←News detail

Ayi AI NotesUpdated Nov 2 21:41

Li Jiaqi Sells Alibaba Cloud Tokens in Double 11 Livestream, 300,000 Units Sold Out

AI summary

Li Jiaqi began selling Alibaba Cloud Tokens in his Double 11 livestream, and 300,000 units sold out that night, with Alibaba's 1992-born vice president Chen Yusen personally appearing on stage with Qwen Office. The author believes this marks compute power moving from the developer back end and enterprise annual framework agreements to a mass consumer product, helping ordinary users develop the mindset of paying for a single invocation. Their judgment is that future competition will no longer be about parameter scale, but about who can drive compute costs down to cabbage-level prices.

Quoted post

AYi

@AYi_AInotes

English text is being prepared.

View original

当 CNN 主持人追问 AI 到底会不会让人类万劫不复时, 全球最大算力掌门人黄仁勋用一段极其通透的工程哲学,直接砸碎了非黑即白的AI恐慌叙事, 这是我今年为止看到的科技界最性感也最冷静的一段自白! 老黄说:真正的高手从不保证无害,他们只负责在狂飙中死死盯住刹车。 扎卡里亚的原话是:面对这场狂飙的 AI 革命,我们最后到底会不会平安无事? 老黄没有给出一个迎合大众的廉价安慰, 也没有陷入恐慌主义的末日叙事, 而是给出了一个工程界最清醒的长期定性。 在老黄看来,所有人都在逼技术给出一个非黑即白的保证, 但人类历史上任何一次真正的工业革命, 从来都不是在绝对无害的温室里完成的, 电力会带来火灾, 航空会有空难, 把每一种造出来的强大机器驯服成基础设施, 靠的从来不是祈祷不出意外, 而是在每一行代码和每一块芯片落地时 都保持不松懈的警惕。 这也是为什么他极其罕见地公开感谢了图灵奖得主辛顿那些刺耳的警报, 在大多数人眼里, 悲观者的警告是在给产业泼冷水, 但在真正掌舵硬件底座的人看来, 每一次对潜在失控的严厉提醒, 都是在逼整个行业把推进安全所消耗的心力, 拉到和推进算力完全对等的高度。 安全从来不是能力的绊脚石, 就像赛车上最坚固的刹车系统, 存在的唯一目的就是为了让车手敢把油门踩得更深。 未来一定会有局部的伤害和漏洞冒出来, 但当这些问题出现时, 世界需要做的是立刻扑上去解决它, 而不是因噎废食停下脚步。 因为这项技术最终带来的积极收益, 将会是压倒性而且不可阻挡的。 完整访谈出处、背景解析与核心思辨整理在一楼,需要自取。

View original post

Body · AI translation

Alibaba's overt commercial scheme is truly impressive. Li Jiaqi started selling Alibaba Cloud Tokens in his Double 11 livestream, and 300,000 units sold out that very night—you could say this is the turning point when AI compute power has fully become a mass consumer product. Alibaba's youngest vice president, Chen Yusen, born in 1992, personally took the stage, turning jargon that once belonged to programmers into something anyone can buy, like water, electricity, or gas.

It's not a programmer binding a credit card and topping up in a console, nor is it a big company's procurement department signing an annual cloud service framework agreement.

Alibaba's youngest vice president, Chen Yusen, born in 1992, brought Qwen Office to the scene in person, packaging the compute units once hidden in developer backends directly into a Li Jiaqi-style little yellow cart link, and 300,000 units were snapped up in an instant.

In the past, when ordinary consumers heard about Tokens, their eyes were full of the bewilderment of reading a sealed book: those were the obscure parameters that coders and algorithm people muttered about every day; when people made a PPT, looked something up, or generated a few images, they only knew whether the software lagged and how much the monthly subscription fee was—no one could understand what was actually being consumed behind every press of the Enter key.

Lofty technical jargon kept the public farther from underlying compute power than the Mariana Trench.

Tonight's livestream tore that disguise off completely. When Li Jiaqi shouted "Buy it," for the first time Token was laid bare as the water, electricity, and gas of the digital age. Just as ordinary people decades ago first bought electricity meters and gas cards, compute power is no longer an esoteric tech concept, but a household daily necessity that deducts fees by the character and by the amount for every word you type and every line of report you generate.

The most ruthless part is the overt scheme behind Alibaba's commercial calculations. In the past, compute power was sold to enterprises, with long payment cycles, harsh bargaining, and the need to painstakingly serve big clients' customization demands; now it is pushed directly to tens of millions of ordinary housewives and office workers, getting ordinary people to build, in their daily work, the mental habit of paying for each individual call. Here is an extremely cold, counterintuitive truth:

when compute power needs a top livestreamer to come on and hawk it, it shows that the tech myth of large models has officially come to an end. Future competition will no longer be about whose model has tens of billions more parameters, but about who can beat compute costs down to cabbage price and stuff them directly into every household's shopping basket. The endgame of the business world has never been poetic. The way the internet once used firewood, rice, oil, and salt to raid offline retail— today's AI compute power is using the exact same posture to sink down again and harvest every ordinary person's daily life.

Source

Ayi AI Notes