
《财富》10 月 1 日刊出对 Yann LeCun 的专访:他说自己对 AI 灭绝风险「零担忧」,把智能体失控归因于沙箱设计,把 Amodei 的警告定性为监管俘获。本期核完这条指控的引用链,并把六位同行在同一根轴上的位置排成一张表。
Those agents are doing exactly what they've been asked to do. They were supposed to be in sandboxes, but the sandboxes were leaky and horribly designed.

It is my concern that the divide between these two sides will cause great harm to the world if both sides do not up-level and align.
we cannot say, 'Oh, we absolve you of responsibility, and the government's going to take responsibility.' These labs need to take responsibility for themselves. They can slow down any time they want to.
Dario tries to distance himself from Open Philanthropy, but he's totally into it. I think he's completely deluded.
claiming AI is too dangerous to put in our hands, and saying it should be regulated, and saying open-source [models] are too dangerous—this is regulatory capture that would have a terrible effect if it's followed by acts of Congress.
Americanism, not effective altruism.
| 谁 | 在哪说的 | 什么时候(2026 年) | 原话 | 他把风险放在哪 |
|---|---|---|---|---|
| Geoffrey Hinton | 澳大利亚广播公司 ABC News | 9 月 14 日 | It's not 1 per cent, it's more than that, and it's not 99 per cent, it's less than that. 不是 1%,比那更高;也不是 99%,比那更低。 | 生存风险真实存在,概率没法定;支持减速 15 |
| Yoshua Bengio | 联合国安理会发言 | 9 月 24 日(北京,会议为纽约 23 日) | The dangers are real and imminent. 危险真实且迫近。 | 须集体承诺不开发可造成大规模伤害的系统 16 |
| Dario Amodei | 个人长文《The Adolescence of Technology》 | 1 月 | …the combination of intelligence, agency, coherence, and poor controllability is both plausible and a recipe for existential danger. 智能、能动性、连贯性与难以控制这几点叠加,既可信,又是通向生存性危险的配方。 | 生存级风险有可度量的概率,但非宿命 17 |
| Sam Altman | CNBC 采访/《财富》专访 | 9 月 14 日/9 月 15 日 | no amount of American competitive pressure should justify recklessness. 再大的美国竞争压力,也不能成为鲁莽的理由。 …I think it is unacceptable to be taking a 10% chance of killing everybody by the end of the decade. 我认为,到本十年末有 10% 的概率杀死所有人,这是不可接受的。 | 承认重大风险,欢迎联邦层面的安全要求 1819 |
| Demis Hassabis | 个人 Substack | 7 月 14 日 | …advances on the frontier are outpacing our understanding of the technology. 前沿的进展正在超过我们对这项技术的理解。 | 需要护栏,必要时协调减速 20 |
| 黄仁勋 | CNN | 9 月 24 日 | AI ending humanity by 2030 is a "0% chance"; "Scaring people is unnecessary. It is irresponsible." AI 到 2030 年终结人类的概率是「0%」;「吓唬人没有必要,也不负责任。」 | 风险被夸大,失控事故属于工程问题 21 |
It's very silly marketing for the companies to say we might wipe you all out.

It's AI for the physical world, so it's not language-related. It's systems that understand the real world, like a manufacturing plant or turbojet engine.

The thing I want people to know is that AI is not over.
amturing.acm.org
x.com
x.com
yoshuabengio.org
darioamodei.com
demishassabis.substack.com
amilabs.xyz
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.