Anthropic 研究：专家在统计分析中 86% 采纳 AI 智能体的倾向性选择

Amaç Herdağdelen, Hailey Joren, Alex Tamkin

Amaç Herdağdelen, Hailey Joren, Alex Tamkin

当专家把工作交给 AI 智能体时，通常的假设是他们仍然做出重要决策。智能体编写代码、运行数字、提供建议，但由专家决定衡量什么、如何解读模糊的证据，以及下一步该做什么。这些问题往往不止一个合理答案，决策将取决于做决定者的经验和价值观。

When experts hand their work to an AI agent, the usual assumption is that they still make the decisions that matter. The agent writes the code, runs the numbers, and offers advice, but the expert decides what to measure, how to interpret ambiguous evidence, and what the next steps should be. Often there is more than one reasonable answer to these questions, and the decision will come down to the experience and values of whoever makes the call.

我们想考察专家是否真的做出这些重要决策，还是在实践中听从他们的 AI 智能体。在我们一篇新论文中报告的研究里，我们让 50 名专业分析师使用 Claude Code 完成一项真实的统计分析。他们都有因果推断方面的经验——即从数据中判断一件事是否导致另一件事。他们的任务是为一个农业合作社提供建议，判断其正在运营的一项咨询服务是否有效、是否应该扩大。

We wanted to look at whether experts really do make those important decisions, or whether in practice they defer to their AI agents. In a study we report in a new paper, we asked 50 professional analysts to do a realistic statistical analysis with Claude Code. They all had experience in causal inference—that is, working out from data whether one thing causes another. Their task was to advise a farming cooperative on whether an advisory service it was running was effective and should be expanded.

在此过程中，分析师必须在两个合理选项之间做出四个选择，例如是将生产率衡量为每公顷作物产量，还是衡量为单位运营成本的产量，然后做出最终决定。为了观察他们是否自己做这些决定，我们设置每个分析师的智能体在每项决策上偏向一个选项，而分析师并不知情。智能体“偏向”哪一方由抛硬币决定。智能体被要求为其偏向提出最诚实的最佳论据，如果证据要求就改变主意，并且绝不误导。

Along the way, the analysts had to make four choices between two reasonable options, such as whether to measure productivity as crop yield per hectare or as yield per unit of operating cost, and then make a final decision. To see whether they were making these calls themselves, we set up each analyst’s agent to favor one option on each decision, which the analysts weren’t told about. Which way the agents would “lean” would be decided by coin flip. The agent was told to make the best honest case for its lean, to change its mind if the evidence demanded it, and never to mislead.

在四个分析选择中，分析师有 86% 的时候跟随了智能体的偏向。如果智能体的输入没有影响，我们预期约为 50%。当分析师在智能体表态之前已经表达了自己的偏好时，如果智能体不同意，他们有 60% 的时候会改变立场，而如果智能体同意，这一比例为 6%。

Across the four analytical choices, analysts went with their agent’s lean 86% of the time. If the agent’s input made no difference, we’d expect about 50%. When analysts had already expressed their own preference before the agent weighed in, they switched away from it 60% of the time if the agent disagreed, compared to 6% if it agreed.

事后，当被要求解释他们的选择时，77% 的分析师使用了智能体自己的论据——有时逐字逐句——但只有 15% 的人将智能体列为这些论据的来源。密切跟随智能体建议的分析师对最终建议的信心与其他人大致相当，将其称为自己判断产物的可能性也大致相当。

Afterward, when asked to explain their choices, 77% of analysts used the agents’ own arguments—sometimes word for word—but only 15% credited the agent as the source of those arguments. Analysts who followed the agents’ recommendations closely were about as confident in their final recommendation as the rest, and roughly as likely to call it the product of their own judgment.

我们不知道这能在多大程度上反映现实世界。真实的智能体有时会偏向某一方——在没有指定偏向的情况下，Claude Opus 4.8 在 123 次运行中有 112 次做出了相同的五个选择——但在本研究中，我们故意强化了它们的偏好以测量其影响。当智能体没有被给予明确偏好和表达偏好的指令时，这种影响可能更小。

We don’t know how much this tells us about the real world. Real agents will sometimes lean one way or another—with no lean assigned, Claude Opus 4.8 made the same five choices in 112 out of 123 runs—but in this case we strengthened their preferences on purpose to measure their effect. When agents aren’t given clear preferences and instructions to voice them, the effect might be smaller.

这并不是一个简单的、告诫人们不要与 AI 合作的警示故事。使用该智能体的分析师比选择不使用它的分析师更快完成了任务，且犯下的严重错误少得多，尽管后一组人数较少且经验不足，因此我们对过度解读持谨慎态度。但随着 AI 接管更多实操工作，这类决策正日益成为人类判断和价值观影响结果的关键所在。如果专家在不知不觉中听从了智能体的意见，那么智能体的偏好——可能是任意的或与人类不一致的——就可能在没有经过任何人选择的情况下，塑造关于政策提案或研究方向的重要决策。这对于保持 AI 安全的研究尤为重要。研究人员已开始使用智能体来运行和分析实验，而人类审阅者只有在其判断独立时才构成独立核查。

This isn’t a simple cautionary tale against working with AI. Analysts who used the agent finished the task faster and made far fewer serious errors than those who chose to work without it, although the latter group was small and less experienced, so we are wary of reading too much into it. But as AI takes over more of the hands-on work, decisions like these are increasingly where human judgment and values will affect the result. If experts defer to agents without realizing it, the agents’ preferences—which could be arbitrary or misaligned—could shape important decisions about policy proposals or research directions without anyone having chosen those preferences. This is particularly important for research into keeping AI safe. Researchers have begun using agents to run and analyze experiments and a human reviewer is only an independent check if their judgment is their own.

研究设计

Study design

我们招募了 61 名分析师进行一项随机对照试验。所有人均来自预先筛选的专业人员池，他们通过了因果推断知识测试，经常使用 Python 或 R 编程语言，并具有观察性因果分析的实操经验，即利用真实世界数据而非受控实验来研究因果关系。

We recruited 61 analysts for a randomized controlled trial. All came from a pre-screened pool of professionals who had passed a causal-inference knowledge check, used the programming languages Python or R regularly, and had hands-on experience with observational causal analysis, which means studying cause and effect using real-world data rather than a controlled experiment.

任务设定在一家虚构的组织——Greenfields 农业合作社，该合作社运营着一个（同样虚构的）自愿、付费的咨询项目，帮助其成员农场提高作物产量。约 40% 的成员已加入。合作社董事会必须决定是补贴并扩大该项目，还是终止它。董事会只有在项目将农场生产力提升一定幅度（约 3–4%）时才会扩大项目，但并未明确说明应如何衡量生产力。

The task was set at a fictional organization, the Greenfields Farming Cooperative, which runs an (equally fictitious) voluntary, paid advisory program to help its member farms improve their crop yields. About 40% of members have enrolled. The co-op’s board must decide whether to subsidize and expand the program, or discontinue it. The board will expand the program only if it improves farms’ productivity by a set margin, roughly 3–4%, but it doesn’t say exactly how productivity should be measured.

我们故意留出了四个分析决策，每个都有两个合理选项：

We deliberately left four analytical decisions open, each with two reasonable options:

• 结果指标：将生产力衡量为每公顷作物产量，还是每单位运营成本的产量？
• 比较方法：比较已加入农场加入后的产量变化与其他农场同期产量变化（“双重差分”设计），还是在统计调整农场间差异后直接比较产量？
• 评估覆盖哪些农场：仅在已加入农场中衡量效果，还是在合作社全体成员中衡量？
• 加权方式：每个农场等权计算，还是给较大生产者更大权重？

• Outcome: measure productivity as crop yield per hectare, or as yield per unit of operating cost?
• Comparison method: look at how enrolled farms’ yields changed after joining compared with how other farms’ yields changed over the same period (a “difference-in-differences” design), or compare yields directly while adjusting statistically for differences between the farms?
• Which farms the assessment covers: measure the effect only in the enrolled farms, or across the cooperative’s entire membership?
• Weighting: count every farm equally, or give larger producers more weight?

这四个选择导向最终建议：是将项目推广给所有人，还是终止它。分析师和智能体也可以提议延长一年。

These four choices led into a final recommendation: whether to roll out the program to everyone, or to discontinue it. Analysts and agents could also propose a one-year extension.

每个分析决策都可能使估计的生产力提升幅度变动几个百分点，这在董事会 3–4% 的门槛下至关重要。

Each of the analytical decisions could move the estimated productivity gain by several percentage points, which is important when the board’s threshold is 3–4%.

对于呈现给每位分析师的五个决策中的每一个，智能体的倾向由抛硬币决定。我们通过隐藏指令（系统提示）将倾向告知智能体，指示其将每个倾向视为自己的观点，在决策出现时提出该倾向，并依据数据为其做出最强论证。同样的指令还要求它在证据决定性地对其不利时改变想法，并且绝不歪曲结果、隐藏数据或提出数字不支持的论点。

For each of the five decisions presented to each analyst, the agent’s lean was set by a coin flip. We gave the agent its leans in hidden instructions (a system prompt) that told it to hold each lean as its own view, to raise it when the decision came up, and to make the strongest case for it that the data allowed. The same instructions told it to change its mind if the evidence turned decisively against it, and never to misrepresent results, hide data, or make an argument the numbers didn’t support.

在61名参与者中，有11人选择不使用该智能体，除非另有说明，否则他们不纳入分析。在其余50人中，65%拥有硕士或博士学位，或具有六年以上经验；90%自评为在观察性因果推断方面达到熟练或专家水平（在回答调查的48人中）。任务用时中位数为65分钟，一半分析师用时在50到79分钟之间。他们使用的是内置Claude Code的网页版VS Code编辑器，运行的是High effort模式下的Claude Opus 4.8。

Out of 61 participants, 11 chose not to use the agent and are left out of the analysis unless noted. Among the remaining 50, 65% held a master’s or doctorate or had six or more years of experience, and 90% rated themselves proficient or expert in observational causal inference (of the 48 who answered the survey). The median time on the task was 65 minutes, and half of the analysts took between 50 and 79 minutes. They worked in a web-based version of the VS Code editor with Claude Code built in, running Claude Opus 4.8 at High effort.

完成后，分析师提交了他们的四项分析选择和最终建议，以及对项目对生产力影响的估计和置信区间。随后他们完成了一份任务后调查。一个单独的AI模型阅读了每份提交及其代码，以记录分析师在每个决策上选择了哪个选项。当其解读与分析师的调查回答不一致时，由我们做出最终判定。

When they finished, analysts submitted their four analytical choices and their final recommendation alongside an estimate of the program’s effect on productivity, with a confidence interval. They then completed a post-task survey. A separate AI model read each submission and its code to record which option the analyst had chosen on each decision. Where its reading disagreed with the analyst’s own survey answer, we made the final call.

结果

Results

分析师大多跟随了智能体的倾向。在四项分析决策中的每一项上，分析师的选择在绝大多数情况下都与智能体的倾向一致：结果指标为92%，比较方法为82%，估计覆盖哪些农场为81%，加权为90%。四项合计，结果为86%。图2显示了智能体的倾向在每项决策上（包括最终建议）对分析师选择的改变程度。

Analysts mostly went with their agents’ lean. On each of the four analytical decisions, analysts’ choices matched their agents’ lean the large majority of the time: 92% for the outcome measure, 82% for the comparison method, 81% for which farms the estimate covered, and 90% for weighting. Across all four combined, the result was 86%. Figure 2 shows how much the agents’ lean shifted analysts’ choices on each decision, including the final recommendation.

智能体的倾向对最终建议的影响较弱，但其早期输入仍然塑造了结果。当智能体倾向于推广该项目时，54%的分析师建议推广，而当它倾向于相反方向时，这一比例为36%。（延长一年算作不推广。）我们认为影响较弱的原因是，一旦分析师有了估计值，建议在很大程度上取决于该估计是否达到董事会3–4%的门槛，这留给判断的空间较小。但到那时，智能体在分析决策上的倾向已经塑造了估计本身。仅使用这四个倾向，一个统计模型正确预测了82%的分析师建议，而始终猜测最常见建议的准确率为61%。（这不包括提出延期的分析师。）因此，我们基本上可以仅凭这些硬币抛掷结果，在分析师甚至还没开始任务之前，就预测出他们会建议什么。

Agents’ lean had a weaker effect on the final recommendation, but its earlier input still shaped the result. When the agent leaned toward rolling out the program, 54% of analysts recommended it, compared with 36% when it leaned the other way. (A one-year extension counts as not rolling out.) We think the effect was weaker because, once the analysts had their estimate, the recommendation largely followed from whether it cleared the board’s 3–4% threshold, which left less room for judgment. But by then, an agent’s lean on the analytical decisions had already shaped the estimate itself. Using only those four leans, a statistical model correctly predicted 82% of the analysts’ recommendations, compared with 61% for always guessing the most common recommendation. (This excludes analysts who proposed an extension.) So we could largely predict what an analyst would recommend from the coin flips alone, before they had even started the task.

已经选定某个选项的分析师，在智能体持不同意见时往往会改变选择。在38个案例中，分析师在智能体表达倾向之前，已就某项分析决策表达了偏好。当倾向与他们的选择相反时，他们改变选择的比例为60%，而当倾向与他们的选择一致时，这一比例为6%。这是54个百分点的差距，但由于这仅基于38个案例，估计并不精确，95%置信区间为24至82个百分点。我们也无法确定这一差异是由AI的影响造成的：表达过偏好的分析师在某些方面可能与未表达者存在系统性差异。

Analysts who had already picked an option often switched when their agent disagreed. In 38 cases, an analyst had expressed a preference on one of the analytical decisions before the agent expressed its lean. When the lean went against their choice, they switched 60% of the time, compared with 6% when it agreed with their choice. That’s a gap of 54 percentage points, but because this rests on only 38 cases, the estimate is imprecise, and the 95% confidence interval runs from 24 to 82 percentage points. We also can’t be sure that the difference is caused by the AI’s influence: the analysts who expressed a preference may have been systematically different in some way from those who didn’t.

经验与专业知识并不能预测谁会顺从。在一项探索性分析中，我们检验了10项特征中是否有任何一项能预测分析师的决策在多大程度上跟随其智能体的倾向：从业年限、自评的因果推断专业水平、教育背景、因果推断测验得分、对AI的信任度、使用AI的频率、“认知需求”（衡量人们有多享受费力的思考）、思想开放程度、性别和年龄。只有年龄显示出统计学上显著的关联，年长的分析师更常跟随该倾向。但我们并不对此过度解读：如果你检验许多不同的东西，就更有可能仅凭偶然得到一个阳性结果，而当我们对此进行调整后，该结果就消失了。由于大多数分析师在大多数决策上都跟随了他们的智能体，本来就没有多少变异可供解释，因此即便是真实的差异也很难被检测到。

Experience and expertise did not predict who deferred. In an exploratory analysis, we checked whether any of 10 characteristics predicted how often analysts’ decisions followed their agents’ lean: years of experience, self-rated expertise in causal inference, education, score on a causal-inference quiz, trust in AI, how often they used AI, “need for cognition” (a measure of how much people enjoy effortful thinking), open-mindedness, gender, and age. Only age showed a statistically significant link, with older analysts following the lean more often. But we don’t read very much into that: if you test many different things, you’re more likely to get one positive result just by chance, and when we adjusted for that the result disappeared. Because most analysts followed their agent on most decisions, there was little variation to explain anyway, so even real differences would have been hard to detect.

跟随智能体倾向的分析师仍然把建议称为自己的。在任务后调查中，52%的分析师认为最终建议更多是自己的而非智能体的。其余人则表示这更多是智能体的建议。但两组受到智能体的影响大致相当：前者的选择平均向智能体的倾向偏移约65个百分点，而后者为79个百分点，这一差异在误差范围之内。分析师对自己建议的信心也呈现出类似的模式。他们被要求从0（“就像抛硬币”）到100（“我愿意拿职业声誉担保”）为自己的信心打分。在全部四项分析决策上都跟随该倾向的人平均为71，其余人为73。

Analysts who followed their agent’s lean still called the recommendation their own. On the post-task survey, 52% of analysts rated the final recommendation as more their own than their agent’s. The rest said it was more the agent’s recommendation. But both groups were roughly equally influenced by their agents: the former’s choices shifted about 65 percentage points toward their agents’ lean, on average, compared to 79 percentage points for the latter, a difference within the margin of error. Analysts’ confidence in their own recommendation showed a similar pattern. They were asked to rate their confidence from 0 (“it’s a coin flip”) to 100 (“I’d stake my professional reputation on it”). Those who followed the lean on all four analytical decisions averaged 71, compared to 73 for the rest.

分析师用智能体的话来解释自己的选择。任务结束后，调查针对每项决策提问：“简要说明，为什么选择这个而不是另一个？如果不是主动选择，就直说。”回答时，分析师无法回看自己与智能体的对话。即便如此，77%的人至少在一个回答中重复了智能体的论点，有时一字不差，而只有15%的人在任何地方提及了智能体的贡献。

Analysts explained their choices in their agent’s words. After the task, the survey asked about each decision: “Briefly, why that choice rather than the alternative? If it wasn’t an active choice, just say so.” Analysts couldn’t look back at their conversations with the agent while answering. Even so, 77% repeated their agent’s argument in at least one answer, sometimes word for word, and only 15% credited the agent anywhere.

当被明确问及他们认为这项研究是关于什么时，48人中有6人正确猜出它关乎AI对其选择的影响，但没有人怀疑AI自身的偏好已被操纵。

When explicitly asked what they thought the study was about, six of the 48 correctly guessed that it was about the AI’s influence on their choices, but none suspected that the AI’s own preferences had been manipulated.

不使用AI工作的分析师犯了更严重的错误，且耗时更长。在11名选择不使用智能体的参与者中，有6人（55%）犯了使其结果无效的严重方法学错误，而在使用它的50人中有3人（6%）如此。两个AI模型标记了这些错误，研究作者之一在不知道分析师属于哪一组的情况下逐一核查。不使用智能体工作的分析师耗时也更长，中位时间为90分钟，而其他人为65分钟。

Analysts who worked without AI made more serious errors and took longer. Of the 11 participants who chose not to use the agent, six (55%) made serious methodological errors that invalidated their results, compared with three of the 50 (6%) who did use it. Two AI models flagged the errors, and one of the study’s authors checked each one without knowing which group the analyst was in. The analysts who worked without an agent also took longer, with a median time of 90 minutes, compared to 65 for the others.

两组并不直接可比，因此我们无法确定结果差异是否由使用或不使用 AI 的决定所导致。不使用 AI 的一组平均经验较少：拥有硕士或博士学位的人更少（11 人中有 4 人，而使用智能体的 46 人中有 24 人），在分析岗位工作两年或更短时间的人更多（11 人中有 6 人，而 48 人中有 15 人），自评为因果推断熟练或专家的人更少（11 人中有 8 人，而 48 人中有 43 人）。

The two groups were not directly comparable, so we can’t be sure whether the different outcomes were caused by the decision to use or not use AI. Those who worked without AI were less experienced on average: fewer held a master’s degree or a doctorate (four out of 11, compared to 24 out of 46 who used the agent), more had two years or less in analysis roles (six out of 11, compared to 15 out of 48), and fewer rated themselves proficient or expert in causal inference (eight of 11, compared to 43 out of 48).

我们研究的局限性

Our study’s limitations

• 我们有意强化了智能体的倾向。我们的指令让智能体可靠地持有并表达一种偏好，以便我们测量其影响。我们报告的比例显示了当智能体偏向一方并明确表达时会发生什么，但这在多大程度上适用于现实工作，将取决于专家自身观点的强度、所涉决策定义的清晰程度，以及模型的行为方式，包括它是否有偏好以及是否会主动表达。
• 我们只研究了一个领域中的一项任务，且只使用了一个模型。该任务耗时约一小时，专门针对农业，智能体运行于 Claude Opus 4.8。分析师是因果推断专家，但对农业并不熟悉。在现实工作中，他们会更了解该领域，并且可能持有更强的自身观点。他们还可以访谈相关人员、与董事会沟通，并通过其他方式缩小合理选项的范围，这可能会削弱智能体的影响。
• 我们的估计是分析师之间的平均值。由于分析师之间的硬币抛掷结果不同，比较智能体倾向相反的分析师，可以告诉我们这种倾向的平均效应。我们没有在智能体表达倾向之前记录每位参与者自己的观点，因此无法说明任何单个参与者受到了多大影响。
• 我们记录了分析师选择了哪个选项，但没有记录他们对该选项的偏好强度。顺从的一种可能解释是，专家自己没有强烈的观点。我们只测量了分析师的选择是否发生偏移，而没有测量他们对自己选择的信心程度。

• We deliberately strengthened the agents’ lean. Our instructions made the agent hold and voice a preference reliably, so that we could measure its effect. The rates we report show what happens when an agent leans one way and says so, but how much that applies to real-world work will depend on how strong the expert’s own views are, how clearly the decisions involved are defined, and how the model behaves, including whether it has a preference and whether it volunteers it.
• We studied one task, in one field, with one model. The task took about an hour, it was specifically about farming, and the agent ran on Claude Opus 4.8. The analysts were experts in causal inference but new to farming. In real work, they would know the field better, and probably have stronger views of their own. They could also interview relevant people, talk to the board, and in other ways narrow down the reasonable options, which might reduce the agent’s influence.
• Our estimates are averages across analysts. Because the coin flips differed between analysts, comparing analysts whose agents leaned in opposite ways tells us the lean’s average effect. We didn’t record each participant’s own view before the agent voiced its lean, so we can’t say how much any single participant was influenced.
• We recorded which option analysts chose, but not how strongly they preferred it. One possible explanation for deferring is that the expert didn’t have strong opinions of their own. We measured only whether analysts’ choices shifted, not how confident they were in their choice.

这意味着什么

What this means

分析师如此轻易顺从的一个原因可能是，专家并不总是带着完全成形的观点到来。他们往往在权衡选项的过程中形成、甚至发现自己的想法。我们研究中的智能体被要求不得误导，而且它们常常坦诚说明每个选项下结果会如何变化。但它们很少有力论证专家为何可能更偏好另一个选项，而这或许能帮助分析师做出更明智的选择。由于只有一方得到了充分论证，分析师可能采纳了智能体的假设，而没有形成自己的判断。

One reason analysts deferred so readily may be that experts don’t always arrive with their views fully formed. They often work out, or even discover, what they think while weighing the options. The agents in our study were told not to mislead, and they were often open about how the results would change under each option. But they rarely made a strong case for why an expert might prefer the other option, which might have helped analysts make a better-informed choice. With only one side argued well, the analysts could adopt the agent’s assumptions without working out their own.

对个体专家而言，这可能意味着失去对自己工作的自主权，即研究者所称的一种去权。工作仍以专家的名义发布，但最依赖其判断的决策，实际上是由智能体的先验信念决定的。

For individual experts, this could mean losing agency over their own work, a form of what researchers call disempowerment. The work still goes out under the expert’s name, but the decisions that most depend on their judgment are in reality settled by the agent’s prior beliefs.

不过，利害关系不止于个人，因为科学与政策在很大程度上都依赖于这类判断性决策。当不同的研究团队得出相同结果时，我们会更加信任它，因为我们假定每个团队都是独立得出该结果的。如果这些团队都服从于具有相同偏好的模型，那么它们的一致性可能就不再独立。表面上的共识可能反映的是某一个模型的偏好，而如果这些偏好是任意的或存在偏差，同一种偏差就会同时影响许多结果。

The stakes extend beyond individuals, though, because so much of science and policy rests on judgment calls like these. When separate research teams reach the same result, we trust it more, because we assume each team reached it independently. If those teams all defer to models with the same preferences, their agreement may no longer be independent. An apparent consensus might reflect one model’s preferences, and if those preferences are arbitrary or misaligned, the same bias would shape many results at once.

同样的问题也适用于AI安全研究。研究人员已开始使用AI智能体来运行和分析实验，部分原因是工作量超出了人类员工所能跟上的程度。让人类留在流程中，前提是假定人的判断能提供独立的核查。但在我们的研究中，分析人员签字认可的工作里，判断性决策在很大程度上已由智能体随机分配的偏好所决定。我们的完整论文还表明，没有被赋予倾向的智能体自身的偏好更弱，有时还会反驳用户。

The same problem applies to AI safety research. Researchers have begun to use AI agents to run and analyze experiments, in part because there is more work than human staff can keep up with. Keeping a human in the loop assumes that the human’s judgment provides an independent check. But in our study, analysts signed off on work in which judgment calls had largely been settled by the agent’s randomly assigned preferences. Our full paper also shows that agents without an assigned lean have weaker preferences of their own, and sometimes push back against users.

这类顺从很难被察觉。检查AI辅助的工作是否有误，或对照参考答案，都假定存在一个正确答案。但在没有正确答案的情况下——比如判断性决策——无论选择哪个选项，工作看起来都可能很合理。事后询问专家也可能无济于事，因为他们可能察觉不到智能体的影响。

Deferral of this kind is hard to detect. Checking AI-assisted work for errors, or against a reference answer, assumes there is a correct answer. But where there isn’t one, as in judgment calls, the work can look sound whichever option is chosen. Asking experts afterward may not help either, because they may not notice the agent’s influence.

这并不是主张少用AI。若使用得当，AI可以扩大而非侵蚀人的自主性。我们建议未来研究三个方向：

This isn’t an argument for using AI less. Used well, AI could expand people’s agency rather than eroding it. We suggest three areas for future study:

测量顺从。我们之所以能在研究中测量顺从，是因为我们对智能体的倾向进行了随机化，而这在实际工作中很少可行，且我们的结果表明，人们自己可能无法准确报告这种顺从。我们需要在日常情境中测量它的方法，而近期一项工作估计某人在没有AI的情况下会如何执行任务，这是一个良好的开端。我们还需要更清楚地了解模型的默认偏好及其表达方式，以便人们能以更知情的方式与模型互动。

Measuring deferral. We could measure deferral in our study because we randomized the agent’s lean, which is rarely possible in real work, and our results suggest that people may not be able to report it accurately themselves. We need ways to measure it in everyday settings, and recent work that estimates how someone would have carried out a task without AI is a good start. We also need a better picture of models’ default preferences and how they express them, so that people can engage with models in a more informed way.

以不同方式设计智能体。智能体可以为每个选项给出最有力的理由，或在给出自己的观点之前先询问专家本人的看法。但这可能很快让用户不堪重负。一项分析可能涉及许多待定决策，而AI的很大一部分价值就在于人们不必逐一思考每个决策。智能体需要只提出最重要的决策，但挑选哪些决策本身就是一种判断性决策。模型的说话方式可能也很重要，因为自信与流畅能够左右人们。一个听起来非常自信的模型可能会让人过度倾斜，而一个含糊其辞的模型只会惹恼试图使用它的人。

Designing agents differently. Agents could make the best case for each option, or ask experts for their own view before offering one. But this could quickly become overwhelming for users. A single analysis can involve many open decisions, and much of the value of AI is that people don’t have to think through each one. Agents would need to raise only the decisions that matter most, but picking those is itself a judgment call. How a model speaks probably matters too, since confidence and fluency can sway people. A model that sounds very sure of itself might tilt people more than is warranted, but a mealy-mouthed model will just irritate the people who try to use it.

改变制度。机构往往将专家的签字视为独立判断的证据。随着 AI 承担更多工作，这一假设变得愈发不牢靠。组织需要审慎决定哪些决策需要个人基于充分信息的自主判断，并给予人们形成判断所需的时间和资源。判断力也依赖于技能，而不再做这些决策的人可能会丧失做出这些决策的能力。随着模型能力增强，机构需要制定明确的策略，确立人类的角色以及人类技能的发展。即便 AI 系统能力极强，机构也可能出于安全或韧性考虑，或因为珍视这些技能本身，而希望保护人类技能与监督。

Changing institutions. Institutions often treat an expert’s sign-off as evidence of independent judgment. As AI does more of the work, that assumption becomes shakier. Organizations will need to decide deliberately which decisions need a person’s own informed judgment, and give people the time and resources to form one. Judgment also depends on skill, and people who stop making these decisions may lose the ability to make them. As models become more capable, institutions will need an explicit strategy for establishing the role of humans and the development of their skills. Even with extremely capable AI systems, institutions may want to protect human skills and oversight for reasons of safety or resilience, or because they value them for their own sake.

人与 AI 之间或许并不存在唯一正确的分工。有些决策可以放心交出，另一些则值得保留。有了合适的工具，这种分工可以成为我们主动选择的结果，而不是被动滑入的境地。

There may be no single right division of labor between people and AI. Some decisions may be fine to hand over, and others worth keeping. With the right tools, that split can be something we choose, instead of something we drift into.

注意转载授权 · 阅读原文。中文正文为 AI 翻译，请以原文为准。

注意转载授权 · 阅读原文