←News detail

AYi AI NotesUpdated Nov 2 19:35

Jensen Huang responds to CNN: AI safety is like the brakes on a race car

AI summary

In a CNN interview, Zakaria pressed Jensen Huang on whether the AI revolution will ultimately leave humanity safe and sound. He responded from an engineering perspective: electricity brought fires, aviation brought plane crashes, and taming powerful machines depends on constant vigilance rather than guarantees of absolute harmlessness. He rarely publicly thanked Turing Award winner Hinton for his warnings, and believes that safety and the advancement of computing power should receive equal effort—safety is the brake that lets the driver dare to press the accelerator hard, not a stumbling block.

Quoted post

AYi

@AYi_AInotes

Stripe co-founder John Collison and OpenAI CEO Sam Altman, founders of two hundred-billion-dollar companies, sat in a bar drinking beer and talked about a very interesting question: when in this round of the AI wave did it most resemble the early internet era? Altman did not mention the trillion-dollar computing power race, nor did he discuss Wall Street's capital stories, but instead told a real detail about an eight-to-ten-year-old child. That child spent only ten minutes and whipped up a complete 3D game on the spot, and afterward looked calm, as if the world should have been this way all along. Adults are used to treating large models as office plugins for improving efficiency, calculating code completion rates and the percentages of cost reduction and efficiency gains. But the real technological rupture often happens in the next generation; they have not even had time to learn what syntax errors and compilation environments are, yet the intuition in their minds already precedes all professional skills, and has directly grown into a world they can play with at will. This irreversible sense of dependence is also happening among top engineers. Altman mentioned that inside OpenAI, Codex once malfunctioned and briefly went down, and engineers across the office suddenly fell into an almost instinctive panic. Everyone suddenly realized that they could hardly imagine going back to the days of typing raw code line by line. Actually, the most fascinating part of the early internet was never the later monopoly of giants, but that savage era when no one needed anyone's approval, and a person alone in a bedroom could connect their imagination to the whole world. When the threshold for generating code is completely flattened to within ten minutes, skill recedes into the lowest background noise, and the only thing that can widen the gap between people is the pure desire you originally had to make something happen. The full interview background, the source of the original video, and the details of the wonderful conversation are compiled on the first floor; help yourself.

View original

两个百亿美金公司的创始人@stripe 联合创始人 John Collison 和OpenAI CEO Sam Altman 坐在酒吧里喝着啤酒,聊到了一个很有意思的问题, 这一轮 AI 浪潮到底在什么时候最像当年的早期互联网。 Altman 没有提万亿市值的算力竞赛, 也没有聊华尔街的资本故事, 而是讲了一个八到十岁小孩子的真实细节。 那个孩子只花了十分钟就现场搓出了一个完整的 3D 游戏,而且做完之后一脸平静,觉得这个世界本就该是这个样子。 成年人习惯把大模型当成提高效率的办公插件, 算着代码补全率和降本增效的百分比。 但真正的技术断层,往往发生在下一代人身上, 他们甚至还没来得及学会什么是语法错误和编译环境, 脑子里的直觉就已经先于所有专业技能, 直接长成了可以随意把玩的世界。 这种不可逆的依赖感在顶级工程师身上同样在发生,Altman 提到 OpenAI 内部有一次 Codex 发生故障短暂宕机, 整个办公室里的工程师突然陷入了一种近乎本能的慌张,大家猛然发现, 自己竟然已经很难想象重新回到一行行敲原始代码的日子。 其实早期互联网最迷人的地方, 从来不是后来的巨头垄断, 而是那个谁也不需要经过任何人批准、 一个人在卧室里就能把想象力连上全世界的蛮荒时代。 当生成代码的门槛彻底被抹平到十分钟以内, 技能就退成了最底层的背景噪音, 唯一能拉开人与人之间差距的, 只剩下你原本有多想把一件事做出来的纯粹渴望。 完整访谈背景、原片出处与精彩对谈细节整理在一楼, 需要自取。

View original post

Body · AI translation

When the CNN host pressed him on whether AI will ultimately bring utter ruin upon humanity, Jensen Huang, the man at the helm of the world's largest computing power, used a remarkably lucid piece of engineering philosophy to shatter the black-and-white AI panic narrative outright, and this is the sexiest and most sober confession from the tech world I have seen all year!

Huang said: true masters never guarantee harmlessness; they are only responsible for keeping their eyes fixed on the brakes amid the breakneck speed.

Zakaria's exact words were: in the face of this breakneck AI revolution, will we ultimately come through safe and sound? Huang did not offer a cheap comfort that panders to the masses, nor did he fall into the doomsday narrative of alarmism, but instead gave the clearest long-term characterization from the engineering world.

In Huang's view, everyone is forcing technology to give a black-and-white guarantee, but every genuine industrial revolution in human history was never completed in a greenhouse of absolute harmlessness,

electricity brings fires, aviation brings crashes, turning every powerful machine we build into infrastructure depends never on praying that no accident will happen, but on, as every line of code and every chip is deployed, maintaining unrelenting vigilance.

This is also why he, extremely rarely, publicly thanked Turing Award winner Hinton for those jarring alarms. In most people's eyes, the pessimist's warnings are pouring cold water on the industry, but in the eyes of the person truly steering the hardware foundation, every stern reminder of potential loss of control is forcing the entire industry to raise the effort spent on advancing safety to a level fully equal to that spent on advancing computing power.

Safety has never been a stumbling block to capability, just like the sturdiest brake system on a race car, whose sole purpose for existing is to let the driver dare to press the accelerator deeper.

The future will certainly see localized harms and vulnerabilities emerge, but when these problems appear, what the world needs to do is immediately pounce on them and solve them, rather than giving up eating for fear of choking and stopping in its tracks.

Because the positive benefits this technology ultimately brings will be overwhelming and unstoppable.

The full interview source, background analysis, and core reflections are compiled in the first comment; take them if you need them.

Source

AYi AI Notes