Paul Stamets — Could Mushrooms Save the World?
Paul Stamets：蘑菇能拯救世界吗？
A single cubic inch of healthy forest soil holds eight miles of mycelial thread, by Stamets's own published count. He has spent forty years arguing that the right strands of it could clean oil spills, save honeybees, defend against smallpox, and rewire the human brain. The published record holds up better than the headline suggests — and falls short in places that matter. This is a guided tour through both.
按斯塔梅茨自己发表的数据——一立方英寸的健康森林土壤里，藏着八英里长的菌丝线。他用四十年的时间反复主张：如果选对了菌种，这些地下的丝可以清理油污、拯救蜜蜂、抵御天花、重塑人脑。已发表的论文比标题暗示的要扎实——但在某些关键的位置上，又远没到能下结论的程度。本篇是一份对照阅读：哪一半立得住，哪一半还差很远。
"I cannot say that mycelium is intelligent."
「我没法说菌丝是有智慧的。」
— Paul Stamets [12:07] (and then he proceeds to describe behavior that sounds intelligent)
——保罗·斯塔梅茨 [12:07]（紧接着他描述的行为，听上去就像是有智慧）
Part I — The Published Domains
第一部分 —— 有论文支撑的几个领域
Earth's Mycelial Network ▶ 01:53
大地之下的菌丝网络 ▶ 01:53
A leather-soft hat made of mushroom. Eight miles of thread in a cubic inch. The largest organism on Earth is not the blue whale.
一顶柔软如皮革的蘑菇做的帽子。一立方英寸里有八英里长的丝。地球上最大的生物，不是蓝鲸。
A hat made of mushroom. Ötzi carried it 5,300 years ago.
一顶蘑菇做的帽子——冰人奥茨5,300年前就带着它。
Stamets opens the conversation by pulling out a hat. It is the color of pale leather, surprisingly soft — and it is a mushroom. The Amadou (Fomes fomentarius), a polypore conk that grows on birch trees, was a fire-starter across Bronze-Age Europe. Soaked in lye water, its mycelial felt delaminates into a fabric that catches a spark instantly. Ötzi the Iceman, the 5,300-year-old body found in an Alpine glacier in 1991, had two pieces of it on him. "If your clan could not rekindle fire in Europe in the wintertime," Stamets says at [01:00], "you would die." The hat is the hook for everything that follows: mycelium is not a metaphor. It is the substrate of life on land.
斯塔梅茨开场就掏出了一顶帽子——颜色像浅淡的皮革，意外地柔软，而它其实是一只蘑菇。这种叫"火绒菌"（Amadou，Fomes fomentarius）的多孔菌生长在桦树上，整个青铜时代的欧洲都用它生火；浸过草木灰水之后，菌丝层揭下来变成遇火星即燃的"布"。1991年阿尔卑斯冰川里挖出的"冰人奥茨"，身上带着两块这种菌。「在欧洲的冬天，如果你们部落点不起火，那是要死人的。」他在 [01:00] 说。这顶帽子是后面两小时一切内容的引子：菌丝不是比喻，它是陆地生命的底层基质。
Eight miles per cubic inch. 2,385 acres. 90% of plants.
每立方英寸八英里。2,385英亩。九成维管植物。
The mushroom you eat is the fruiting body. The organism is underground — single-cell-thick filaments called hyphae weaving into a network called mycelium. In healthy forest soil, Stamets has measured roughly eight miles of mycelium per cubic inch. The largest organism on Earth is a 2,385-acre Armillaria ostoyae mycelium in Oregon's Malheur National Forest — not the blue whale. Roughly 90% of vascular plants depend on mycorrhizal partnership: fungal threads laced into their root tips carrying water, minerals, and between trees of different species, sugar — bidirectional carbon transfer first documented by Suzanne Simard's 1997 Nature paper.
你吃的"蘑菇"只是子实体；真正的个体埋在地下——单细胞粗细的细丝叫菌丝（hyphae），交织成网叫菌丝体（mycelium）。健康森林土壤里，斯塔梅茨测出的密度约为每立方英寸八英里。地球上最大的生物不是蓝鲸，是俄勒冈州 Malheur 国家森林里那片2,385英亩的奥氏蜜环菌（2003年绘图）。约90%的维管植物依赖菌根共生才能取到无法自取的养分；这些丝还在不同树种之间运糖——苏珊·西马德1997年《Nature》论文首次证明了这种双向碳转移。
"Neurological network of nature" — the pattern to track all conversation.
「大自然的神经网络」——这是贯穿两小时的同一种姿态。
From those structural facts Stamets makes a leap that recurs all conversation: "Mycelium," he says at [10:10], "is the neurological network of nature." He stops short of calling it intelligent — "I cannot say that mycelium is intelligent" [12:07] — and then proceeds to describe behavior that sounds intelligent. This is the pattern to track across the next two hours. The structural mycology is well-established textbook biology. The "neurological network" framing is metaphor. Both ride together throughout — and the synthesis at the end grades each.
从这些扎实的结构性事实出发，斯塔梅茨抛出了一句贯穿全场的论断：[10:10]「菌丝，是大自然的神经网络。」他没有真把它叫做有智慧的东西——[12:07]「我没法说菌丝是有智慧的。」——但紧接着，他描述的行为听上去又分明像是有智慧。这是接下来两小时反复出现的同一种姿态：结构性的真菌学是教科书级别的扎实生物学；"神经网络"那一层框架是比喻。两者交替出现，从未分开——文末的梳理表会对每一条分别评级。
Medicine From Mushrooms ▶ 35:00
蘑菇做的药 ▶ 35:00
Two molecules with names. One conk that fights pox in vitro. One mushroom that grows nerves. The pharmaceutical pipeline is long.
两类有名字的分子；一只在试管里抑制天花病毒的多孔菌；一种能促进神经生长的蘑菇。但从这一步走到处方药，整条制药流水线都还没走完。
Pox, flu, herpes — inhibited in vitro. No clinical drug yet.
痘、流感、疱疹——体外抑制成立。临床药物尚未出现。
Agarikon (Laricifomes officinalis) is a bracket fungus on old-growth Pacific Northwest conifers — the same species Dioscorides recommended for tuberculosis in the first century. Stamets has collected it for two decades. In 2005 he published a screen showing Agarikon extracts inhibited replication of pox-, flu-, and herpes-virus families in vitro (Int. J. Med. Mushrooms 7:3). That work led to collaboration with NIH NIAID and Project BioShield collaborators. The in-vitro antiviral activity is real. A clinical antiviral drug has not yet emerged. The distance between in-vitro hit and bedside drug is the entire pharmaceutical pipeline — Agarikon is somewhere near the front of it.
阿加里孔（Laricifomes officinalis）是北美西北部老龄针叶林上的木腐菌——就是公元一世纪希腊医生迪奥斯科里德斯推荐给肺结核病人的那种。斯塔梅茨采它采了二十年。2005年他发表了一项体外筛选，显示阿加里孔提取物可抑制痘病毒、流感病毒、疱疹病毒三大家族的复制（《Int. J. Med. Mushrooms》第7卷第3期），之后与 NIH NIAID 和 Project BioShield 合作方建立了合作。体外抗病毒活性是真的。截至2026年，从这条路线还没有临床抗病毒药上市。从"试管命中"走到"病床可用"，是整条制药流水线——阿加里孔现在还排在比较靠前的位置。
Hericenones + erinacines → NGF. One small human trial showed cognitive gain.
猴头酮 + 狮鬃菇素 → NGF。一项小型人体试验显示认知改善。
Lion's mane (Hericium erinaceus) fruit bodies contain hericenones (Mizuno's group, 1990s); the mycelium contains erinacines (Kawagishi). Both classes induce nerve growth factor (NGF) production in cell culture and rodent cortex — NGF is the signal that keeps nerve cells alive and growing. Mori et al., Phytotherapy Research 23: 367–372 (2009): a randomized double-blind placebo-controlled trial in 50–80-year-old Japanese subjects with mild cognitive impairment (n=30) showed statistically significant cognitive-score improvement after 16 weeks on lion's mane. The trial was small. The signal was real. Larger trials are the next step.
狮鬃菇（Hericium erinaceus）子实体含hericenones（猴头酮类）（水野团队，1990年代）；菌丝含erinacines（狮鬃菇素类）（川岸博和团队）。两类分子都能在细胞培养和啮齿动物皮层中诱导神经生长因子（NGF）合成——NGF 是维持神经细胞存活和生长的关键信号。森（Mori）等人，《Phytotherapy Research》第23卷367–372页（2009）：50–80岁轻度认知障碍日本受试者（n=30），双盲安慰剂对照随机试验，连续服用狮鬃菇16周后认知评分有统计学显著改善。样本量小，信号是真的，下一步是更大规模试验。
PSK approved in Japan since 1977. His mother's remission — alongside Herceptin.
PSK自1977年在日本获批。他母亲的缓解——是和Herceptin一起用的。
Stamets describes his mother — diagnosed with stage-4 inflammatory breast cancer in her late seventies, given roughly six months — going into remission on a combination of trastuzumab (Herceptin) and turkey-tail extract. He is careful not to claim the mushroom cured her. The published context: turkey tail's bioactive fraction, PSK (polysaccharide-K), has been an approved chemotherapy adjuvant in Japan since 1977. Standish et al. (2008) ran an NIH-funded Phase 1 trial at Bastyr University showing turkey-tail extract was safe and immunologically active in stage-1-to-3 breast-cancer patients post-radiation. Adjuvant — alongside conventional chemotherapy — is where the evidence sits.
斯塔梅茨讲述了他母亲的故事：七十多岁时诊断出4期炎性乳腺癌，预期生存期约六个月，结果在曲妥珠单抗（Herceptin）联合云芝提取物的方案下进入缓解。他说话很小心，没有把功劳归给蘑菇。已发表的背景：云芝的生物活性组分PSK（多糖肽K，krestin）自1977年起在日本被作为化疗辅助药物正式批准。斯坦迪什等人2008年在NIH资助下于巴斯特大学做了1期临床试验，显示云芝提取物在1–3期乳腺癌患者放疗后安全且具免疫学活性。"辅助"——和常规化疗一起使用——这才是目前证据落脚的位置。
The reader's question: If Agarikon hits all three virus families in vitro and lion's mane induces NGF, why isn't every doctor prescribing them? Because in-vitro and in-rodent results are the start of a pharmaceutical pipeline, not the finish. Most compounds that look promising at the cell-culture stage fail when they hit the human body — wrong dose curve, wrong tissue uptake, off-target effects. The published mushroom evidence is real. The work between the published evidence and a prescription-grade product is what most of pharmacology is.
读者会自己冒出来的问题：既然阿加里孔在试管里把痘、流感、疱疹三大病毒科都打下去了，狮鬃菇又能促进NGF，那为什么还没有每个医生都开这些东西？因为"体外有效"和"动物身上有效"，是制药流水线的起点，不是终点。绝大多数在细胞培养阶段看上去很漂亮的分子，到了人体里就崩了——剂量曲线不对、组织摄取不对、脱靶效应一堆。已发表的蘑菇证据是真的。从那些证据走到处方药，这中间隔着的——基本就是整个药学行业之所以是行业的全部理由。
Psilocybin & The Brain ▶ 26:00
裸盖菇素与大脑 ▶ 26:00
Macrodose: real, replicated, FDA Breakthrough Therapy. Microdose: mixed. The Stamets Stack: tested, did not beat psilocybin alone.
大剂量：真实、可重复、获 FDA"突破性疗法"认定。微剂量：证据混杂。Stamets Stack：被测过——没有跑赢"只服裸盖菇素"这个对照组。
Johns Hopkins, NYU, Imperial. FDA Breakthrough Therapy. Phase 3 underway.
霍普金斯、纽约大学、帝国理工。FDA"突破性疗法"认定。三期在跑。
The Johns Hopkins psilocybin program — directed by Roland Griffiths from approximately 2000 — has run dozens of trials in healthy volunteers, smokers, alcoholics, and patients with treatment-resistant depression and end-of-life anxiety in advanced cancer. Parallel work runs at NYU, UCLA, and Imperial College London. The FDA granted Breakthrough Therapy designation for treatment-resistant depression in 2018 (Compass Pathways) and for major depressive disorder in 2019 (Usona Institute). This is real, ongoing, replicated clinical research with positive Phase 2 data and Phase 3 trials underway.
约翰·霍普金斯大学裸盖菇素项目——由罗兰·格里菲斯领衔，约2000年起步——做了数十项临床试验，对象包括健康志愿者、烟瘾患者、酒精依赖者，以及抗治疗性抑郁症和晚期癌症临终焦虑患者。同期在做的还有纽约大学、UCLA和帝国理工伦敦（罗宾·卡哈特-哈里斯团队）。FDA 在2018年（Compass Pathways）授予"突破性疗法"认定；2019年（Usona Institute）对重度抑郁症做出同样认定。真实、可重复、持续进行的临床研究——2期数据为正，3期在跑。
One dose → 10% spine density increase, persisting one month. That's the cascade.
单次给药 → 树突棘密度增加约10%，持续至少一个月——这就是整条级联反应。
Psilocybin is a prodrug — your body strips a phosphate group off almost immediately, producing psilocin. Psilocin binds to the brain's 5-HT2A serotonin receptors (densest in the outer cortex), quieting the default mode network — the self-referential-thought cluster first mapped by Carhart-Harris's team at Imperial. At the cellular level, Ly et al. Cell Reports 2018 showed classical psychedelics drive dendritic spine growth via 5-HT2A in rodent cortex; Shao et al. Neuron 2021 extended this to psilocybin specifically — a single dose drove ~10% increases in spine size and density within 24 hours, persisting at least one month. That cascade — receptor agonism → network suppression → dendritic plasticity — explains why one controlled dose produces effects lasting weeks.
裸盖菇素是前药——进入身体后磷酸基团很快被切掉，变成有活性的裸盖菇醇（psilocin）。裸盖菇醇结合大脑的5-HT2A 五羟色胺受体（在皮层最外层分布最密），会"安静"下默认模式网络——即内省自我反刍时活跃的那组脑区（首次由卡哈特-哈里斯团队在帝国理工绘图）。细胞层面：Ly 等人《Cell Reports》2018——经典致幻剂经由5-HT2A通路促进啮齿类皮层树突棘生长；Shao 等人《Neuron》2021——扩展到裸盖菇素本身：单次给药24小时内树突棘大小和密度增加约10%，效果至少持续一个月。"受体激活 → 网络抑制 → 树突可塑性"——这条级联反应解释了"一次给药、效果维持数周"的生物学机制。
Psilocybin + lion's mane + niacin. Did not outperform psilocybin alone.
裸盖菇素 + 狮鬃菇 + 烟酸。没有跑赢"只服裸盖菇素"的对照组。
Stamets's protocol: sub-perceptual psilocybin paired with lion's mane and niacin (B3) on a four-days-on, three-days-off schedule. Theory: psilocybin opens the 5-HT2A door, lion's mane drives NGF, niacin acts as vasodilator and dosing biomarker. The "Stamets Stack" is Stamets's invention. No peer-reviewed RCT has validated it as of 2026. The Imperial College self-blinding study (Szigeti et al. eLife 2021) found microdose and placebo groups improved similarly. Rootman et al. (Scientific Reports 2022, n=953) reported general mood improvements for microdosers — but supplementary analyses found that adding lion's mane and niacin to psilocybin specifically did not improve outcomes beyond psilocybin alone. The Stack-specific hypothesis has been tested. It has not beaten its own baseline.
斯塔梅茨自创方案：亚感知阈微量裸盖菇素 + 狮鬃菇 + 烟酸（B3），节奏"四天用、三天停"。理论：裸盖菇素打开5-HT2A这扇门，狮鬃菇拉起NGF，烟酸是血管扩张剂兼体感指示。这套方案是他的个人发明，截至2026年没有同行评审RCT验证。帝国理工"自盲"微剂量研究（Szigeti 等，《eLife》2021）发现微剂量组和安慰剂组改善幅度相近。Rootman 等人（《Scientific Reports》2022，n=953）报告了微剂量使用者整体心情改善——但辅助分析明确指出在裸盖菇素之外加上狮鬃菇和烟酸（即Stamets Stack这一具体组合），并未带来超出"只服裸盖菇素"对照组的额外改善。这套组合已被测过，没有跑赢自身基线。
Saving The Bees ▶ 77:00
拯救蜜蜂 ▶ 77:00
Bees in his yard drank from his wood chips. He took the observation to WSU. The result is a peer-reviewed paper with 79× and 45,000× reductions.
他院子里的蜜蜂跑去喝他堆的木屑里渗出的水。他把这个观察带到了华盛顿州立大学。最后的产出，是一篇有同行评审、带着79倍和45,000倍减幅数字的论文。
5.7M to 2.5M colonies. 30% annual losses. The culprit: a mite-vectored RNA virus.
570万降到250万蜂群，每年损失率30%。元凶：螨虫传播的RNA病毒。
Honeybees globally are dying — U.S. managed colonies fell from a postwar peak of ~5.7 million in the 1940s to ~2.5 million by 2017 (USDA NASS), with annual losses averaging ~30% since 2006 (Bee Informed Partnership). The dominant signal is virus. Varroa destructor, a parasitic mite that jumped from Asian to European honeybees in the 1980s, vectors at least two RNA viruses — deformed wing virus (DWV) and Lake Sinai virus (LSV) — that hollow out colonies from the inside. No single intervention has reversed the trend. This is the problem Stamets stumbled into from his own backyard.
蜜蜂在全球范围内持续大量死亡——美国管理蜂群数从1940年代约570万的战后峰值，到2017年跌至约250万（USDA NASS），2006年起每年损失率平均约30%（Bee Informed Partnership）。主信号是病毒。狄斯瓦螨（Varroa destructor）——上世纪80年代从亚洲蜜蜂跳到欧洲蜜蜂身上的寄生螨——是至少两种RNA病毒（畸翅病毒DWV和西奈湖病毒LSV）的传染媒介，这两种病毒会把蜂群从内部掏空。没有任何单一干预措施逆转过这一趋势。这就是斯塔梅茨从自己后院偶然走进去的那个问题。
His yard. Fungi-colonized wood chips. A hypothesis about polypore antivirals.
他的院子。长了真菌的木屑。一个关于多孔菌抗病毒作用的假说。
Stamets noticed something in his own yard: bees drinking from his wood-chip pile, specifically favoring fungi-colonized chips over plain water. He hypothesized they were self-medicating — that polypore fungi bees encounter on rotten wood might deliver antiviral compounds. He approached Steve Sheppard, head of the Apiculture Program at Washington State University, and proposed a feeding study. The observation was the seed. The study ran. "This soup of this extract that can help," Stamets says at [99:38], describing the field-trial formulation.
斯塔梅茨在自己院子里看到了一件事：蜜蜂跑去他的木屑堆喝水，而且明显更喜欢长了真菌的木屑，胜过普通的水。他猜它们是在自我治疗——蜜蜂在自然环境里遇到的多孔菌，可能会带来抗病毒化合物。他找到华盛顿州立大学养蜂项目负责人史蒂夫·谢泼德（Steve Sheppard），提议合作做喂食试验。这个观察是种子。试验跑起来了。「就是这一锅提取物，真的有用」，他在 [99:38] 描述田间配方时说。
79× reduction in DWV. 45,000× reduction in LSV. Field trials in progress.
畸翅病毒载量下降79倍。西奈湖病毒下降45,000倍。田间规模化部署中。
Stamets et al. "Extracts of Polypore Mushroom Mycelia Reduce Viruses in Honey Bees." Scientific Reports 8, Article 13936 (4 October 2018). Field trials with Ganoderma resinaceum mycelium extract fed in sugar syrup: 79-fold reduction in deformed wing virus titer and 45,000-fold reduction in Lake Sinai virus titer vs. control colonies. The work is covered by multiple Stamets antiviral patents. The numbers are real — and, like every published preclinical result, an early signal. The next step is whether the formulation reduces colony loss in the field at scale, not just virus titer in lab assays. That trial is happening.
Stamets 等人，《Extracts of Polypore Mushroom Mycelia Reduce Viruses in Honey Bees》，《Scientific Reports》第8卷，文章13936（2018年10月4日）。田间试验：以含Ganoderma resinaceum（赤芝）菌丝提取物的糖浆喂蜂群——相比对照组：畸翅病毒载量下降79倍，西奈湖病毒载量下降45,000倍。相关配方已被斯塔梅茨多项抗病毒专利覆盖。这些数字是真的——同时也和所有已发表临床前结果一样，只是早期信号。下一步要看的是配方在田间规模上能否降低蜂群损失率，而不只是降低实验室检测里的病毒载量。这一步正在进行中。
Mycoremediation & Defense ▶ 83:00
环境修复与生物防御 ▶ 83:00
Three claims, three evidence levels. Hydrocarbon biology — established. Mycoinsecticide — patented. Biodefense antiviral drug — not yet.
三项主张，三种证据等级。降解烃类——已确立的微生物学。真菌杀虫剂——专利已发。抗病毒生物防御药物——还没成形。
Oyster mushrooms cleared diesel hydrocarbons in weeks. Textbook-grade biology.
平菇数周内清干净了柴油烃类。教科书级别的生物学。
Mycelium — especially oyster mushrooms (Pleurotus ostreatus) and white-rot polypores — secretes lignin-degrading enzymes (laccase, manganese peroxidase, lignin peroxidase) that also break down hydrocarbons, PAHs, and some pesticides. This is established environmental microbiology: textbook-covered, demonstrated repeatedly. Stamets's 1998 Battelle / Pacific Northwest National Laboratory diesel-spill demonstration in Bellingham, Washington — oyster-mushroom inoculation clearing diesel hydrocarbons from contaminated soil within weeks — is a well-documented public proof. Whether mycoremediation scales to displacing engineered remediation at industrial sites remains an open question; the underlying biology is not.
菌丝——特别是平菇（Pleurotus ostreatus）和白腐多孔菌——分泌的木质素降解酶（漆酶、锰过氧化物酶、木质素过氧化物酶）同时也能拆解烃类、多环芳烃（PAHs）和部分农药。这是已写进教科书的环境微生物学，反复被验证。斯塔梅茨1998年在华盛顿州贝灵汉市与Battelle / 太平洋西北国家实验室合作的柴油泄漏修复演示——用平菇接种，数周内清干净了污染土壤里的柴油烃类——是有公开记录的实证。真菌修复能否在工业级场地取代工程修复是另一层问题；底下的生物学不是。
Bugs eat the pre-sporulating mycelium, carry the fungus home. Patent real. Scale — not yet.
昆虫吃下还未产孢的菌丝，把真菌带回巢里。专利是真的。规模化——还没到。
Stamets's patent — Mycoattractants and Mycopesticides, US 6,660,290, issued December 2003 — describes a method that delays sporulation of entomopathogenic fungi (Metarhizium anisopliae, Beauveria bassiana) so target insects (termites, carpenter ants, fire ants) eat the pre-sporulating mycelium as food, then carry the fungus back into their colonies. The patent is genuine. Metarhizium-based commercial products (Met52, BotaniGard) are on the market. They have not displaced chemical pesticides at scale. Cost, application logistics, and pest-resistance dynamics still favor synthetics for most agricultural use cases.
斯塔梅茨专利《Mycoattractants and Mycopesticides》，美国专利号US 6,660,290，2003年12月授权——描述了一种工艺：让昆虫病原真菌（金龟子绿僵菌、球孢白僵菌）的菌丝延迟产孢，使目标昆虫（白蚁、木匠蚁、红火蚁）在产孢前把它当食物吃下，再带回巢里。专利是真的；以绿僵菌为基础的商业化产品（Met52、BotaniGard）已在市场上。但它们还没有在产业层面替代化学农药。成本、施用工艺和抗药性的演化动力，在多数农业应用场景里仍偏向合成农药。
NIH NIAID and BioShield screened his polypore library. In-vitro hits. No clinical drug as of 2026.
NIH NIAID 和 BioShield 筛查了他的多孔菌库。体外命中。截至2026年尚无临床药物。
Stamets's polypore library was screened by NIH NIAID and Project BioShield collaborators against bioterror priority agents. Several extracts showed in-vitro activity against pox-, flu-, and herpes-virus families — the same classes documented in his 2005 Agarikon publication. The screening is documented. A clinical antiviral drug has not yet emerged from this line as of 2026. Peer-reviewed papers, patents, and Phase 1-grade mechanism work have come of it. A blockbuster drug has not. The pipeline distance between in-vitro hit and approved drug is, again, where most of pharmacology lives.
斯塔梅茨的多孔菌库被NIH下属NIAID（国家过敏和传染病研究所）以及Project BioShield合作方拿去筛查生物恐怖优先级病原。多种提取物在体外对痘病毒、流感病毒、疱疹病毒科显示出活性——与他2005年阿加里孔论文记录的同一类别。这段筛查有公开记录。截至2026年，从这条线还没有成形的临床抗病毒药上市。结出了一批同行评审论文、一批专利、以及一期级别的机制研究——但还没有重磅药物。从"试管命中"走到"已批准药物"这条距离，又一次落在药学行业存在的理由里。
Part II — Speculation, Synthesis & Open Question
第二部分 —— 推测、综合判分、开放问题
Stamets's Speculations ▶ 115:00
斯塔梅茨的推测前沿 ▶ 115:00
A neurological metaphor that isn't biology. A panspermia leap that requires another leap. A cognitive-evolution story that lacks evidence.
一种被叫作"神经"的比喻——但它不是生物学。一种"泛种论"的跳跃——而它还需要再跳一次。一种关于认知进化的故事——但它没有证据。
"Neurological network of nature." Factual basis real. The leap to "neurological" — metaphor.
「大自然的神经网络」。事实基础是真的。"神经"那一步——是比喻。
Stamets returns to this metaphor repeatedly throughout the conversation. The factual basis is real: mycelial networks do transport nutrients between trees of different species (Simard 1997 Nature et seq.); they do respond to environmental signals; some networks span thousands of acres. The leap from "responsive distributed network" to "neurological" is metaphor, not biology. Hyphal cells have no synapses, no neurotransmitters, no central processing. The metaphor is generative — it has produced research questions worth investigating — but it is not a literal description. The synthesis table grades it accordingly.
斯塔梅茨在整场对话中反复回到这个比喻。事实基础是真的：菌根网络确实在不同树种之间转运养分（Simard 1997《Nature》及其后续研究）；它们确实对环境信号有响应；部分网络跨越数千英亩。但从"分布式响应网络"跳到"神经网络"，是比喻，不是生物学。菌丝细胞里没有突触，没有神经递质，也没有中枢处理。这个比喻很有启发——确实带出过值得研究的问题——但它不是字面描述。文末梳理表对此评级。
Spores survive low-Earth orbit (confirmed). Spores seeding planets across interstellar space — unsupported.
孢子在低地球轨道存活（已确认）。孢子能跨星际距离播种行星——没有支撑。
Stamets's farthest claim draws from Hoyle and Wickramasinghe's panspermia hypothesis: that fungal spores, with tough chitinous walls, could survive interstellar travel and seed life on planets they reach. The spore durability part is partly true — certain fungal and lichen spores survive vacuum and low temperatures; the BIOMEX experiment (ESA EXPOSE-R2, ISS exterior, 2014–2016) confirmed survival of certain lichens after ~18 months of space exposure. The leap from "spores survive low-Earth orbit" to "spores seed planets across interstellar distances" is unsupported. Panspermia is speculative; mushroom-vehicled panspermia is speculation on speculation.
斯塔梅茨最远的推测来自霍伊尔和维克拉马辛格（Hoyle & Wickramasinghe）的泛种论：真菌孢子壁含几丁质、足够坚韧，理论上能熬过星际航行并在所到行星播种生命。孢子耐受性这部分有一点事实基础——某些真菌和地衣孢子确实在真空和低温下存活；欧洲航天局BIOMEX实验（EXPOSE-R2平台，国际空间站外部，2014–2016年）确认了部分地衣在约18个月空间暴露后复苏。但从"孢子在低地球轨道存活"跳到"孢子能跨星际距离播种行星"，没有支撑。泛种论本身是推测；以蘑菇为载体的泛种论，是推测之上的推测。
Psilocybin mushrooms drove rapid human brain expansion. No archaeological or genetic support.
裸盖菇素蘑菇推动了人脑的快速扩张。没有考古学或遗传学支持。
Terence McKenna's Food of the Gods (1992): early Homo sapiens on the African savanna stumbled across psilocybin-containing mushrooms in cattle dung, and the resulting cognitive plasticity drove rapid brain expansion. Stamets references this approvingly. The hypothesis has no archaeological or genetic support. The brain-expansion timing predates the savanna-dwelling phase the hypothesis depends on; there are no preserved psilocybin-mushroom remains in the relevant strata; the genetic signal of cognitive change is gradual, not sudden. It is a charming origin story. On present evidence, it is not a true one.
特伦斯·麦肯纳《Food of the Gods》（1992）：非洲草原上的早期智人在牛粪堆里偶遇含裸盖菇素的蘑菇，由此带来的认知可塑性推动了人脑的快速扩张。斯塔梅茨以认可的口吻引用了它。这套假说没有考古学也没有遗传学层面的支持：人脑快速扩张的时间线早于假说依赖的"非洲草原阶段"；相关地层里没有裸盖菇素蘑菇的化石残留；遗传层面看到的认知变化信号是渐进的，不是骤变。这是一个迷人的起源故事，但按目前的证据，它不是真的。
Stamets is simultaneously the lead author on a Scientific Reports bee-virus paper, the holder of multiple antiviral patents, and a man who tells the JRE audience a sincere story about Star Trek writers giving him the bee-virus discovery in a dream [118:34]. Both are real about him. The Spark refuses to render him as either pure scientist or pure mystic. The synthesis is what does justice to both.
斯塔梅茨同时是一篇《Scientific Reports》蜜蜂病毒论文的通讯作者、多项抗病毒专利的持有人——也是那个会在乔·罗根面前认真讲述「《星际迷航》编剧在梦里把蜜蜂病毒研究的灵感塞给我」[118:34] 的同一个人。两面都真实。本篇拒绝把他写成"纯粹的科学家"或"纯粹的神秘主义者"——而是用下面这张梳理表，对两面都有交代。
Claim vs. Published Evidence
逐条主张 vs. 已发表的证据
Six claims. Each row pairs one of Stamets's specific claims with the strength of the published evidence as of 2026 — and what is actually on the record.
六条主张。每一行把斯塔梅茨的一项具体主张，与截至 2026 年已发表证据的强度，并排放在一起——并写出公开记录里"实际能查到的内容"。
Four evidence levels
四种证据等级
- Established Established
- Peer-reviewed, replicated, or clinically authoritative on the specific claim.
- Preliminary Preliminary
- Early-stage signal — small trial, single lab, animal model. Replication needed.
- Speculative Speculative
- Plausible but not yet supported by peer-reviewed evidence; or evidence has so far been negative.
- Metaphor Metaphor
- Evocative framing, not a literal biological claim.
- 扎实 扎实
- 就该具体主张而言，已有同行评审、可重复，或临床层面的权威证据。
- 初步 初步
- 早期信号——小样本、单一实验室、动物模型。需要重复验证。
- 推测 推测
- 有合理性，但尚无同行评审证据支持；或现有证据为阴性。
- 隐喻 隐喻
- 富有启发的框架，但不是字面意义上的生物学描述。
Two of Stamets's six core claims are established on solid published ground (Agarikon antivirals in vitro; bee-virus mycelium-extract reduction in field trials). One is established at the clinical level (macrodose psilocybin trials). One is preliminary (lion's mane in human cognition). One is speculative, and has been tested with negative results so far (the Stamets Stack). One is metaphor rather than biology (mycelium-as-neurological-network). The published record runs further than skeptics generally credit. The headline runs further still.
六条核心主张里：两条建立在扎实的已发表基础上（阿加里孔体外抗病毒；菌丝提取物在田间试验中显著降低蜜蜂体内病毒）；一条在临床层面已经扎实（大剂量裸盖菇素试验）；一条仍是初步（狮鬃菇对人体认知的作用）；一条属于推测，且现有研究已经给出阴性结果（Stamets Stack）；一条更接近隐喻而不是生物学（菌丝是神经网络）。已发表的论文跑得比怀疑者一般以为的要远；标题党跑得还要更远一些。
The Verdict Is Not the Spark's Job
下判决不是这篇文章该做的事
The title of this Spark is the title Stamets has given his life's work: could mushrooms save the world? The published record says: in some specific, narrow, well-defined ways, mushrooms are already helping save parts of the world — bees that would otherwise die from RNA viruses, soils that would otherwise stay contaminated, possibly the depressed and dying who would otherwise stay depressed and dying. The headline-grade version of the question — whether the strands underground are a kind of planetary intelligence preparing to rescue us from ourselves — is still up for grabs.
这篇文章的标题，就是斯塔梅茨给自己一生工作贴的标签：蘑菇能拯救世界吗？已发表的论文给出的回答是：在某些具体、范围很小、定义很清楚的层面上，蘑菇已经在拯救世界的一部分——它们正在救一些本来会死于 RNA 病毒的蜜蜂、清理一些本来会一直被污染下去的土壤，可能还在拯救一些本来会一直困在抑郁和临终焦虑里的人。但标题党级别的那一问——地下那些丝是不是某种"行星级智慧"，正准备把我们从我们自己手里救出来？——这一问，到今天还没有结论。
The synthesis above gives readers what to watch. Bees first. Macrodose psilocybin trials next. Agarikon and lion's mane on a longer pharmaceutical timeline. The Stamets Stack, possibly never. The neurological-network metaphor — generative but not literal.
上面那张梳理表，把"接下来要看什么"列得清清楚楚。先看蜜蜂；其次是大剂量裸盖菇素的临床试验；阿加里孔和狮鬃菇要走更长的制药线；Stamets Stack 也许永远不会"跑赢";"神经网络"这层比喻——继续作为有启发的比喻存在，但不要把它当字面读。
So: which of Stamets's specific claims do you think will, in ten years, have moved from "preliminary" or "speculative" to "established" — and which do you think will stay where they are?
所以：斯塔梅茨的这些具体主张里，你觉得十年之后，哪一些会从"初步"或"推测"挪到"扎实"？又有哪一些，会一直停在原地？
The Spark refuses to answer that for you. The synthesis above is what the published record currently says. The verdict is yours.
这篇文章不替你回答这个问题。上面那张梳理表，已经把已发表的论文目前能说的都说完了。剩下的判断，留给你自己。