Nuanced View of AI Bubble Concerns

Nuanced View of AI Bubble Concerns | Generated by AI

Home 2026.07

Question: Is this view reasonable?

Answer:

Broadly, yes, but I’d make it more nuanced. Right now you’re mixing together several independent claims, some of which are widely accepted and others that are much more speculative.

A cleaner version would be:

我觉得现在谈 AI 泡沫有点像 1998-2000 年谈互联网泡沫。

AI 确实存在一些泡沫，比如估值过高、部分公司商业模式还没跑通、还有 OpenAI、英伟达、云厂商之间形成的资本循环。但这不代表整个 AI 产业是泡沫。

真正重要的是，AI 正在成为新的基础设施，像互联网、电力一样，会带来一轮新的生产力革命。即使未来资本市场调整，技术本身还是会继续发展。

相比之下，我觉得这种泡沫比房地产泡沫健康得多。房地产更多是存量资产价格上涨，而 AI 投资最终会沉淀成 GPU、数据中心、模型、软件和新的生产力。

同时很多趋势都在并行发生：美国经济增速放缓但依然保持科技领先，中国在制造业和 AI 上持续追赶，欧洲在 AI 基础模型和算力竞争中相对落后。未来几年，全球价值会围绕 AI、自动化和数字化重新分配。

A few comments on the specific statements:

“应该没有泡沫” → I’d avoid saying no bubble. There are clearly bubble-like characteristics in parts of AI (valuations, VC funding, startups with weak moats). A better phrasing is “有局部泡沫，但不是整个产业都是泡沫。”
“OpenAI 和英伟达循环交易” → There is an ecosystem where GPU vendors, cloud providers, AI labs, and investors reinforce each other’s growth, but calling it “循环交易” can sound like you’re alleging something improper. I’d instead say “资本和产业形成了正反馈循环” or “资本投入、算力采购和模型发展互相促进”.
“美国下行衰落” → This is the weakest claim. The U.S. has fiscal issues and slower GDP growth than before, but it’s still leading in frontier AI, semiconductors, software, and venture capital. I’d say “美国经济增长放缓，但科技创新仍然领先。”
“中国上行变强” → Reasonable if you qualify it. Manufacturing, EVs, batteries, robotics, and open-source AI have strengthened, while real estate remains a structural drag.
“欧洲在AI时代掉队” → Compared with the U.S. and China in frontier foundation models and compute, this is a common assessment. Europe remains strong in research and industrial automation but has fewer hyperscalers and frontier AI companies.

The main idea—that AI is causing a global reallocation of capital, talent, and economic value—is a defensible macro view. The only part I’d soften is making absolute geopolitical statements (“美国衰落”, “没有泡沫”), because the evidence there is much less definitive than the technological trend itself.

Back Donate