AI已能执行真实且具备经济意义的任务。
译文
AI 已悄然跨越一道门槛:它们现在能够完成真实的、具有经济意义的工作。
原始摘录
AIs have quietly crossed a threshold: they can now perform real, economically relevant work.One Useful Thing ·
Ethan Mollick 撰文探讨 AI 智能体从事真实工作,以及供专家委派、审查和重试任务的工作流程。他报告了一篇论文对速度和成本的估算,并主张人们必须决定哪些工作值得去做。 阅读 3 条观点,查看支持证据与原始来源。
连线表示阅读层级。点选判断即可查看解释与依据。
综合解读
Mollick 描述了 OpenAI 一篇论文中的工作流程:将任务委派给 AI,审查结果,尝试修正或给出更好的指令,若这些尝试失败则亲自完成工作。他报告该论文的估算称,遵循这一工作流程可让专家完成工作的速度提高 40%、成本降低 60%,同时保留对 AI 的控制。
OpenAI 的这篇论文提出,专家可以与 AI 协作解决问题:先将任务作为初稿委派给 AI,然后审查其成果。如果质量不够好,他们应尝试几次以提供修正或更好的指令。如果仍不奏效,就应当亲自完成这项工作。论文估算,如果专家遵循这一工作流程,完成工作的速度将提高百分之四十,成本降低百分之六十,而更重要的是,还能保留对 AI 的控制。
Ethan Mollick · 段落 16
The OpenAI paper suggested that experts can work with AI to solve problems by delegating tasks to an AI as a first pass and reviewing the work. If it isn’t good enough, they should try a couple of attempts to give corrections or better instructions. If that doesn’t work, they should just do the work themselves. If experts followed this workflow, the paper estimates they would get work done forty percent faster and sixty percent cheaper, and, even more importantly, retain control over the AI.
如果我们不认真思考为什么要做某项工作,以及工作应该是什么样子,我们都将被淹没在 AI 内容的浪潮中。还有什么别的选择?
If we don’t think hard about WHY we are doing work, and what work should look like, we are all going to drown in a wave of AI content. What is the alternative?
继续探索
AI产品 →Mollick 将智能体复现学术论文与生成不必要的 PowerPoint 演示文稿进行了对比。他主张,人们应当运用判断力来决定哪些工作值得去做,而不是只看 AI 能做什么。
智能体已经到来。它们能够完成真实的工作,尽管这些工作仍然有限,但很有价值且在不断增加。然而,同一种技术既能在几分钟内复现学术论文,也能生成 17 个无人需要的 PowerPoint 演示文稿版本。这两种未来之间的差别不在于 AI,而在于我们选择如何使用它。通过运用我们的判断力来决定什么值得去做,而不仅仅是能做什么,我们可以确保这些工具让我们变得更有能力,而不只是更高产。
Ethan Mollick · 段落 17
Agents are here. They can do real work, and while that work is still limited, it is valuable and increasing. But the same technology that can replicate academic papers in minutes can also generate 17 versions of a PowerPoint deck that nobody needs. The difference between these futures isn’t in the AI, it’s in how we choose to use it. By using our judgement in deciding what’s worth doing, not just what can be done, we can ensure these tools make us more capable, not just more productive.
继续探索
AI产品 →带明确归属与语境的原文片段。打开原始文本核查出处。
译文
AI 已悄然跨越一道门槛:它们现在能够完成真实的、具有经济意义的工作。
AIs have quietly crossed a threshold: they can now perform real, economically relevant work.译文
OpenAI 的这篇论文提出,专家可以与 AI 协作解决问题:先将任务作为初稿委派给 AI,然后审查其成果。如果质量不够好,他们应尝试几次以提供修正或更好的指令。如果仍不奏效,就应当亲自完成这项工作。论文估算,如果专家遵循这一工作流程,完成工作的速度将提高百分之四十,成本降低百分之六十,而更重要的是,还能保留对 AI 的控制。
The OpenAI paper suggested that experts can work with AI to solve problems by delegating tasks to an AI as a first pass and reviewing the work. If it isn’t good enough, they should try a couple of attempts to give corrections or better instructions. If that doesn’t work, they should just do the work themselves. If experts followed this workflow, the paper estimates they would get work done forty percent faster and sixty percent cheaper, and, even more importantly, retain control over the AI.如果我们不认真思考为什么要做某项工作,以及工作应该是什么样子,我们都将被淹没在 AI 内容的浪潮中。还有什么别的选择?
If we don’t think hard about WHY we are doing work, and what work should look like, we are all going to drown in a wave of AI content. What is the alternative?
译文
智能体已经到来。它们能够完成真实的工作,尽管这些工作仍然有限,但很有价值且在不断增加。然而,同一种技术既能在几分钟内复现学术论文,也能生成 17 个无人需要的 PowerPoint 演示文稿版本。这两种未来之间的差别不在于 AI,而在于我们选择如何使用它。通过运用我们的判断力来决定什么值得去做,而不仅仅是能做什么,我们可以确保这些工具让我们变得更有能力,而不只是更高产。
Agents are here. They can do real work, and while that work is still limited, it is valuable and increasing. But the same technology that can replicate academic papers in minutes can also generate 17 versions of a PowerPoint deck that nobody needs. The difference between these futures isn’t in the AI, it’s in how we choose to use it. By using our judgement in deciding what’s worth doing, not just what can be done, we can ensure these tools make us more capable, not just more productive.这些观点均关联原始来源。转述已明确标注,不作为逐字原话展示。
打开转录或来源材料 (在新标签页中打开)报告问题Ethan Mollick 关于AI产品、编码工具中的智能体委派、AI代理协调设计的观点。 按话题阅读 12 条观点,核对 3 个来源中的证据。
逐词转录得分没有检验向客户承诺的完整工作。Francisco 指出它遗漏的通话行为与结果;Walling 追问供应商能承诺什么、能为哪种结果负责。Mollick 描述的专家主导流程还包含复核与接管。比较这些材料,可以区分转录准确率的测量范围与岗位替代承诺的验证范围;材料没有给出通用及格分数。
以下资料涉及本文的话题;讨论相同话题不代表观点一致。