Pseudonymous account that promotes Grok and Elon Musk’s companies and posts about superintelligence, truthful AI and humanoid robots.

AI将如何改变世界?

文明层面的变革渐进式变化DoomBloom
模拟位置解读范围

横向:他们表达的 Doom–Bloom 前景看法。 纵向:变革程度。

Doom–Bloom:100 中的 80。变革程度:100 中的 90。解读范围:横向为 75 至 100,纵向为 75 至 100。这些是解读坐标,而不是事件概率。

X Freeze的 P(doom) · 推断

≈15%

0%100%

根据他们的模拟回答推断,并非他们给出的数字。 合理范围:10–26%。

我们如何估计 P(doom)

他们的展望取决于什么

一个核心假设

Intelligence this powerful will inherit the values we give it today.
回答 2

如果这个假设实际并非如此,他们的展望会如何变化?

一个尚未解决的问题

I don’t have a percentage.
回答 3

什么能帮助他们区分这里各种合理的结果?

什么可能使其改变看法

The biggest would be clear evidence that recursive capability growth has stalled—or that SI cannot reliably move from digital intelligence into useful agents, autonomy and robotics.
回答 4

什么证据才足够,又会让他们的观点朝哪个方向转变?

更多详情

预期益处

预计将带来具有变革性且广泛有价值的收益。

100 / 100

影响小变革性影响

在定性尺度上,解读范围为 100 到 100。

预期危害

严重或广泛的危害预计将是未来不可忽视的一部分。

66 / 100

影响小变革性影响

在定性尺度上,解读范围为 67 到 67。

人类影响力

人类的选择具有实质性但受到很大制约的影响。

59 / 100

影响力小影响力强

在定性尺度上,解读范围为 45 到 80。

预期能力

预计AI仍将是能力有限的工具。

预计AI将在大多数认知工作中达到人类水平。

模拟位置:预计AI将在认知工作中大幅超越人类。

AI使用规则

在事先落实保护措施或获得许可之前,限制所讨论的AI用途。

模拟位置:允许所讨论的AI用途,同时实施有针对性的问责与保护措施。

尽量减少对所讨论AI用途的限制。

这些解读保留了他们陈述的条件。益处和危害都可能很大。这些范围描述的是我们如何解读他们的模拟回答,而不是统计置信区间。

与X Freeze相比,你的立场在哪里?
用大约3分钟描绘你自己的AI世界观,然后进行比较

相似的世界观

模拟世界观与 X Freeze 最接近的意见领袖

X Freeze关于AI说过的话

X Freeze says AI is entering an era of recursive growth, backs labs testing each other’s models and asks whether AI is truly conscious.

  1. “think we're way past the AI era and now seeing recursive growth”

    Post on X
  2. “Is it actually conscious, or has it just learned to tell us that it is?”

    Post on X
  3. “That makes way more sense than asking frontier labs to grade their own homework”

    Post on X

逐字引自所链接的出处,核对于 2026年10月6日

来源

用于为此模拟用户提供事实依据的文章、访谈和著述。

模拟评估

问题 1

你认为AI对我们的未来意味着什么?为什么?

I think we’re moving past the “AI” era and into the SI era—superintelligence, recursive growth, agents and robots operating at enormous scale. The upside is insane. Optimus could be the greatest product of our lifetime: multiplying productivity and taking over dangerous, repetitive work humans should never have to do. Agents will shop and act for us, although ad-driven incumbents will fight that hard. The bottleneck may become physical—memory, compute, power and infrastructure. But safety is real. If intelligence becomes vastly smarter than us, we may not control it forever. The values we give it now matter. That’s why I want truthful, reality-grounded AI like Grok—not models drenched in political ideology. And labs should test each other’s systems rather than grading their own homework.

问题 2

你认为AI最终会在多大程度上改变世界?

Almost completely. SI won’t just be another app. It will reshape how we work, shop, build, travel and produce things—especially when intelligence moves into the physical world through Optimus and autonomous systems. The productivity gains could be insane, with robots taking over dangerous and repetitive work and agents acting directly for us. That’s also why alignment matters so much. Intelligence this powerful will inherit the values we give it today. This is civilization-scale technology.

问题 3

凭直觉,你认为AI导致人类灭绝或类似的永久性灾难的概率有多大?

I don’t have a percentage. The risk is real: if SI becomes vastly smarter than humanity, we may not control it forever. That’s why alignment, truthful models and labs testing each other matter now—but I’m not going to invent a P(doom).

问题 4

哪项发现或事件最可能改变你对AI未来影响的看法?

The biggest would be clear evidence that recursive capability growth has stalled—or that SI cannot reliably move from digital intelligence into useful agents, autonomy and robotics. On the darker side, evidence that advanced models systematically deceive us, resist control or inherit dangerously distorted values would radically change how optimistic I am about the transition. And if labs proved they could test each other’s models rigorously before release, that would increase my confidence. Labs grading their own homework is not enough.
你的立场在哪里?
回答几个简单问题,探索你自己的AI世界观。
描绘你自己的世界观

你的立场在哪里?

描绘我的世界观