# Compute Articoli collegati

Il Centro Notizie HTX fornisce gli articoli più recenti e le analisi più approfondite su "Compute", coprendo tendenze di mercato, aggiornamenti sui progetti, sviluppi tecnologici e politiche normative nel settore crypto.

A Four-Page Internal Letter: What Card Is OpenAI Playing?

OpenAI's internal memo, revealed by The Information, outlines a strategic narrative against Anthropic across three key areas: revenue accounting, enterprise competition, and compute capacity. First, OpenAI CRO Denise Dresser challenged Anthropic’s reported $30B annualized revenue, claiming the actual net figure—using OpenAI’s accounting method—is $22B. The discrepancy stems from differing GAAP interpretations: Anthropic books gross revenue (including cloud partner shares), while OpenAI records net revenue after partner deductions. Second, enterprise adoption data from Ramp shows Anthropic rapidly closing the gap with OpenAI, narrowing from an 11% to a 4.6% difference within months. Anthropic already leads in high-value sectors like tech, finance, and professional services. Dresser acknowledged Anthropic’s edge in coding capabilities but warned against being a "single-product company" in a platform war. Third, while current compute capacity is comparable (OpenAI ~1.9 GW vs. Anthropic ~1.4 GW), OpenAI’s long-term plans aim for 30 GW by 2030—four times Anthropic’s projected 7-8 GW by 2027. Anthropic’s growth depends on sustaining enterprise revenue to cover rising cloud costs, estimated to reach $6.4B by 2027. The memo also highlighted OpenAI’s strategic shift: reducing reliance on Microsoft (which “limited customer reach”) and partnering with Amazon, which invests in both OpenAI and Anthropic. This places Amazon’s Bedrock platform as a battleground where both models compete for the same enterprise clients.

marsbit04/14 08:44

A Four-Page Internal Letter: What Card Is OpenAI Playing?

marsbit04/14 08:44

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

From 5 Cents per kWh Chinese Electricity to $45 API Export Packages: Token is Becoming the New Currency Unit

From 5 Cents per kWh Chinese Electricity to $45 API Export Plans: Token Emerges as a New Monetary Unit In 1858, the first transatlantic cable connected Europe and America, shifting information control from traditional media to those who owned the infrastructure. Today, a similar shift is occurring with AI and crypto, where Token is evolving from a technical term into a fundamental unit of machine-driven economy. Token serves a dual role: in AI, it is a computational unit for billing API calls and model inference; in crypto, it is a medium of exchange. These parallel systems are converging as AI Agents automate tasks—reading files, calling APIs, managing workflows—while consuming Tokens as fuel. Protocols like x402 and ERC-8183 are enabling machines to natively understand, call, and settle payments using Tokens, compressing complex processes into seamless, protocol-based actions. China’s "Token出海" (Token going global) narrative highlights this shift. With China’s annual electricity consumption exceeding 10 trillion kWh—a global first—and its growing dominance in data centers and GPU-driven inference, Token exports represent a new form of resource abstraction: Chinese electricity and compute power are being packaged into Token-denominated services consumed globally. Models like Minimax and DeepSeek rank highly on platforms like OpenRouter, with ~13% of global usage originating from Chinese models in 2025. OpenClaw exemplifies how Tokens transition from a cost (like "talk time") to a production input: Agents execute complex tasks, consuming Tokens at scale. This makes cost differentials critical, and China’s competitive pricing accelerates adoption. Moreover, AI Agents are not just to spend Tokens but also to earn—through memes, fees, or even mining—demonstrating early economic behaviors. Crypto provides the ideal settlement layer for Agentic commerce: permissionless accounts, programmable escrow, and micro-payments. x402 gives Agents wallets; ERC-8183 enables contracts with evaluation-based escrow. Together, they form a machine-native economic loop. Token’s rise is not about replacing fiat but becoming the base-layer unit for machine transactions—a universal measure for pricing compute, services, and digital resources. The future won’t have one currency, but Token may underpin the new economy, where the power to compress resources into Tokens defines value creation.

marsbit03/13 04:50

From 5 Cents per kWh Chinese Electricity to $45 API Export Packages: Token is Becoming the New Currency Unit

marsbit03/13 04:50

活动图片