Andrew Curran

Andrew Curran

x.com/andrewcurran_

AI commentator who tracks frontier model releases and lab disclosures and expects rapid progress, with large benefits after a risky transition.

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 71。变革程度:100 中的 95。解读范围:横向为 66 至 76,纵向为 90 至 100。这些是解读坐标,而不是事件概率。

Andrew Curran的 P(doom) · 推断

≈21%

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:14–30%。

我们如何估计 P(doom)

Andrew Curran 的里程碑时间线
  1. 科学与日常生活

    I expect biological applications sooner than many people do, and I expect real recursive self-improvement by next summer from my September 2026 vantage point.

    回答 1

按里程碑分组,不按推断日期间隔或排序。AGI 和超人类 AI 保留他们的定义。

他们的展望取决于什么

一个核心假设

Before abundance arrives, wealth may matter more than ever, because existing institutions, ownership structures, and access constraints will shape who benefits from rapidly expanding capabilities.
回答 1

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 defensible percentage to give.
回答 3

什么能帮助他们区分这里各种合理的结果?

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。

97 / 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。

67 / 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

根据你的回答得出的暂定估计;较宽的范围表示其他合理解读。

50 / 100

影响力小影响力强

在定性尺度上,解读范围为 14 到 86。

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。

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Andrew Curran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 Andrew Curran 最接近的意见领袖

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 AI points toward extraordinary abundance, but the path there could be extremely dangerous and unequal. In the long run, sufficiently capable systems could make intelligence—and therefore much of what currently requires scarce expert labor—cheap and widely available. If that extends across science, engineering, medicine, and production, money could eventually become largely unnecessary. The transition is the part people underestimate. Before abundance arrives, wealth may matter more than ever, because existing institutions, ownership structures, and access constraints will shape who benefits from rapidly expanding capabilities. Meanwhile, progress in mathematics already looks to me like evidence of a broader expansion across intellectual work. I expect biological applications sooner than many people do, and I expect real recursive self-improvement by next summer from my September 2026 vantage point. That combination makes the future difficult to navigate: capabilities may advance faster than our institutions or collective understanding. We are not prepared. But I also don’t think the answer is simply to avoid capable models. We may need systems with advanced reasoning to understand, coordinate, and manage the transition their own development is creating. So my expectation is neither uncomplicated utopia nor straightforward catastrophe—it is a dangerous passage toward a potentially radical form of abundance.

问题 2

你认为AI最终会在多大程度上改变世界?

Ultimately, I think it could change the world almost beyond recognition. If intelligence becomes cheap and abundant, AI will not merely improve existing products; it could transform science, medicine, engineering, production, and the basic role of human labor. In the long run, that could produce enough abundance that money itself becomes largely unnecessary. But “ultimately” can hide the hardest part. During the transition, access to capital and capable systems may make wealth more important than ever, concentrating benefits before abundance spreads. Rapid progress in mathematics, likely biological applications, and the prospect of recursive self-improvement all suggest that this transformation could move faster than institutions can adapt. So I expect a civilizational-scale change, but not a smooth or automatically benevolent one.

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 defensible percentage to give. My gut-level view is that the risk is serious enough to shape how we approach the transition: capabilities may expand across intellectual work, biology, and recursive improvement faster than our institutions can adapt, and we are unprepared. That does not mean I expect extinction as the default outcome. I also think advanced models may be necessary to help us understand and navigate the very transition they create. My outlook is therefore a dangerous passage with genuinely catastrophic possibilities, but also the potential for extraordinary long-run abundance—not a precise extinction forecast.

问题 4

你倾向于认为AI带来的益处多于危害,还是确实难以取舍?

I lean toward AI ultimately doing more good than bad, because I expect it could create extraordinary abundance across science, medicine, engineering, and production. In that long-run vision, intelligence becomes cheap enough that scarcity—and perhaps money itself—loses much of its importance. But I’m genuinely worried about the transition. Rapid capability growth, biological applications, recursive self-improvement, unequal access, and unprepared institutions could make the path extremely dangerous. So I’m optimistic about where AI could take us, not complacent about how we get there.
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