A cluster of crypto and tech developments landed on July 27, led by Changxin’s blockbuster market debut and Circle’s latest infrastructure push. According to crypto KOL AB Kuai.Dong, Country Garden invested RMB 2 billion in Changxin in September 2021 at RMB 2.22 per share, securing 901 million shares, but later exited at cost by the end of 2024 as it grappled with a property-sector liquidity crisis. Had it held the position until now, the return would have reached 23x. Changxin then closed its first trading day up 465.8% at RMB 49, with a total market value of RMB 3.28 trillion. In crypto, Circle said it acquired IBM’s blockchain patent portfolio, covering more than 680 patent families and nearly 1,000 granted patents tied to core blockchain technology, banking, financial services, insurance, enterprise infrastructure, supply-chain verification and secure cloud operations. In the U.S., the Senate’s latest merged text of the Digital Asset Market Clarity Act entered a critical two-week window before the August recess, with ethics provisions emerging as the main sticking point. Onchain data also drew attention: 6.93 million HYPE are set to be unstaked over the next seven days on Hyperliquid, while the platform’s nine largest whale addresses were collectively sitting on about $48.635 million in unrealized losses.

NVIDIA said it has formed the Open Secure AI Alliance with a group of founding members including Adobe, CrowdStrike, IBM, Microsoft, Red Hat, SpaceXAI and organizations such as the Linux Foundation, according to the NVIDIA Blog. The alliance is focused on building and sharing open-source models, agent harnesses and security tools for cybersecurity and AI safety defense. Initial contributions named by the group include NVIDIA’s NOOA agent framework, the SPIFFE/SPIRE zero-trust identity standard with HPE participation, Hugging Face’s Safetensors model weight format, IBM and Red Hat’s Lightwell supply-chain signing patch, Microsoft’s MDASH multi-model vulnerability scanning system, and SpaceXAI’s open-source Grok Build coding agent alongside plans to open Grok model weights. The alliance also called on regulators to treat open models and tools as defensive assets rather than as a regulatory burden.

NVIDIA CEO Jensen Huang used a 70-minute Axios interview to make a blunt case that Wall Street has twice misunderstood what stronger Chinese AI models mean for the chip market. His argument was simple: better models drive wider use, wider use demands more data centers, and more data centers mean more compute infrastructure. In that framework, models such as DeepSeek and Kimi are not bearish for NVIDIA. They are part of the demand engine. Huang also rejected AI doomsday claims in unusually direct language, calling predictions that AI will destroy humanity or wipe out half of U.S. jobs "nonsense." To support that view, he cited examples including a roughly 20% increase in radiologists, a roughly 10% increase in paralegals, and about 50% growth in manufacturing jobs in recent years, though the interview did not provide source attribution for those figures. On industry structure, Huang said the semiconductor sector needs to grow 5x to 10x over the next decade and argued that a bubble is unlikely in the next five years because supply remains constrained across chips, memory, land, power, and construction labor. He also said NVIDIA’s sales in China are "approximately zero," while adding that any return would be an honor. The interview ranged well beyond chips, covering open versus closed models, AI economics, robotics, agents, regulation, and U.S. industrial policy.

NVIDIA co-founder and CEO Jensen Huang used a 70-minute interview recorded at the company’s Fort Worth, Texas factory to push back on several of the biggest arguments now shaping the AI industry. He said Wall Street had misread the effect of Chinese models such as DeepSeek and Kimi, arguing that stronger models drive broader adoption, and broader adoption leads to more computing demand, more NVIDIA systems, and more data centers. Huang also rejected doomsday claims around AI, calling predictions that AI will destroy humanity or wipe out half of U.S. jobs “nonsense,” while citing examples he said show employment growth in radiology, legal support, and manufacturing. On industry capacity, he said the semiconductor sector will need to expand by 5x to 10x over the next 10 years and that a bubble is unlikely within five years because supply remains constrained across chips, memory, land, electricity, and construction labor. Huang also spoke about Donald Trump, open models, Anthropic’s Mythos, robotics, and the coming era of billions to trillions of always-on AI agents, while warning that governments could still make mistakes if they let science-fiction narratives drive overregulation.

OpenClaw AI agents hit 160K GitHub stars after the 2026.2.2 security update. Solana and Base see surging real-world applications; Moltbook runs a $30K USDC hackathon pushing autonomous agents from experiments to practical workers.

Ripple has introduced the XRPL AI Starter Kit to help developers build AI agents that can pay with XRP and RLUSD, entering an x402 market where USDC still accounts for most activity and volume.

Tether has released its open-source Mining Development Kit, aiming to unify hardware, power, and monitoring systems for Bitcoin miners from home setups to gigawatt-scale farms.

Ethereum co-founder Vitalik Buterin has abandoned all cloud AI services, running a fully local and sandboxed AI system. He warns that roughly 15% of AI agent skills contain malicious instructions and details his private LLM setup using an Nvidia 5090 GPU and Qwen3.5 model.