenAI. Anthropic proposes a coordinated pause on AI development—if a verifiable mechanism to ensure all labs comply can be established.

marsbit06/05 00:27

Breaking News! Anthropic Calls for a Universal Pause in AI Research

marsbit06/05 00:27

From 'Old Guys' to 'New Favorites': How AI Is Revaluing Old Infrastructure from Dell to Nokia?

From "Vintage Tech" to "New AI Darlings": How AI Revalues Old Infrastructure One year ago, tech giants like Dell, Nokia, Cisco, and Western Data were seen as slow-growth, low-valuation stories, far from the AI spotlight dominated by players like Nvidia. Now, these legacy tech stocks are gaining market attention, sparking debate on whether this is genuine industry revaluation or a temporary narrative. As AI moves from model parameters to real-world data centers, the market is recognizing companies with proven delivery and infrastructure capabilities. This shift marks a change in the AI investment thesis: from pure model and GPU focus to the complex systems engineering required for deployment. Companies like Dell, HPE, and Corning are being revalued not for being "sexy" AI innovators, but for their decades of accumulated expertise in supply chains, enterprise delivery, and infrastructure—assets that have become critical in the AI buildout phase. The revaluation is unfolding across three key infrastructure lines: 1. **Servers & System Integration:** Dell and HPE are emerging as crucial system integrators or "general contractors" for AI data centers, translating GPU orders into complete, deployable server racks integrated with power, cooling, and networking. 2. **Networking & Connectivity:** AI's scale demands robust high-speed connections. Corning (fiber optics), Nokia (AI-RAN, 6G), and Cisco (data center switches) are gaining importance for enabling efficient data transfer within and between AI clusters. 3. **Storage:** Beyond high-speed memory (HBM/DRAM), the AI data explosion is driving demand for high-capacity hard drives (HDDs) from companies like Western Digital and Seagate to handle training data, logs, and cold storage cost-effectively. For this revaluation to be substantive and not just a narrative, three criteria are key: 1) Concrete AI-related order and revenue growth (e.g., Dell's AI server sales), 2) Upward revisions to company financial guidance, and 3) Sustainable improvements in profit quality, not just top-line revenue spikes. In essence, AI's transition to a real construction phase is re-pricing "old assets" against "new demand." The opportunity, however, is selective. Only those legacy firms that are demonstrably integrated into the capital expenditure chains of data center and enterprise AI deployment are likely to experience a true "logic re-rating" rather than just a temporary valuation bounce.

marsbit06/04 11:41

From 'Old Guys' to 'New Favorites': How AI Is Revaluing Old Infrastructure from Dell to Nokia?

marsbit06/04 11:41

活动图片

Hot Categoriesright-arrow

Hot Tagsright-arrow

Project Updatesright-arrow

Indepth Researchright-arrow