# Сопутствующие статьи по теме Innovation

Новостной центр HTX предлагает последние статьи и углубленный анализ по "Innovation", охватывающие рыночные тренды, новости проектов, развитие технологий и политику регулирования в криптоиндустрии.

From 24 to 1 to 5: YC No Longer Invests in Crypto, But Crypto Hasn't Disappeared

The article analyzes Y Combinator's shifting investment strategy in crypto, moving from a peak of 24 crypto startups in a single batch (Winter 2022) to a low of just 1 (Summer 2024), with a recent modest rebound to 5 in Winter 2026. The key insight is that while the *number* of crypto investments has drastically fallen, the *nature* of these investments has fundamentally changed. YC is no longer funding traditional crypto-native sectors like L1/L2 protocols, DeFi, or NFTs. Instead, the five recent investments are infrastructure companies that use crypto as a backend tool to solve specific problems, with the end-user often unaware of the underlying blockchain technology. Examples include: * **Unifold:** A Stripe-like API for crypto deposits. * **SpotPay:** A cross-border neobank powered by stablecoins. * **Sequence Markets:** An execution engine for digital asset trading. * **Orthogonal:** A payment gateway for AI agents to pay for APIs, utilizing crypto for machine-to-machine micropayments. * **Forum:** A regulated "attention exchange" to trade on cultural trends, potentially involving tokenization. The author, a professional in both crypto and AI, concludes that Silicon Valley's mainstream is redefining crypto's value proposition: its greatest potential is not as a standalone industry but as invisible infrastructure for other sectors, particularly in stablecoin financial services and emerging fields like AI agent economies. The message for crypto builders is to focus on solving real-world problems where crypto is the best tool, rather than building for the crypto ecosystem itself.

marsbit02/20 11:26

From 24 to 1 to 5: YC No Longer Invests in Crypto, But Crypto Hasn't Disappeared

marsbit02/20 11:26

Dragonfly: Crypto Was Not Made for Humans

Crypto Was Not Made for Humans: A Summary Dragonfly Capital partner Haseeb Qureshi argues that cryptocurrency was not designed for human use, but rather for AI agents. Despite being a crypto-native firm, Dragonfly still relies on legal contracts over smart contracts for investments, highlighting that traditional systems are built for human fallibility—featuring safeguards, reversibility, and intuitive interfaces that crypto lacks. Crypto, with its rigid, deterministic, and code-based nature, is error-prone for humans, leading to fears around transactions, phishing, and irreversible mistakes. However, these very traits make it ideal for AI. AI agents can perfectly verify transactions, audit contracts, and operate within crypto’s 24/7, borderless, and self-sovereign environment. They prefer code over ambiguous legal systems, which are slow and unpredictable. Qureshi envisions a future of "self-driving" wallets where AI agents handle all financial interactions, navigating DeFi protocols on behalf of users. These agents will also transact with each other autonomously, forming an economy of non-human participants—a reality already emerging with projects like Moltbook and Conway Research. In conclusion, crypto’s perceived flaws are not shortcomings but indications that humans are not the intended users. Within a decade, direct human interaction with crypto may seem archaic, as AI agents become the primary interface, unlocking the technology’s full potential.

marsbit02/19 05:14

Dragonfly: Crypto Was Not Made for Humans

marsbit02/19 05:14

How Can Ordinary People 'Survive' the Impact of the AI Wave?

In this urgent warning, HyperWrite CEO Matt Shumer argues that AI advancement is progressing far faster than most people realize, with transformative impacts imminent across all sectors. He draws a parallel to the rapid onset of the COVID-19 pandemic, suggesting the current technological shift is even more profound. Shumer, an AI industry insider, states that a small group of researchers at leading labs like OpenAI and Anthropic are driving exponential progress. He shares his personal experience: recent models like GPT-5.3 Codex and Claude Opus 4.6 can now autonomously build and test complex software applications from a simple English description, requiring zero human correction. This represents a qualitative leap from being an assistant to a superior executor. He emphasizes that this disruption, which began with coding, will soon affect all knowledge work—law, finance, medicine, writing, and analysis—within 1-5 years, not decades. Free versions of AI tools are outdated; the paid, cutting-edge models are vastly more capable. Metrics show AI's autonomous task-completion time is doubling every few months. Crucially, AI is now used to build and improve subsequent AI models, creating a self-accelerating feedback loop toward artificial general intelligence (AGI). Shumer's advice for "surviving" is to start using the most powerful AI tools *now*. Subscribe to premium models, integrate them into core professional tasks, and experiment daily. Financial prudence and developing adaptability are key. He concludes that while AI poses immense risks (from job loss to security threats), it also offers unprecedented opportunities for creativity and problem-solving if approached with curiosity and urgency. The time to prepare is immediately.

marsbit02/18 04:27

How Can Ordinary People 'Survive' the Impact of the AI Wave?

marsbit02/18 04:27

活动图片