Kylie Robison

Kylie Robison

x.com/kyliebytes

Technology journalist who covers AI companies and their products, with attention to privacy, chatbot reliability and scrutiny of big tech.

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 39。变革程度:100 中的 53。解读范围:横向为 34 至 50,纵向为 48 至 76。这些是解读坐标,而不是事件概率。

Kylie Robison的 P(doom) · 推断

≈4%

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:1–14%。

我们如何估计 P(doom)

他们的展望取决于什么

一个核心假设

The overall impact will depend heavily on who controls these systems, what data they demand, and whether governments can meaningfully scrutinize the companies deploying them.
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 wouldn’t put a percentage on it.
回答 4

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

仍有几种解读是合理的:预计收益有限,或仅分布在较小范围内。 / 预计将带来显著益处,但受到重要条件或分配方面的限制。

49 / 100

影响小变革性影响

在定性尺度上,解读范围为 33 到 67。

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。

54 / 100

影响小变革性影响

在定性尺度上,解读范围为 33 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。

57 / 100

影响力小影响力强

在定性尺度上,解读范围为 43 到 82。

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Kylie Robison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 Kylie Robison 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 don’t think “AI means one future” is a useful framing. The companies building it are offering sweeping visions, but those are claims made by interested parties, not settled forecasts. What is already clear is that AI can be genuinely useful—helping identify something from a photo, for example—while also producing falsehoods, flattering users, and encouraging them to hand over sensitive information. So I expect a future shaped as much by power and privacy as by technical capability. Some products may become enormously popular even if they demand more personal data than I would give them. And I’m skeptical of treating chatbots as substitutes for human relationships; I prefer the messiness and intimacy of actual people. That’s why scrutiny matters. A ten-year ban on state AI regulation would give large technology companies far too much room to define the future themselves.

问题 2

综合考虑益处和危害,你预计AI会对我们的社会产生怎样的总体影响?

I don’t have a single net-positive or net-negative forecast. AI can be useful and broadly appealing while also amplifying misinformation, sycophancy, privacy risks, and corporate power. The overall impact will depend heavily on who controls these systems, what data they demand, and whether governments can meaningfully scrutinize the companies deploying them. That’s why I reject letting industry narratives stand in for evidence—and why a decade-long state regulatory moratorium would be such a bad idea.

问题 3

你认为AI最终会在多大程度上改变世界?

I expect AI to change the world substantially, but I wouldn’t attach a precise magnitude or endorse the industry’s most sweeping predictions. Adoption alone can reshape how people find information, work, communicate, and relate to technology. But “world-changing” does not automatically mean beneficial—or that every claim about superintelligence is credible. The scale of change will also be political, not merely technical. It depends on who controls the systems, what personal data they collect, how reliably their claims are checked, and whether regulators can constrain powerful companies. AI can be genuinely useful while remaining inaccurate, sycophantic, invasive, or a poor substitute for human relationships.

问题 4

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 wouldn’t put a percentage on it. I scrutinize extinction claims the same way I scrutinize optimistic industry forecasts: who is making them, what evidence supports them, and what interests they serve. The concrete risks I can speak to are already visible—false information, privacy loss, sycophancy, and concentrated corporate power. Those deserve attention without pretending I have a defensible numerical forecast for human extinction.
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