Shanghai Fubei Pet Products Co., Ltd. submits listing application to Hong Kong Stock Exchange 上海福贝宠物用品股份有限公司向港交所提交上市申请书

Silicon Valley's tech giants shout about revolutionizing the world while secretly locking down their employees' token usage—it's like suddenly cutting the lights to check the electricity meter at the height of a party. After burning through billions, they're now pinching pennies on electricity bills? The so-called AI revolution ends up with even internal employees' computing power being meticulously budgeted. No matter how big this bubble inflates, someone has to pinch it first. Microsoft last week launched "Skills Self-Evolution," touting it as cultivating skills like training neural networks, but ordinary people just laugh bitterly: you're not even willing to let your own tokens be used openly, yet you talk about skill evolution? It's like a market auntie teaching you to code—before you remember any technical highlights, you already smell the scent of expired promotions.

Meanwhile, domestically, MiniMax and Zhipu are rushing to list on the A-shares market. Their frantic haste feels as if missing this window would mean the AI track is closing forever. The STAR Market has become a hot commodity, but the question is: are these companies truly technologically robust, or are they eyeing the enthusiasm of domestic retail investors? An AI programming company valued at $26 billion claims to have a "fully Chinese team" supporting the world's highest valuation—wait, $26 billion? USD? The number is so large it makes people wonder if a decimal point was missed. Programming tools are certainly useful, but can they rival the GDP of a mid-sized country? When capital goes into a frenzy, not even Newton's coffin lid can hold it down.

Then there's Guangdong debunking rumors about using AI for grading college entrance exam papers—a slap in the face to the notion of AI being omnipotent. Last year, there were rumors that AI would replace teachers; this year, officials officially debunked it: grading papers is still too mature for machines. What's more ironic is that such rumors go viral precisely because they expose the public's contradictory mindset of fearing and mythologizing AI simultaneously. People worry about losing their jobs to AI while expecting it to solve all problems—and the result? Even grading papers still requires humans. In the face of the college entrance exam, AI remains an intern who only knows how to press buttons.

Speaking of the "one-person company" boom, some earn millions annually while others see income shrink by 90%. This isn't business innovation—it's a large-scale performance art show. Pigs flying in the wind can only soar so high; once the wind stops, they all crash into meat patties. AI tools have lowered the barrier to entrepreneurship but not the cruelty of market competition. Those flaunting income screenshots might just be celebrating survivorship bias, while more people silently delete their accounts behind the scenes. Technology gives everyone a weapon, but most use it to shadowbox.

Looking back at Fubei Pet's listing and Yongmaotai receiving subsidies, these traditional enterprise stories seem remarkably serene amidst the AI frenzy. A pet supplies company rushes to list on the Hong Kong stock exchange; an aluminum company relies on government subsidies to stay afloat—they're not talking about disruption, just quietly making money. Perhaps this is the true backdrop of the economy: no matter how hot AI炒作 gets, meals must be eaten bite by bite, and dog food must be sold bag by bag. Those chasing trends often forget how to walk.

Microsoft's skill evolution, the sky-high valuations of AI programming companies, domestic companies' listing sprints—all these pieces form a bizarre tableau: technology leaps forward while simultaneously spinning its wheels. Token limits act like a bucket of cold water, waking some bubbles, but more bubbles are swelling elsewhere. We always anticipate AI bringing miracles, only to find it excels most at manufacturing new anxieties. Next time you hear about a "revolutionary breakthrough," perhaps you should first ask: Is your token balance still enough?

硅谷那些大厂一边喊着要颠覆世界，一边悄悄给员工的Token使用量上锁——这画面简直像在派对最嗨时突然关灯查电表。烧了几十亿美元之后，现在开始心疼电费了？所谓的AI革命，到头来连内部员工的算力都要精打细算，这泡沫吹得再大，也得有人先掐住呼吸。微软上周扔出个“Skills自我进化”，吹得天花乱坠，说是像训练神经网络一样培养技能，但普通人听了只会冷笑：你们连自家Token都舍不得敞开了用，还谈什么技能进化？这就像菜市场大妈教你怎么写代码，技术亮点没记住，先闻到一股过期促销的味道。

国内这头也没闲着。MiniMax和智谱急着回A股上市，这股火急火燎的劲儿，仿佛错过了这波窗口，下一秒AI赛道就要关门大吉。科创板成了香饽饽，但问题是，这些公司到底是真技术硬核，还是看中了国内韭菜的热情？估值260亿美元的AI编程公司，号称“全华班”撑起全球最高——等等，260亿？美元？这数字大得让人怀疑是不是少打了个小数点。编程工具固然有用，但能抵得上一个中型国家GDP吗？资本狂欢起来，连牛顿棺材板都压不住。

再看广东辟谣高考用AI改卷这事儿，简直是对AI万能论的一记耳光。去年还在传AI要取代教师，今年就官方打脸：批卷这种事儿，机器还嫩了点。但更讽刺的是，这种谣言能火，恰恰暴露了公众对AI既恐惧又迷信的矛盾心态。一边怕被AI抢了饭碗，一边又指望它解决所有问题——结果呢？连改个卷子都得真人上阵，AI在高考面前，还是个只会按按钮的实习生。

说到“一人公司”爆火，有人年赚百万，有人收入缩水九成。这哪是商业创新，分明是大型行为艺术现场。风口上的猪飞得再高，风一停全摔成肉饼。AI工具降低了创业门槛，但没降低市场竞争的残酷性。那些晒收入截图的，可能只是幸存者偏差的狂欢，更多人在后台默默注销账号。科技给了每个人武器，但大多数人用来打的是空气拳。

回头再看福贝宠物上市和永茂泰拿补助，这些传统企业新闻在AI喧嚣中显得格外宁静。宠物用品公司冲刺港股，铝业公司靠政府补助补血——它们没在谈论颠覆，只是默默赚钱。或许这才是经济的真实底色：AI炒得再热，饭还得一口口吃，狗粮还得一袋袋卖。那些追逐风口的，最后往往忘了怎么走路。

微软的技能进化、AI编程的天价估值、国内公司的上市冲刺，所有这些拼凑出一幅光怪陆离的图景：技术一边在飞跃，一边在原地打转。Token限制像一盆冷水，浇醒了一些泡沫，但更多泡沫正在别处膨胀。我们总在期待AI带来奇迹，结果发现它最擅长制造新的焦虑。下次再听到“革命性突破”时，不妨先问问：你们的Token余额还够用吗？