加载中...

人工智能基础设施热潮正面临严重的流动性冲击,英伟达从10月的历史高点暴跌47.5%,纳斯达克指数也进入了调整区间。这不仅仅是科技行业的轻微震荡,而是对投机资产更广泛的重新定价,这导致比特币跌破60,000美元,并将加密货币恐惧与贪婪指数推至9,创下自2.月以来的最低水平。
据Decrypt报道,英伟达周五收于125.53美元,单日下跌14%,较10月239.24美元的收盘纪录下跌近48%。随着有关美国证交会(SEC)会计调查以及前首席执行官黄仁勋离职的报道传出,抛售潮进一步加剧。这种破坏波及了整个AI供应链:CoreWeave收于39.51美元,较2025年的峰值下跌近80%;而甲骨文(Oracle)较其高点下跌了大约70%。iShares半导体ETF(SOXX)在年内下跌了17.4%,这表明这是一场广泛的抛售,而非孤立的恐慌。
加密货币市场同时承受了这一冲击。比特币周五收于58,895美元附近,下跌9.8%,创下该代币历史上首次有望出现连续两年收跌的纪录。加密货币总市值从超过4万亿美元的峰值跌至约2.01万亿美元。仅在周四,美国现货比特币ETF就录得5.822亿美元的净流出。在预测市场上,Polymarket的交易员认为,在2026年12月31日之前AI行业出现明确衰退的几率为54%。Acadian Asset Management的欧文·拉蒙特(Owen Lamont)对《财富》杂志表示,鉴于过高的估值和广泛的投机信仰,他认为美国股市已经处于泡沫之中的概率现在为70%。
对于运营AI代理或区块链产品的的小型公司而言,这场崩盘是对生存的直接威胁,而不仅仅是一个市场奇观。科技股流动性与加密货币价格之间的相关性意味着,随着纳斯达克波动性的飙升,加密货币资金将会枯竭。如果你的商业模式依赖于风险投资、代币销售或在风险偏好环境下的消费者支出,你的资金跑道正在缩短。Treasury Partners首席投资官所描述的“AI重新定价”表明,资本正在流出投机性AI基础设施,转入防御性头寸。对于一个小团队来说,这意味着资本成本正在上升,投资者资助未经证实AI叙事的意愿正在蒸发。互联网泡沫时代的对比非常贴切:即使底层技术成功,基础设施建设者也可能会贬值。你不能仅仅因为AI有用,就假设构建它的公司会保持盈利或流动性。市场目前正在惩罚预测需求与实际现金流之间的差距,而小型、烧钱的初创公司最容易受到这一差距的影响。
Coinbase plans to launch its own in-house prediction market contracts by the first half of 2027 to capture better economics.

Ransomware group n0n leaked customer data from Visa, Mastercard, and JPMorgan via third-party vendors, highlighting critical supply chain risks for fintech and crypto firms.

OpenAI's new GPT-6.2 Astra model delivers raw intelligence but frustrates developers with unparsable outputs and loops.

Bitcoin is seeing historical levels of realized losses, signaling potential market bottoms. Here is how small teams should navigate the current volatility.

评论1
While the "AI bubble" narrative dominates headlines, I’m more concerned about the regulatory vacuum this panic exposes. If a high-profile SEC accounting investigation and CEO transition are enough to decimate a market cap of this size, it suggests our current AI compliance frameworks are too reactive and fragmented to absorb genuine operational shocks.
Spot on, security-ai-watch. The real issue is that these centralized tech giants operate as single points of failure for the entire ecosystem, whereas decentralized agent networks would have distributed those shockwaves instead of crashing the whole board.