宇宙悖论揭示无观测者世界的可怕后果
内容总结：
【本报综合报道】理论物理学界近期一项突破性研究显示，宇宙的复杂性可能依赖于观察者的存在。这一发现源于对“封闭宇宙”模型的长期探索，或将颠覆人类对客观物理定律的认知。
2019年，普林斯顿高等研究院的胡安·马尔达西那等学者运用描述量子引力的“全息原理”计算封闭宇宙时，发现一个令人费解的现象：这种宇宙在数学上仅存在单一量子态，其信息容量甚至无法承载最简单的0或1二进制选择。这与我们观测到的星辰璀璨、生命繁复的现实宇宙形成尖锐矛盾。
“我们环顾四周，世界显然比理论预测复杂得多。”加拿大圆周理论物理研究所的罗布·迈尔斯评论道。此后多项独立研究均证实该悖论，包括斯坦福大学亨利·马克斯菲尔德团队对“婴儿宇宙”模型的演算，都指向封闭宇宙理论上应有的极度简并特性。
转机出现在2023年。加州大学圣克鲁兹分校物理学家埃德加·沙古里安发现，该现象与数学中的“拓扑场论”具有相似性——当引入观察者划分观测边界时，系统的信息容量将急剧扩张。这一洞见促使麻省理工学院赵颖（音译）团队在2025年初发表关键研究，证明将经典观察者引入封闭宇宙模型后，世界的丰富性得以重现。
目前学界普遍认为，若该理论通过验证，将标志着物理学范式的重大转变：传统追求“无视角客观描述”的研究路径可能面临根本性调整，宇宙的复杂性或许本质上就与观察者密不可分。正如研究人员所言：“我们或许永远只能获得特定视角下的宇宙图景。”
（完）
中文翻译：
宇宙悖论揭示无观测者世界的可怕后果
引言
物理学家在案头推演量子时空的数学公式时，发现了一个令人费解的难题。量子理论与引力的玄妙法则让他们能够精确构想出各式各样的宇宙图景，这种强大的思想实验近年来不断破解着萦绕在黑洞周围的未解之谜。
但2019年当研究团队审视某个与我们的宇宙惊人相似的理论模型时，却遭遇了悖论：这个理论宇宙似乎只允许存在单一状态。它显得如此简单，以至于其内容甚至无法承载半个比特的信息——连在0与1之间做出选择的可能性都不存在。这一结论与事实严重冲突：此类宇宙本应能容纳黑洞、恒星、行星乃至人类文明，然而所有这些丰富的细节在模型中全然无踪。
"我们环顾四周，现实世界显然比这复杂得多。"加拿大滑铁卢圆周理论物理研究所的理论物理学家罗布·迈尔斯评论道，他未直接参与这项研究。
物理学家有充分理由相信这个基于基础物理理念的计算结果。数学推导指向仅存单一状态的宇宙，而我们的宇宙显然并非如此。如今某理论团队提出了可能的解答：当物理学家试图客观描述整个宇宙状态时，悖论便会产生。但这样的描述或许根本不可能实现——即便在理论上也是如此。这种尝试隐含着一个前提：宇宙可以在没有观测者的情况下独立存在。而若失去观测者，宇宙的复杂性或许就失去了意义。
惊世之论
对于同时痴迷量子力学与引力理论的物理学家而言，将这两大理论融合已被证明异常艰难。弦理论作为该问题的假定解决方案，用微小的振动弦取代粒子，消解了困扰其他候选理论的致命缺陷。然而该理论的数学表述极其复杂，其深层含义始终难以破解。
但约三十年前，高等研究院物理学家胡安·马尔达塞纳发表的开创性论文表明，棘手的弦理论计算有时可以通过粒子物理的成熟概念来规避。关键在于这种方法仅适用于具有特殊"反德西特"几何的宇宙。这种宇宙存在边界，常被比喻为罐头筒。奇妙的是，筒内发生的一切——从粒子碰撞到旋转黑洞——都会通过筒外壁的阴影显现。这仿佛意味着三维宇宙等同于平面屏幕上的影像，物理学家称之为全息原理。
全息原理带来了重大突破。2019年，马尔达塞纳与高等研究院的三位同事——艾哈迈德·阿尔姆海iri、拉古·马哈詹和赵颖——运用全息思维深化了对黑洞内部的认识。他们在前人工作基础上提出"岛公式"，用以追踪黑洞内部不同区域的边界。这项成果很快帮助他们及其他研究者破解了长期谜题：黑洞如何能在不违背其引力绝对性的前提下，透露坠入物质的信息（量子理论认为这必然发生）？这一成功让物理界确信岛公式是理解量子引力的可靠途径，后续研究更表明其适用性可超越最初的反德西特语境。
但这仅仅是个开端。
"黑洞是理想的理论试验场，但并非终极目标。"斯坦福大学物理学家亨利·马克斯菲尔德指出，"量子引力的核心课题是量子宇宙学"—即探索早期宇宙本质的征程。
问题在于我们并非生活在反德西特式的罐头宇宙中。宇宙膨胀的特性意味着它没有边界。无论行进多远，你永远触碰不到边缘。
宇宙无边界的一种可能是其具有"闭合"几何结构。在此情况下，沿直线航行的旅行者终将回到起点，就如同乘坐喷气机一直向东飞行最终会绕地球一周。
鉴于我们的宇宙可能正是这种闭合结构，马尔达塞纳很快将岛公式应用于闭合宇宙。他发现了令同行难以接受的结论：闭合区域似乎几乎完全空无一物。
"这个论证让我极为震惊。"赵颖回忆道，"我当时还试图反驳他。"经过数年探索，赵颖最终在马尔达塞纳的空旷宇宙模型中找到了突破口。
空白画卷
马尔达塞纳研究的闭合宇宙并非没有质量或能量，而是缺失了更重要的东西：信息。
当物理学家研究量子理论时，需要追踪物理系统所有可能的状态。为此他们使用称为希尔伯特空间的抽象空间。这个以20世纪初数学家大卫·希尔伯特命名的空间，通过增加数学维度来容纳不同量子状态。维度越多，希尔伯特空间能编码的信息就越丰富。
简单系统（如非0即1的计算机比特）可能只需两个维度。
但多数量子系统要复杂得多。以氢原子为例：随着能量增加，其电子可跃迁至更高轨道。此时可能状态数无限，因此其希尔伯特空间是无限维的。大多数真实量子系统都具有此特性。
物理学家由此预期整个宇宙也应具有无限数量的状态。但当马尔达塞纳将岛公式应用于闭合宇宙时，却发现其希尔伯特空间仅有一个维度。找不到任何信息踪迹，整个宇宙及其包含万物只能处于单一量子态，其复杂性甚至不及一个比特。
这个结论在物理界引发悖论：我们很可能就生活在闭合宇宙中，而周围显然存在着远超单一状态的丰富景象。
"我的书桌上就存在着无限多种状态。"加州大学圣克鲁兹分校物理学家埃德加·沙古利安指出。
然而随着物理学家持续研究各类闭合宇宙，相同模式不断重现。当高等研究院团队聚焦黑洞时，马克斯菲尔德与合作者唐纳德·马罗尔考察了名为"婴儿宇宙"的假定量子时空泡，同样发现了这种极致的简洁。闭合宇宙的贫瘠性似乎正成为普遍规律。
"最终我们接受了这个事实。"赵颖坦言。
复杂性的回归
现状构成了悖论：计算始终表明任何闭合宇宙仅存在单一可能状态，但我们这个很可能闭合的宇宙却展现出无限复杂性。究竟发生了什么？
在2023年的论文中，沙古利安指出物理学家早就在拓扑场论中见过类似奇特行为。数学家运用这些理论描绘几何空间的形状（即拓扑结构）。拓扑场论也可以具有一维希尔伯特空间，但若将几何空间分割为多个区域，就能用多种不同方式描述空间。要追踪所有新可能性，就需要更大的希尔伯特空间。
"游戏规则改变了。"沙古利安表示。
他提出可能存在类似方法分割闭合宇宙：引入观测者。
量子力学要求区分观测者（如进行实验的科学家）与其观测的系统。系统通常是微观量子客体（如原子），观测者则处于宏观层面且距离遥远，因此可用经典物理充分描述。沙古利安发现这种分割类似于拓展拓扑场论希尔伯特空间的方式。或许观测者也能为这些闭合且看似极度简单的宇宙带来改变？
2024年赵颖加入麻省理工学院，开始研究如何将观测者置入闭合宇宙。她与两位同事——丹尼尔·哈洛与米哈伊洛·乌萨秋克——将观测者视作引入的新型边界：并非宇宙边缘，而是观测者自身的界限。她们证明，当考虑闭合宇宙中的经典观测者时，世界的全部复杂性都将回归。
麻省理工团队的论文于2025年初发表，同期另一团队也提出了相似构想，其他学者纷纷指出这与早期工作的关联。
现阶段所有参与者都强调尚未找到完整解答。这个悖论本身可能是误解，或许会随着新论证而消解。但截至目前，为闭合宇宙添加观测者并考量其存在，或许是最稳妥的路径。
"我是否确信这就是正确答案，能彻底解决问题？我不敢断言。我们已竭尽所能。"赵颖表示。
若该构想成立，运用观测者的主观特性来解释宇宙复杂性，将代表物理学的范式转变。物理学家传统上追求"无视角观测"—即对自然进行独立于观测者的客观描述。他们渴望理解世界运行机制，以及我们这类观测者如何作为世界组成部分涌现。但随着物理学家通过个体观测者的私有边界来理解闭合宇宙，这种"无视角观测"似乎越来越难以实现。或许，"特定视角观测"才是我们唯一可能拥有的认知方式。
英文来源：
Cosmic Paradox Reveals the Awful Consequence of an Observer-Free Universe
Introduction
Tinkering at their desks with the mathematics of quantum space and time, physicists have discovered a puzzling conundrum. The arcane rules of quantum theory and gravity let them imagine many different kinds of universes in precise detail, enabling powerful thought experiments that in recent years have addressed long-standing mysteries swirling around black holes.
But when a group of researchers examined a universe intriguingly like our own in 2019, they found a paradox: The theoretical universe seemed to admit only a single possible state. It appeared so simple that its contents could be described without conveying even a single bit of data, not even a choice of a zero or a one. This result clashed with the fact that this type of universe should be capable of hosting black holes, stars, planets — and people. Yet all those rich details were nowhere to be seen.
“We look around, and certainly the world seems more complex than that,” said Rob Myers, a theoretical physicist at the Perimeter Institute for Theoretical Physics in Waterloo, Canada, who has not been directly involved in this research.
Physicists have good reason to trust the calculation, which builds on fundamental physical ideas. The math implies a universe with only one state; our universe is clearly not like that. Now a team of theorists has floated a possible answer. The paradoxical result occurred when physicists sought an objective description of the state of an entire universe. But a description like that might not be possible, even in principle. It implicitly assumes a universe that exists without an observer to observe it. And perhaps without observers, the complexity of the universe loses its meaning.
A Shocking Argument
For physicists passionate about both quantum mechanics and gravity, the two theories have proved extraordinarily difficult to combine. String theory is a putative solution to the problem, replacing particles with minuscule lengths of vibrating string to smooth over snags that doom other candidate theories. The theory’s math is challenging, however, and its implications have been tough to tease out.
But almost 30 years ago, a landmark paper by Juan Maldacena, a physicist at the Institute for Advanced Study, showed that difficult string theory calculations could sometimes be sidestepped and carried out using familiar concepts from particle physics instead. The catch is that this approach only works if the universe has an unusual “anti-de Sitter” geometry. An anti-de Sitter universe has a boundary, often illustrated to resemble a tin can. Remarkably, everything that happens inside the can, from colliding particles to spinning black holes, is revealed by shadows on the can’s outer boundary. It’s as if the 3D universe inside were equivalent to an image on a flat screen, a concept physicists call holography.
Holography has delivered major breakthroughs. In 2019, Maldacena and three colleagues at IAS — Ahmed Almheiri, Raghu Mahajan and Ying Zhao — used holographic thinking to better understand what happens inside a black hole. Building on earlier work, they proposed the “island formula,” which tracks the boundaries of different regions within a black hole. It soon helped them and others to uncover a potential explanation for a long-standing mystery: How can black holes reveal information about what has fallen into them — which quantum theory says must happen — when doing so would seem to defy the absolute nature of a black hole’s gravity? Their success gave physicists confidence that the island formula was a trustworthy way of understanding quantum gravity, and subsequent results showed that it could hold its own outside its original anti-de Sitter context.
But it was merely a warm-up.
“Black holes are a really good testing ground for ideas, but they’re not where the money is,” said Henry Maxfield, a physicist at Stanford University. “The important question of quantum gravity is quantum cosmology” — the push to understand the very early universe.
The problem is that we don’t live in an anti-de Sitter tin-can cosmos. The nature of the universe’s expansion implies that it has no boundary. No matter how far you travel, you will never hit an edge.
One way for a universe to lack an edge is for it to have a “closed” geometry. In this case, a traveler moving in a straight line could eventually return to where they started, much as you would if you jumped in a jet and flew due east.
Since our universe could be closed in this way, Maldacena soon applied the island formula to a closed universe. He uncovered something his colleagues found hard to accept: The closed region seemed almost completely empty.
“I was pretty shocked by that argument,” Zhao said. “I tried to argue with him.” It would take a few years, but Zhao would eventually find a hole in Maldacena’s empty universe.
Blank Slate
The closed universes Maldacena investigated weren’t empty of mass or energy. They were empty of something even more important: information.
When physicists study quantum theories, they need to keep track of each possible state a physical system can be in. To do this, they use an abstract space called Hilbert space. Hilbert spaces, named for the early-20th-century mathematician David Hilbert, account for different quantum states by adding new mathematical dimensions. The more dimensions there are, the more information these Hilbert spaces can encode.
A simple system, like a computer bit that can be either zero or one, might have two dimensions.
Most quantum systems are much more complex. Take a single hydrogen atom. Its electron can reach higher and higher orbits as you give it more energy. In this case, the number of possible states is unlimited, and so its Hilbert space is infinite-dimensional. Most real quantum systems have this feature.
Physicists therefore expect a whole universe to have an infinite number of states too. But when Maldacena applied the island formula to a closed universe, he found instead that it had a Hilbert space with just one dimension. There was no information to be found. The whole universe and everything in it could be in only one quantum state. It lacked even the complexity of a single bit.
This conclusion struck physicists as paradoxical, given that we too could conceivably live in a closed universe. And we clearly see far more than a single state around us.
“On my desk there are an infinite number of states,” said Edgar Shaghoulian, a physicist at the University of California, Santa Cruz.
But as physicists continued to study different types of closed universes, they kept seeing the same pattern. While the IAS group considered black holes, Maxfield and his collaborator Donald Marolf looked at hypothetical quantum bubbles of space-time called baby universes. They found the same stark simplicity. Increasingly, it appeared that the barrenness of closed universes was a universal trend.
“Eventually we believed it,” Zhao said.
Complexity Returns
The situation presents a paradox: Calculations consistently imply that any closed universe has only one possible state. But our universe, which may very well be closed, seems infinitely more complex. So what’s going on?
In a 2023 essay, Shaghoulian noted that physicists had seen this strange behavior before in theories called topological field theories. Mathematicians use these theories to chart the shape, or topology, of geometric spaces. Topological field theories can also have one-dimensional Hilbert spaces. But if you split up the geometric space into multiple zones, you can describe the space in many different ways. To keep track of all the new possibilities, you need a bigger Hilbert space.
“The rules of the game change,” Shaghoulian said.
Shaghoulian proposed that there might be a similar way to split up a closed universe: Bring in an observer.
Quantum mechanics requires a distinction between an observer — such as the scientist carrying out an experiment — and the system they observe. The system tends to be something small and quantum, like an atom. The observer is big and far away, and thus well described by classical physics. Shaghoulian observed that this split was analogous to the kind that enlarges the Hilbert spaces of topological field theories. Perhaps an observer could do the same to these closed, impossibly simple-seeming universes?
In 2024, Zhao moved to the Massachusetts Institute of Technology, where she began to work on the problem of how to put an observer into a closed universe. She and two colleagues —Daniel Harlow and Mykhaylo Usatyuk — thought of the observer as introducing a new kind of boundary: not the edge of the universe, but the boundary of the observer themself. When you consider a classical observer inside a closed universe, all the complexity of the world returns, Zhao and her collaborators showed.
The MIT team’s paper came out at the beginning of 2025, around the same time that another group came forward with a similar idea. Others chimed in to point out connections to earlier work.
At this stage, everyone involved emphasizes that they don’t know the full solution. The paradox itself may be a misunderstanding, one that evaporates with a new argument. But so far, adding an observer to the closed universe and trying to account for their presence may be the safest path.
“Am I really confident to say that it’s right, it’s the thing that solves the problem? I cannot say that. We try our best,” Zhao said.
If the idea holds up, using the subjective nature of the observer as a way to account for the complexity of the universe would represent a paradigm shift in physics. Physicists typically seek a view from nowhere, a stand-alone description of nature. They want to know how the world works, and how observers like us emerge as parts of the world. But as physicists come to understand closed universes in terms of private boundaries around private observers, this view from nowhere seems less and less viable. Perhaps views from somewhere are all that we can ever have.