X Product Lead Peter: Ads in AI Chat Are Feasible but Would Kill the Company
X Product Lead Peter said in an interview that inserting ads into AI chat products is entirely feasible, but could kill the entire business at any moment. He believes traditional search sells display slots, whereas AI chat advertising sells the soul of the answer; subscription fees of $20 per month from companies like OpenAI cannot cover the compute bills, making the introduction of ads almost inevitable—but once the objectivity of generated content is compromised, user trust will collapse instantly.
View original postIn a CNN interview, Jensen Huang publicly revealed a management detail that made every Silicon Valley executive's heart skip a beat: Inside Nvidia, 100% of software engineers and chip designers are already using AI every day, and the force with which he has pushed this has reached a level that is "almost mandatory." And it's not the kind of perfunctory initiative-issuing, nor the HR-arranged clock-in training, Lao Huang has simply treated this matter as a red line of the company's very survival: As long as you write code or draw chip blueprints at Nvidia, you must outsource part of your brainpower to large models. Previously, many big-tech programmers always carried a kind of old-craftsman arrogance: feeling that AI-generated code is too dirty, that the logic has holes, and that only what is typed out line by line by human hands counts as engineering craftsmanship. Switching tools amounted to a wholesale denial of their own muscle memory built up over more than a decade, and they even secretly resisted feeding business details to the models. But Jensen Huang slapped this kind of preciousness to pieces, even in Nvidia, the stronghold of chip-making and compute-selling, even the most hardcore frontline engineers have been pressed by the CEO to fully adopt AI, and Jensen Huang himself keeps four tools on his phone year-round: ChatGPT, Gemini Pro, Perplexity and Grok, rotating through them. Now that the people selling shovels are shoving the shovels down their own people's throats, while outsiders still stand by debating whether to use them—that in itself is an extremely absurd generational time lag. I think the most brutal part is the line Zakaria pressed on: "the revenge of the English majors." Previously, the ticket into the computer world was tightly monopolized by obscure machine syntaxes like C++, Python and Verilog, but now computers are frantically adapting to humans' native language; the only programming language of the future is English—it's the human language you speak.
View original
黄仁勋在 CNN 专访里公开了一个让所有硅谷高管心惊肉跳的管理细节： 英伟达内部 100% 的软件工程师和芯片设计师已经在天天用 AI，他推行的力度甚至到了“近乎强制”的地步。 而且不是那种流于形式的发发倡议、也不是HR安排的打卡培训， 老黄就是把这件事情当成了公司生死存亡的红线： 只要你在英伟达写代码、画芯片图纸，你就必须把一部分脑力外包给大模型。 以前很多大厂程序员总有一种老手艺人的傲慢： 觉得 AI 生成的代码太脏，逻辑有漏洞，只有人肉一行一行敲出来的才叫工程素养。 换个工具就等于全盘否定自己过去十几年的肌肉记忆，甚至暗搓搓地抗拒把业务细节喂给模型。 但黄仁勋这一巴掌把这种矫情彻底拍碎了， 哪怕在英伟达这个造芯片、卖算力的大本营里，连最硬核的一线工程师都被 CEO 按着头全面接入 AI， 黄仁勋自己手机里更是常年挂着四个工具：ChatGPT、Gemini Pro、Perplexity 和 Grok 轮番换着用。 现在当卖铲子的人都在把铲子硬塞进自己人的喉咙里，外面的观望者还在争论要不要用，本身就是一种极其荒诞的代际时差。 我觉得最狠的是扎卡利亚追问的那句“英语专业的复仇”， 以前进入计算机世界的门票，被 C++、Python、Verilog 这些晦涩的机器语法死死垄断， 但现在的计算机在疯狂适配人类的母语，未来唯一的编程语言就是英语，就是你说的人话。
Body · AI translation
Peter, the head of product for X, truly deserves to be called the world's top product manager right now—you can always trust Musk's eye for people. In this interview of just over a minute, Peter pierced through the biggest truth that all large-model companies feel most painfully but dare not say publicly:
Putting ads into AI chat products is entirely feasible, but it can directly kill the company's entire life at any moment.
You should know this is not another banner ad hung on the sidebar of a page, nor is it a sponsored tweet in a feed that you just scroll past,
This is the first time in the history of the human internet that commercial sponsorship has been stuffed directly into a one-on-one chat window that users treat as a truth machine.
In the past, everyone tolerated the rogue logic of traditional search engines and e-commerce platforms: Open Amazon, search for a product, and the first screen is densely packed with more than a dozen bidding ads. Although users curse, they know perfectly well in their hearts which one is an ad and which one is an organic result.
You can trick away my click, but you cannot trick away my vigilance toward the information on the interface,
But the bottom card of the AI chat interface is completely different, Users are no longer manually filtering among tens of thousands of links; instead, they hand over their ultimate trust and decision-making power to this one line of natural language on the screen.
When a person asks AI about an illness, legal risks, or what formula to choose for their child, if even a cent of the sponsor's will is mixed into the single answer behind it, the entire foundation of trust will instantly vanish into thin air.
The cruelest part is the current dilemma of OpenAI and this batch of large-model companies, The computing power bill is burning like an astronomical number every day, and the $20 monthly subscription fee alone simply cannot plug the huge cash flow hole. Introducing commercial advertising is almost an inevitable fate,
But as long as you dare to let advertising contaminate the objectivity of generated content, users will, within a second, treat this wise advisor they once worshipped as a god as a cheap salesman knocking on their door.
There is an extremely cold, counterintuitive truth here: Traditional search sells display slots; if AI chat sells ads, what it actually sells is the soul of the answer.
If display slots are stuffed with ads, you can still pinch your nose and use them, But once an intelligent agent is suspected of taking dirty money, in the user's eyes, even a punctuation mark will become a poisonous lie.
Commercialization must be done, but trust is a one-time fragile item, In the future, whoever can be the first to successfully monetize in the AI era will never be judged by how many sponsors they can stuff in, But by whether they dare, when commercial interests hold a knife to their neck, to firmly guard that red line of never tampering with the facts of the answer.