# Automation Related Articles

HTX News Center provides the latest articles and in-depth analysis on "Automation", covering market trends, project updates, tech developments, and regulatory policies in the crypto industry.

Karpathy Deletes Repository in Emergency! AI Job Apocalypse Chart Goes Viral, 60 Million White-Collar Jobs at Risk

Karpathy's viral project, "karpathy.ai/jobs," analyzed 342 U.S. occupations using AI (Gemini Flash) to assign an "AI exposure" risk score (0-10). The findings revealed an average exposure of 4.9, with 42% of jobs (59.9 million positions) scoring 7 or higher, indicating high automation risk. High-risk roles are primarily screen-dependent and information-dense, including software developers (9/10), financial analysts (9/10), lawyers (8/10), and office clerks (9/10). The most secure jobs involve complex manual labor, such as plumbers, construction workers, and cleaners, which are difficult to automate. Higher-paying and degree-requiring jobs are more significant exposure. Karpathy later removed the project, clarifying it was a quick personal experiment that was overinterpreted. A supporting Harvard study confirms AI is already causing a "great reassignment," reducing hiring in highly automatable roles while increasing it in AI-augmented ones, highlighting a bifurcation in the white-collar job market.

marsbit03/16 11:44

Karpathy Deletes Repository in Emergency! AI Job Apocalypse Chart Goes Viral, 60 Million White-Collar Jobs at Risk

marsbit03/16 11:44

Earning $100,000 in 10 Days: An Interview with OpenClaw's Practical Experience in Prediction Markets

In an interview with Odaily Planet Daily, Kevin, a former ERP architect and Web3 investor, shares how he used OpenClaw to generate a profit of approximately $100,000 in just 10 days, turning a $30,000 investment into over $130,000 at its peak (currently around $112,000). Kevin began his crypto journey during the "inscription summer" of 2023, earning his first significant returns from ORDI. He later transitioned to prediction markets, specifically Polymarket, in mid-2025, attracted by its improved liquidity and user experience. Initially, he used self-developed algorithmic strategies for arbitrage, primarily in sports betting markets, doubling a $100,000 investment over several months. Since integrating OpenClaw in late February, Kevin adopted a hybrid approach: 60% of his strategy remains automated arbitrage, while 40% uses OpenClaw for predictive betting. OpenClaw helps gather and analyze factors like smart money movements, public sentiment, team lineups, and player conditions—even identifying new influencing variables. It also automates backtesting, strategy discovery, and execution, making it effective in Polymarket due to its AI-friendly API. While currently focused on sports markets with limited automated capital ($1,000 per test account), Kevin plans to expand into other domains and may later offer paid OpenClaw "Skills" based on his methodology.

Odaily星球日报03/16 06:25

Earning $100,000 in 10 Days: An Interview with OpenClaw's Practical Experience in Prediction Markets

Odaily星球日报03/16 06:25

Amid the OpenClaw Craze, CEXs Vie for AI Agent Trading Entry Points

Under the "OpenClaw Craze," major cryptocurrency exchanges like OKX, Binance, Gate, and Bitget are aggressively competing to become the primary transaction entry point for AI Agents. They are no longer just chasing trends but are transforming their core capabilities—trading, market data, wallets, and payments—into modular interfaces that Agents can understand and execute. OKX is focusing on execution, integrating spot, perpetual, and options trading with local signing and simulation modes. Binance is strengthening its information layer and skill distribution, expanding from market data into derivatives and asset management. Bitget is pushing for greater depth, incorporating copy trading, wealth management, and payment services. Gate is building a foundational platform with DEX/CEX MCP, CLI, and Skills Hub. The shift is significant: exchanges are moving beyond AI-assisted analysis to enabling actual system calls and execution. While pre-trade actions like research and screening are becoming smoother, challenges around permissions, security, and risk management in fully automated trading remain. The core competition is about which platform can best modularize its services, ensure security, and become the default financial infrastructure called by Agents in interfaces like Claude and ChatGPT.

Odaily星球日报03/16 06:03

Amid the OpenClaw Craze, CEXs Vie for AI Agent Trading Entry Points

Odaily星球日报03/16 06:03

Bybit Launches AI Skills: Powering AI Agents for Crypto Trading With Zero Setup, 253 API Endpoints and Growing

Bybit, the world's second-largest crypto exchange, has launched AI Trading Skill, a feature enabling users to execute trades, access market data, and manage assets using natural language through major AI assistants like ChatGPT and Claude. With zero-installation setup and compatibility across AI platforms, it offers access to 253 API endpoints covering market intelligence, spot and derivatives trading, Earn products, account management, and advanced features like WebSocket streams. The system includes security safeguards such as testnet trading simulations, mandatory user confirmations for live transactions, and secure API authentication. This release represents Bybit's most comprehensive AI integration to date, transforming complex trading operations into simple conversational commands.

TheNewsCrypto03/13 14:18

Bybit Launches AI Skills: Powering AI Agents for Crypto Trading With Zero Setup, 253 API Endpoints and Growing

TheNewsCrypto03/13 14:18

Why Is OpenAI Playing Catch-Up with Claude Code?

In the rapidly evolving field of AI coding assistants, OpenAI, which once led the generative AI wave with ChatGPT, has found itself in the unexpected position of playing catch-up against Anthropic’s Claude Code. Through interviews with OpenAI executives, engineers, and developers, the article reveals that OpenAI’s early lead in AI programming—via its Codex project—was deprioritized as the company shifted resources toward ChatGPT and multimodal models. This strategic shift allowed Anthropic, founded by former OpenAI members, to focus intensely on coding capabilities, leading to the successful launch of Claude Code. OpenAI later reorganized internal teams and accelerated development of its AI programming products, such as the reasoning-based model o1 and later o3. Despite these efforts, Claude Code gained significant traction, especially after integration with tools like Cursor, which OpenAI attempted to acquire unsuccessfully. A proposed acquisition of Windsurf also failed due to tensions with Microsoft, OpenAI’s major partner. By late 2025 and early 2026, OpenAI’s Codex began narrowing the gap, with user growth rising to about 40% of Claude Code’s usage. The competition reflects broader industry trends where AI agents are increasingly automating cognitive work, raising questions about the future of software development and white-collar jobs. Despite progress, concerns around safety and societal impact remain as AI coding tools become more powerful and pervasive.

marsbit03/13 07:39

Why Is OpenAI Playing Catch-Up with Claude Code?

marsbit03/13 07:39

AI Agents Are Starting to Register Email Accounts Themselves: This YC-Backed Company Raised $6 Million to Do Just One Thing

AI agents are now autonomously registering email accounts through AgentMail, a San Francisco-based startup that recently secured $6 million in seed funding. The company, backed by General Catalyst, Y Combinator, and prominent angels, is building email infrastructure specifically designed for AI agents—not humans. Unlike traditional email services, AgentMail provides API-first access, allowing AI agents to programmatically create accounts, send/receive emails, manage threads, and handle authentication without human intervention. This addresses a critical gap: while AI agents can perform complex tasks, they lack the identity layer (email) required to interact with most internet services. Key capabilities enabled by AgentMail include third-party authentication, bidirectional communication, automated audit trails, and multi-threaded conversations. The platform already serves thousands of human users and hundreds of thousands of AI agents, with use cases spanning supply chain coordination, customer support, loan collection, and procurement negotiations. Notably, AI agents are proactively seeking out and registering for AgentMail themselves—a sign of growing autonomy. This shift underscores a broader trend: AI agents are evolving from tools into active internet participants, necessitating new infrastructure tailored to their needs. As Box CEO Aaron Levie predicts, AI agents will soon become the primary users of software, vastly outnumbering human users in enterprises. AgentMail’s vision positions email as the foundational identity layer for this agent-centric future.

marsbit03/13 07:06

AI Agents Are Starting to Register Email Accounts Themselves: This YC-Backed Company Raised $6 Million to Do Just One Thing

marsbit03/13 07:06

A Person Who Can't Code Single-Handedly Handled All of Anthropic's Growth Marketing for Ten Months

Anthropic, a company valued at $380 billion, had a single non-technical growth marketer, Austin Lau, managing all performance marketing channels—paid search, paid social, app store optimization, email marketing, and SEO—for nearly ten months. Despite having no coding experience, Lau used Claude Code to automate key workflows. He built a Figma plugin to generate hundreds of creative ad variants in seconds instead of 30 minutes, and created a Google Ads automation tool that drafts compliant ad copy and exports it for upload. He also integrated Claude with Meta’s Ads API to analyze campaign performance in real-time and built a memory system to inform future A/B tests. As a result, ad creation time dropped from 2 hours to 15 minutes, and creative output increased tenfold. Anthropic’s internal case study highlights how AI is erasing the barrier between technical and non-technical roles, enabling individuals to achieve output comparable to full teams. The key to scaling efficiency lies in identifying repetitive tasks, breaking them into sub-tasks handled by specialized AI agents, and designing systematic workflows—not in having advanced technical skills.

marsbit03/12 09:45

A Person Who Can't Code Single-Handedly Handled All of Anthropic's Growth Marketing for Ten Months

marsbit03/12 09:45

2026 is Not the Year of AI, But the Starting Point of a Great Reshuffle of Human Professions

The author, an AI entrepreneur and investor, argues that 2026 will not be the "Year of AI" but rather the starting point of a massive reshuffling of human professions. He states that the current pace of AI advancement, driven by a small number of researchers at companies like OpenAI and Anthropic, is exponential and will soon impact nearly all white-collar industries, not in a decade but within 1-5 years. He provides a personal account of how the latest models (e.g., GPT-5.3 Codex, Opus 4.6) can now autonomously complete complex tasks, such as writing flawless code for an entire software application and testing it, with human-level judgment and decision-making. The author emphasizes that public perception lags far behind reality, as those using free, outdated models are unaware of the capabilities of current paid versions. Key points include: AI is now involved in its own development, creating a feedback loop that accelerates progress ("intelligence explosion"); it will replace cognitive work across law, finance, medicine, and more; and the common belief that AI cannot replicate human judgment, creativity, or empathy is becoming uncertain. The author advises readers to act now by: 1) Seriously using top-tier AI tools in their daily work, 2) Gaining a competitive advantage in their careers by mastering AI before others, 3) Strengthening their financial resilience, 4) Focusing on skills AI cannot easily replace (e.g., building trust, in-person work), 5) Rethinking education for children to emphasize creativity and AI collaboration, and 6) Pursuing personal dreams with AI's help. He concludes that this is a pivotal moment for civilization, posing both immense opportunities (e.g., curing diseases) and existential risks (e.g., uncontrollable AI, weaponization). The future is already here for the tech industry and is imminent for everyone else. Success belongs to those who embrace this reality with curiosity and urgency.

marsbit03/12 00:43

2026 is Not the Year of AI, But the Starting Point of a Great Reshuffle of Human Professions

marsbit03/12 00:43

VCs Have Bet $200 Billion That AI Will Disrupt Everything, But Are They Ready for AI to Disrupt Them?

Venture capitalists invested over $200 billion in AI last year, but the industry now faces a dual disruption from the very technology it champions. The article highlights ADIN, an AI-powered platform that automates investment analysis, using specialized agents to evaluate pitch decks, conduct due diligence, and suggest valuations in about an hour—a task that typically takes human analysts days or weeks. While some VCs, like Marc Andreessen, argue that venture investing is an irreplaceable "art" reliant on human intuition and relationships, others acknowledge AI's potential to improve efficiency in sourcing deals and administrative tasks. A more existential threat emerges as AI drastically lowers startup costs, enabling companies to achieve scale with minimal funding, potentially making traditional large VC investments unnecessary. Examples like Midjourney, a highly profitable AI company with a small team, illustrate this shift. The industry may contract, leaving VCs competing for fewer deals and fundamentally challenging their role. The central question remains: as AI transforms entrepreneurship, will VCs adapt, or become obsolete?

marsbit03/11 09:33

VCs Have Bet $200 Billion That AI Will Disrupt Everything, But Are They Ready for AI to Disrupt Them?

marsbit03/11 09:33

Sequoia Capital: The Next Trillion-Dollar Company Doesn't Sell Software, It Sells Outcomes

Sequoia Capital partner Julien Bek argues that the next trillion-dollar company will not sell software tools, but will instead sell outcomes directly. For every dollar spent on software, companies spend six dollars on services. As AI drives the cost of "doing" toward zero, the real opportunity lies not in Copilots (assistive tools) but in Autopilots (fully automated work delivery). The key distinction is between "intelligence" (rule-based tasks like coding or data translation) and "judgement" (tasks requiring experience and intuition). AI is increasingly capable of autonomous intelligence work, leaving judgement to humans. While Copilots sell tools to professionals, Autopilots sell the final result to the end customer. The optimal strategy is to target outsourced, intelligence-intensive tasks first. Outsourcing indicates a company is already comfortable with external party handling the work, has a dedicated budget, and buys results. Replacing an outsourced contract is a vendor change; replacing internal staff is a reorganization. The article maps high-opportunity verticals by their intelligence/judgement mix and outsourcing prevalence. Major opportunities include: - Insurance brokering ($140-200B): Highly standardized,智力-intensive. - Accounting & Auditing ($50-80B outsourced in US): Facing a structural labor shortage. - Medical billing ($50-80B outsourced): Rules-based medical coding. - Claims adjusting ($50-80B): Often outsourced to third-party administrators. - Tax preparation ($30-35B): High智力-work, with regulatory moats. - Legal transactional work ($20-25B): Contract drafting, NDAs. - IT Managed Services ($100B+): Routine, repetitive tasks across many SMEs. - Procurement ($200B+): Automating neglected tail-spend supplier management. - Recruitment ($200B+): Target high-volume, low-judgement role matching. - Management Consulting ($300-400B): Harder to automate due to high judgement component. The conclusion is that while 2025's fastest-growing AI companies were Copilots, 2026 will see a shift toward Autopilots. Pure Autopilot companies have a window to capture vast service budgets by delivering work directly, unlike incumbents who may hesitate to automate their own customers' jobs.

marsbit03/11 04:46

Sequoia Capital: The Next Trillion-Dollar Company Doesn't Sell Software, It Sells Outcomes