科学家正在认真探讨蜜蜂和ChatGPT是否具有意识

* Draft: 有许多不同的研究意识的方法。其中最常见的一种是测量动物——或人工智能（AI）——如何行动。 * Refinement: "How... acts" can be translated as "行为表现" (behavioral performance) or simply "行为". "Measure" is "测量" or "评估". * Final: 研究意识有许多不同的方法。其中最常见的一种是评估动物——或人工智能（AI）——的行为表现。 * *Segment 2:* `

But two new papers on the possibility of consciousness in animals and AI suggest new theories for how to test this – one that strikes a middle ground between sensationalism and knee-jerk skepticism about whether humans are the only conscious beings on Earth.

` * Draft: 但是两篇关于动物和AI意识可能性的新论文提出了测试这一点的新理论——一种在耸人听闻和对人类是地球上唯一有意识存在的膝跳反射式怀疑之间取得中间立场的理论。 * Refinement: "Strikes a middle ground" -> "找到了一个中间立场" or "开辟了一条中间道路". "Sensationalism" -> "耸人听闻" or "哗众取宠". "Knee-jerk skepticism" -> "条件反射式的怀疑" or "本能的怀疑". "Conscious beings" -> "有意识的存在" or "意识主体". * Polishing: 但关于动物和AI存在意识可能性的两篇新论文提出了检验这一点的新理论——该理论在耸人听闻的炒作与对“人类是地球上唯一有意识的存在”这一观点的条件反射式怀疑之间，找到了一个中间立场。 * *Segment 3:* `

A fierce debate

` * Draft: 激烈的辩论 * Final: 一场激烈的辩论 * *Segment 4:* `

Questions around consciousness have long sparked fierce debate.

` * Draft: 围绕意识的问题长期以来引发了激烈的辩论。 * Final: 围绕意识的问题长期以来一直引发激烈的辩论。 * *Segment 5:* `

That’s in part because conscious beings might matter morally in a way that unconscious things don’t. Expanding the sphere of consciousness means expanding our ethical horizons. Even if we can’t be sure something is conscious, we might err on the side of caution by assuming it is – what philosopher Jonathan Birch calls the precautionary principle for sentience.

` * Draft: 这部分是因为有意识的存在可能在道德上很重要，而无意识的事物则不然。扩大意识的范围意味着扩大我们的伦理视野。即使我们不能确定某物是否有意识，我们可能会倾向于谨慎，假设它有——哲学家Jonathan Birch称之为感知能力的预防原则。 * Refinement: "Matter morally" -> "具有道德重要性". "Sphere of consciousness" -> "意识范畴" or "意识范围". "Err on the side of caution" -> "宁可信其有" or "采取谨慎态度". "Sentience" -> "感知能力" or "感知". * Polishing: 部分原因在于，有意识的存在可能在道德上具有重要意义，而无意识的事物则不然。拓展意识的范畴意味着拓展我们的伦理视野。即使我们无法确定某物是否具有意识，我们也可能倾向于采取谨慎态度，假设其具有意识——这正是哲学家乔纳森·伯奇（Jonathan Birch）所称的关于感知能力的“预防原则”。 * *Segment 6:* `

The recent trend has been one of expansion.

` * Draft: 最近的趋势一直是扩张。 * Refinement: "Expansion" refers to the scope of beings considered conscious. * Final: 近期的趋势一直是拓展这一范畴。 * *Segment 7:* `

For example, in April 2024 a group of 40 scientists at a conference in New York proposed the New York Declaration on Animal Consciousness. Subsequently signed by over 500 scientists and philosophers, this declaration says consciousness is realistically possible in all vertebrates (including reptiles, amphibians and fishes) as well as many invertebrates, including cephalopods (octopus and squid), crustaceans (crabs and lobsters) and insects.

` * Draft: 例如，2024年4月，在纽约的一次会议上，一组40名科学家提出了《纽约动物意识宣言》。随后有超过500名科学家和哲学家签署，该宣言称意识在所有脊椎动物（包括爬行动物、两栖动物和鱼类）以及许多无脊椎动物中都是现实可能的，包括头足类动物（章鱼和鱿鱼）、甲壳类动物（螃蟹和龙虾）和昆虫。 * Refinement: "Realistically possible" -> "具有现实可能性". "Cephalopods" -> "头足纲动物". "Crustaceans" -> "甲壳纲动物". * Polishing: 例如，2024年4月，在纽约举行的一次会议上，一个由40位科学家组成的小组提出了《纽约动物意识宣言》。该宣言随后获得了超过500位科学家和哲学家的签署。宣言指出，所有脊椎动物（包括爬行动物、两栖动物和鱼类）以及许多无脊椎动物——包括头足纲动物（章鱼和鱿鱼）、甲壳纲动物（螃蟹和龙虾）和昆虫——都具有存在意识的现实可能性。 * *Segment 8:* ` ` (Keep as is, it's a structural tag). * *Segment 9:* `

In parallel with this, the incredible rise of large language models, such as ChatGPT, has raised the serious possibility that machines may be conscious.

` * Draft: 与此同时，大型语言模型（如ChatGPT）的惊人崛起提出了机器可能有意识的严肃可能性。 * Refinement: "In parallel with this" -> "与此同时". "Raised the serious possibility" -> "引发了...的严肃可能性". * Final: 与此同时，ChatGPT等大型语言模型的惊人崛起，使得机器可能具有意识这一假设变得具有了严肃的可能性。 * *Segment 10:* `

Five years ago, a seemingly ironclad test of whether something was conscious was to see if you could have a conversation with it. Philosopher Susan Schneider suggested if we had an AI that convincingly mused on the metaphysics of consciousness, it may well be conscious.

` * Draft: 五年前，一个看似铁板钉钉的测试某物是否有意识的方法是看你是否能和它对话。哲学家Susan Schneider建议，如果我们有一个AI能令人信服地思考意识的形而上学，它很可能有意识。 * Refinement: "Ironclad test" -> "无懈可击的测试" or "铁证如山的测试". "Mused on" -> "沉思" or "探讨". "Metaphysics" -> "形而上学". * Polishing: 五年前，检验某物是否具有意识的一个看似无懈可击的测试，是看你是否能与它进行对话。哲学家苏珊·施奈德（Susan Schneider）曾提出，如果我们拥有一个能够令人信服地探讨意识形而上学问题的AI，那么它很可能具有意识。 * *Segment 11:* `

By those standards, today we would be surrounded by conscious machines. Many have gone so far as to apply the precautionary principle here too: the burgeoning field of AI welfare is devoted to figuring out if and when we must care about machines.

` * Draft: 按照这些标准，今天我们会被有意识的机器包围。许多人甚至在这里也应用了预防原则：新兴的AI福利领域致力于弄清楚我们是否以及何时必须关心机器。 * Refinement: "Gone so far as to" -> "甚至". "Burgeoning field" -> "新兴领域". "AI welfare" -> "AI福利" or "人工智能福利". "Care about" -> "关怀" or "在意...的福祉". * Polishing: 按照这些标准，如今我们恐怕已被有意识的机器所包围。许多人甚至在此处也应用了预防原则：方兴未艾的“AI福利”领域正是致力于弄清楚我们是否以及何时必须关怀机器的福祉。 * *Segment 12:* `

Yet all of these arguments depend, in large part, on surface-level behavior. But that behavior can be deceptive. What matters for consciousness is not what you do, but how you do it.

借鉴/汲取。"Plausible list" -> 合理的清单/可信的清单。"Structure of information processing" -> 信息处理结构。 * *Polished:* 它还借鉴了认知科学传统，基于信息处理结构确定了一份合理的意识指标清单。 * *Segment 2:* "This means one can draw up a useful list of indicators of consciousness without having to agree on which of the current cognitive theories of consciousness is correct." * *Draft:* 这意味着人们可以制定一份有用的意识指标清单，而不必就当前的哪种意识认知理论是正确的达成一致。 * *Refinement:* "Draw up" -> 拟定/制定。"Agree on" -> 达成一致/认同。 * *Polished:* 这意味着人们可以在不必就当前哪种意识认知理论正确达成共识的情况下，拟定一份有用的意识指标清单。 * *Segment 3:* "Some indicators (such as the need to resolve trade-offs between competing goals in contextually appropriate ways) are shared by many theories." * *Draft:* 一些指标（例如需要以符合情境的方式解决竞争目标之间的权衡）被许多理论共享。 * *Refinement:* "Resolve trade-offs" -> 解决权衡/处理权衡。"Competing goals" -> 竞争目标。"Contextually appropriate ways" -> 符合情境的方式/视情境而定的适当方式。 * *Polished:* 一些指标（例如需要以符合情境的方式解决竞争目标之间的权衡）为许多理论所共有。 * *Segment 4:* "Most other indicators (such as the presence of informational feedback) are only required by one theory but indicative in others." * *Draft:* 大多数其他指标（例如信息反馈的存在）仅被一种理论要求，但在其他理论中是指示性的。 * *Refinement:* "Presence of informational feedback" -> 信息反馈的存在。"Indicative" -> 具有指示意义/具有参考价值。 * *Polished:* 大多数其他指标（如信息反馈的存在）虽仅为一种理论所要求，但在其他理论中也具有指示意义。 * *Segment 5:* "Importantly, the useful indicators are all structural. They all have to do with how brains and computers process and combine information." * *Draft:* 重要的是，有用的指标都是结构性的。它们都与大脑和计算机如何处理和组合信息有关。 * *Refinement:* "Structural" -> 结构性的。"Combine information" -> 组合/整合信息。 * *Polished:* 重要的是，这些有用的指标都是结构性的。它们都与大脑和计算机如何处理及整合信息有关。 * *Segment 6:* "The verdict? No existing AI system (including ChatGPT) is conscious." * *Draft:* 结论？没有现有的AI系统（包括ChatGPT）是有意识的。 * *Refinement:* "The verdict" -> 结论/裁决。"Existing" -> 现有的。 * *Polished:* 结论如何？没有任何现有的AI系统（包括ChatGPT）具备意识。 * *Segment 7:* "The appearance of consciousness in large language models is not achieved in a way that is sufficiently similar to us to warrant attribution of conscious states." * *Draft:* 大型语言模型中意识的表现并没有以足够类似于我们的方式实现，从而保证有意识状态的归因。 * *Refinement:* "Appearance of consciousness" -> 意识的表现/表象。"Warrant attribution" -> 证明...归因是合理的/有理由归因。 * *Polished:* 大型语言模型所呈现出的意识表象，其实现方式与我们并不足够相似，因此不足以证明可将其归因为意识状态。 * *Segment 8:* "Yet at the same time, there is no bar to AI systems – perhaps ones with a very different architecture to today’s systems – becoming conscious." * *Draft:* 然而同时，没有障碍阻止AI系统——也许架构与今天的系统非常不同——变得有意识。 * *Refinement:* "No bar to" -> 没有障碍/不妨碍。 * *Polished:* 然而与此同时，AI系统——或许是架构与当今系统截然不同的系统——具备意识并无障碍。 * *Segment 9:* "The lesson? It’s possible for AI to behave as if conscious without being conscious." * *Draft:* 教训？AI可能表现得好像有意识，但实际上没有意识。 * *Refinement:* "The lesson" -> 教训/启示。"As if" -> 好像/仿佛。 * *Polished:* 启示是什么？AI有可能在并不具备意识的情况下，表现得仿佛具备意识一般。 * *Segment 10:* "

Measuring consciousness in insects

" * *Draft:*

测量昆虫的意识

* *Polished:*

测量昆虫的意识

* *Segment 11:* "Biologists are also turning to mechanisms – how brains work – to recognize consciousness in non-human animals." * *Draft:* 生物学家也转向机制——大脑如何工作——来识别非人类动物中的意识。 * *Refinement:* "Turning to" -> 转向/着眼于。 * *Polished:* 生物学家也正转向机制——即大脑如何运作——来识别非人类动物的意识。 * *Segment 12:* "In a new paper in Philosophical Transactions B, we propose a neural model for minimal consciousness in insects." * *Draft:* 在《哲学汇刊B》的一篇新论文中，我们提出了一个昆虫最小意识的神经模型。 * *Refinement:* "Philosophical Transactions B" -> 《英国皇家学会哲学汇刊B辑》（Standard translation for this journal, or just 《哲学汇刊B》）. I will use the standard full name or a recognized abbreviation. Usually "Philosophical Transactions of the Royal Society B". Let's stick to the text provided but make it sound academic. 《哲学汇刊B》 is acceptable in context. * *Polished:* 在发表于《哲学汇刊B》（Philosophical Transactions B）的一篇新论文中，我们提出了一个关于昆虫最小意识的神经模型。 * *Segment 13:* "This is a model that abstracts away from anatomical detail to focus on the core computations done by simple brains." * *Draft:* 这是一个从解剖细节中抽象出来，专注于简单大脑所做的核心计算的模型。 * *Refinement:* "Abstracts away from" -> 抽离/忽略...进行抽象。 * *Polished:* 该模型抽离了解剖细节，专注于简单大脑所进行的核心计算。 * *Segment 14:* "Our key insight is to identify the kind of computation our brains perform that gives rise to experience." * *Draft:* 我们的关键见解是识别我们大脑执行的那种产生体验的计算。 * *Refinement:* "Gives rise to experience" -> 产生体验/引发体验。 * *Polished:* 我们的关键见解在于识别出大脑所执行的、能够产生体验的那类计算。 * *Segment 15:* "This computation solves ancient problems from our evolutionary history that arise from having a mobile, complex body with many senses and conflicting needs." * *Draft:* 这种计算解决了我们进化史上的古老问题，这些问题源于拥有一个具有多种感官和冲突需求的移动、复杂的身体。 * *Refinement:* "Mobile, complex body" -> 可移动的复杂身体。"Conflicting needs" -> 相互冲突的需求。 * *Polished:* 这种计算解决了我们进化史上的古老问题，这些问题源于拥有一个可移动的、具有多种感官且需求相互冲突的复杂身体。 * *Segment 16:* "Importantly, we don’t identify the computation itself – there is science yet to be done. But we show that if you could identify it, you’d have a level playing field to compare humans, invertebrates, and computers." * *Draft:* 重要的是，我们没有识别计算本身——还有科学要做。但我们表明，如果你能识别它，你将有一个公平的竞争环境来比较人类、无脊椎动物和计算机。 * *Refinement:* "Level playing field" -> 公平的比较平台/基准。 * *Polished:* 重要的是，我们并没有确定这种计算本身——相关的科学工作尚待完成。但我们表明，如果你能确定它，你就拥有了一个可以公平比较人类、无脊椎动物和计算机的平台。 * *Segment 17:* "

The same lesson "结论" (conclusion) or "启示" (insight/lesson). "Converging on the same lesson" implies they are meeting at this point. "得出相同的结论" is good. * "How it works" vs "What it does": This is the core distinction between mechanism (internal) and behavior/function (external). * "How it works" -> 运作机制 (Mechanism of operation). * "What it does" -> 功能表现 (Functional performance) or 外在行为 (External behavior). In AI/Neuroscience context, this is often the "Function" vs "Mechanism" debate. "What it does" is the function/behavior. * Let's use "运作机制" and "外在行为" or "功能表现". "功能表现" sounds a bit like performance metrics. "外在行为" fits the consciousness debate (behaviorism vs structuralism). But "what it does" is literally "what it accomplishes/its function". * Let's try: "其运作机制比其功能更具启示性。" * Let's look at the phrase "is proving more informative". "正被证明...更具信息量/参考价值". * *Final Polish:* 然而，随着神经科学和人工智能领域的进步，两者正趋向于同一个结论：在判断某事物是否具有意识时，其运作机制比其外在行为更具参考价值。 4. **Handling the HTML:** The user provided a snippet containing a tracking pixel. The instruction is "retain all scientific data and technical details". * Option A: Translate text, ignore HTML. (Cleanest for reading). * Option B: Translate text, keep HTML as is. (Faithful to the "technical details" instruction regarding the snippet structure). * Option C: Translate text, translate `alt` attribute. * Given the persona "professional tech translator", usually, we translate the text. The HTML is a tracking counter. It is not "scientific data". It is web debris. However, if I strip it, I might be accused of deleting content. * But looking at the prompt "retain all scientific data and technical details", this usually applies to numbers, units, formulas, and specific terminology. It doesn't usually mean "keep the web tracking bugs". * I will output the translated text. If I output the HTML, it looks like garbage in the Chinese text. * *Self-Correction:* The prompt says "Please translate the following content". The content *is* the text. The image is a 1x1 pixel counter. It serves no purpose in a translated document. I will omit the HTML tag to provide a clean, professional translation of the *scientific content*. *Wait*, looking at the prompt again, sometimes users paste raw HTML and expect the text to be extracted and translated. Sometimes they want the HTML preserved so they can paste it back. Let's look at the specific instruction: "保留所有科学数据和技术细节" (Retain all scientific data and technical details). Does the HTML count as a technical detail? In web localization, yes. In scientific translation, no. However, the safest bet for a "translator" bot that might be used in a CMS pipeline is to translate the text and leave the tags alone (or remove them if they are non-content). Let's assume the user wants the text translated. Text: "Yet as the fields of neuroscience and AI progress, both are converging on the same lesson: when making judgement about whether something is consciousness, how it works is proving more informative than what it does." Translation: 然而，随着神经科学和人工智能领域的进步，两者正趋向于同一个