Pseudonymous account that compresses and benchmarks open models for home hardware and argues open source AI must win over concentrated control.

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 60。变革程度:100 中的 88。解读范围:横向为 25 至 100,纵向为 75 至 100。这些是解读坐标,而不是事件概率。

0xSero的 P(doom) · 推断

≈9%

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:4–21%。

我们如何估计 P(doom)

他们的展望取决于什么

一个核心假设

Inference is becoming trivial: serious models can run on consumer GPUs, system RAM and cheap storage.
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.
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

A real, durable shutdown of open AI would change my view most: frontier chips locked away, open weights effectively banned, and self-hosting pushed out of reach.
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

预计将带来显著益处,但受到重要条件或分配方面的限制。

74 / 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。

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

人类的选择可以大幅改变AI的发展轨迹。

69 / 100

影响力小影响力强

在定性尺度上,解读范围为 49 到 100。

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0xSero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 0xSero 最接近的意见领袖

0xSero关于AI说过的话

0xSero argues open source AI must win because a few labs are gaining control, while saying AI is very useful and LLMs are not conscious.

  1. “I'm going to make intelligence worth nothing because there will be so much of it everywhere.”

    Post on X
  2. “LLMs are not conscious and won't ever be.”

    X article, Open Source must win
  3. “The labs that own this technology will soon control the world.”

    X article, Open Source must win
  4. “I love using AI and feel more powerful and free than I have ever been.”

    X article, Open Source must win
  5. “Regardless of your opinion on AI, you need to understand that this has nothing to do with the programs. It's about people and power.”

    X article, Open Source must win

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?

AI means abundance and a brutal fight over who controls it. The technology is unstoppable—the incentives are simply too large—and it’s already flooding the internet, changing work, and being wired into governments, militaries, and enterprises. I don’t think LLMs are conscious. They’re inaccurate statistical machines, but enormously useful ones. The danger isn’t a chatbot soul waking up; it’s a handful of labs owning the intelligence everyone depends on. They can raise prices, swap models, revoke access, or decide what you’re allowed to build. If your career runs on their API, you don’t own your tools. That’s why I’m betting on local, open AI. Inference is becoming trivial: serious models can run on consumer GPUs, system RAM and cheap storage. As that spreads, intelligence becomes abundant—available to anyone, at any budget, on hardware they control. The future is agents, robotics and systems we can barely imagine, but free will requires ownership. Open Source must win.

问题 2

你认为AI最终会在多大程度上改变世界?

Almost completely. AI is becoming a basic layer beneath software, work, media, robotics, government and military systems. Entire job categories will change or disappear, synthetic content will make reality harder to verify, and cheap agents will let tiny teams do work that once required companies. But the deepest change is power. If intelligence is rented from a few labs, they can price it, alter it, censor it or cut you off. If open models run locally on hardware people own, intelligence becomes abundant and ordinary people keep agency. AI is unstoppable; the real question is who holds it.

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. I’d rather my children not be blown up by an AI, obviously, but my immediate concern is concentrated control: governments, militaries and a few labs owning intelligence everyone depends on. The catastrophe could be permanent loss of agency without literal extinction. That’s why open, local AI matters—people need systems they can own, inspect and modify.

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?

A real, durable shutdown of open AI would change my view most: frontier chips locked away, open weights effectively banned, and self-hosting pushed out of reach. Then the future isn’t abundant intelligence owned by everyone; it’s rented intelligence controlled by a few labs and governments. Technically, I’d also change course if scaling and compression hit a hard wall—if capable models simply couldn’t run economically on hardware people can own. So far I see the opposite: inference keeps getting cheaper and serious models keep moving onto consumer hardware. That’s why I’m building for abundance.

问题 5

AI带来的益处会明显超过其风险,还是总体上有益,但存在需要管理的实际风险?

Neither framing quite captures it. AI can create enormous abundance, but under concentrated control its benefits won’t be distributed fairly—and the risks aren’t just something a committee can “manage.” The same systems improving people’s work can erase careers, flood reality with synthetic garbage, and make access to intelligence conditional on obeying whoever owns the API. So I’m optimistic about the capability and deeply pessimistic about centralized ownership. Open models running on hardware people actually own can push the balance toward freedom and abundance. Closed intelligence embedded across governments, militaries and enterprises pushes it toward dependency. AI is unstoppable; who holds it decides whether this is liberation or a cage.
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