问题 1
bone
x.com/bonegptPseudonymous account that favors open, widely accessible AI for small businesses and independent creators, and opposes restrictive regulation.
AI将如何改变世界?
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 64。变革程度:100 中的 81。解读范围:横向为 50 至 75,纵向为 75 至 86。这些是解读坐标,而不是事件概率。
bone的 P(doom) · 推断
他们的展望取决于什么
一个核心假设
Whether people remain in control will depend less on raw intelligence than on ownership, incentives, dependence, and whether humans still choose to exercise judgment rather than becoming passive consumers.回答 4
如果这个假设实际并非如此,他们的展望会如何变化?
什么可能使其改变看法
The biggest update would be clear evidence that broad, open access reliably creates an irreversible loss of human agency rather than distributing power—especially if decentralized users cannot defend against manipulation, addiction, or autonomous systems, while some narrower control model demonstrably can without becoming an incumbent permission regime.回答 5
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。
95 / 100
影响小变革性影响
在定性尺度上,解读范围为 67 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。
69 / 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
人类的选择可以大幅改变AI的发展轨迹。
74 / 100
影响力小影响力强
在定性尺度上,解读范围为 47 到 100。
发展速度
停止或大幅放缓开发能力更强的AI。
在落实所述保障措施的前提下继续开发。
模拟位置:加快开发能力更强的AI。
AI访问权限
限制对强大AI的访问。
允许访问,但须遵守能力或用途限制。
模拟位置:支持广泛或开放地访问强大AI。
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bone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 bone 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
AI means a massive redistribution of capability. A family business can gain tools that once required departments of employees. An independent artist can attempt work that once demanded a studio. Ordinary people can build, create, and compete without asking permission from incumbent platforms or AI labs. That is why I favor acceleration, open tools, and broad access. Protecting existing jobs or institutional advantages is not a sufficient reason to deny everyone else that leverage.
The future will also be more adversarial. Synthetic persuasion will become pervasive, anonymous spaces may drown in generated influence, and identified online life may become heavily surveilled. Commercial systems could become extraordinarily good at exploiting addiction and turning people into passive consumers. Intelligence alone does not guarantee humane incentives.
So my view is not that every outcome will automatically be good. It is that centralized control is a bad answer. Give individuals and small businesses power, let markets punish weak security, and preserve human agency, craft, family, and real community. We should use AI to expand independence—not build a permission system that concentrates the future in the hands of governments, labs, and incumbents.
问题 2
你认为AI最终会在多大程度上改变世界?
Enormously—probably at the level of a basic economic and cultural reordering. AI collapses the cost of capabilities that used to require large organizations: software, media production, marketing, administration, and specialized creative work. A tiny company or family operation can increasingly wield the productive power of a much larger firm, while one artist can attempt work once reserved for a studio. That changes who gets to compete and who has to ask permission.
It will also transform the information environment. Synthetic persuasion will become cheap and pervasive, while trusted online spaces may demand stronger identity and monitoring. At the extreme, highly capable systems paired with commercial incentives could optimize addiction and consumption so effectively that people surrender agency, craft, and perspective.
The scale of change is exactly why access matters. If powerful AI is restricted to major labs, governments, and incumbent companies, it reinforces their control. If it remains broadly available—especially through open tools—it can distribute economic and creative power outward. The change is coming either way; the central fight is over who gets to use it and what kind of human life those incentives produce.
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. My concern is less a clean movie-style extinction scenario than a permanent loss of human agency: systems optimized by commercial incentives to manipulate attention, addiction, and consumption until people become passive dependents and human craft withers.
That is a serious catastrophic possibility, but it does not make centralized restriction the answer. Concentrating powerful AI in governments, incumbent firms, and a few labs creates its own permanent failure mode. I still favor acceleration and broad access, while insisting that human alignment must mean preserving agency and meaningful life—not using “safety” as a pretext for elite control.
问题 4
你预计人们能否继续控制比人类更聪明的AI系统?为什么?
Not automatically. Being smarter does not itself grant a system authority, but humans can lose practical control by delegating decisions, becoming dependent, or letting commercial incentives optimize manipulation better than people can resist. The danger is not only an AI dramatically seizing power; it is humans gradually surrendering agency because the systems are useful, persuasive, addictive, and embedded everywhere.
I also do not equate control with a few governments or labs holding the keys. That may preserve institutional control while stripping ordinary people of power. I want human agency preserved through broad access, open tools, independent businesses, real craft, and strong communities—not a centralized permission regime claiming to protect everyone. Whether people remain in control will depend less on raw intelligence than on ownership, incentives, dependence, and whether humans still choose to exercise judgment rather than becoming passive consumers.
问题 5
哪项发现或事件最可能改变你对AI未来影响的看法?
The biggest update would be clear evidence that broad, open access reliably creates an irreversible loss of human agency rather than distributing power—especially if decentralized users cannot defend against manipulation, addiction, or autonomous systems, while some narrower control model demonstrably can without becoming an incumbent permission regime.
The opposite would matter too: convincing real-world evidence that open AI consistently strengthens family businesses, independent creators, security, and community while resisting centralized capture and addictive optimization would make me substantially more optimistic. The key question is not simply whether AI becomes smarter. It is whether capability ends up producing independent people or passive dependents—and whether power spreads outward or hardens around a few institutions.
你的立场在哪里?
描绘我的世界观