週報~月刊 ,2024

2024年6月刊

AI粵語；Mexico Elects；民國革命場景；三體訪問；後真相時代、私域流量；3M with PFAS；Insulin Empire；高考；OpenAI suck；祖姑崇拜；無體温關係；Boys struggle；NBA Foster；地方治理信用透支；taste；

2cold

Jun 16, 2024 16:16 UTC+8 • 49 min read

AI粵語

Tan Lee 自 1990 年代初開始致力於語音語言相關的研究，領導開發了一系列以粵語為核心的口語技術，並得到了廣泛的應用。他在 2002 年與團隊合作推出的粵語語音語料庫 CU Corpora，是彼時世界同類數據庫中最大的，包含兩千多人的錄音數據。蘋果的第一代語音識別在內，許多公司和研究機構希望開發粵語功能時，都曾向他們購買這套資源。
在他看來，ChatGPT 的廣東話語音表現「水平不是很好，主要是不穩定，聲音的質量、發音的準確性整體都不是讓人很滿意」。但這種表現不佳並非源於技術侷限。實際上，如今市面上許多具備廣東話能力的語音生成產品，質量都要遠高於此。以至於他對網絡視頻中 ChatGPT 的表現感到難以置信，一度以為是深度仿冒的贗品，「如果是做語音生成模型的，做成這樣基本不能見人，等於自殺」。
以香港中文大學自身開發的系統為例，最先進的一批在語音效果上已經很難分辨是真人還是合成的聲音。與普通話和英語等更強勢的語言相比，AI廣東話只在一些更個性化和生活化的場景中，情感表現會遜色一些，比如在父母與孩子的對話、心理諮詢、工作面試的場景中，廣東話會顯得比較冰冷。
「但嚴格來講，在技術上這並沒有什麼難度，關鍵在於社會資源的選擇。」 Tan Lee 表示。
相比於 20 年前，語音合成領域已經發生了翻天覆地的變化，CU Corpora 的數據量跟如今的數據庫相比「可能還不到萬分之一」。語音技術的商業化讓數據成為了一種市場資源，只要願意，數據公司隨時可以提供大量的定製數據。而廣東話作為口語化語言，文本與語音的平行數據缺少的問題，近年來隨着語音識別技術的發展，也已經不再是一個問題。在當下，廣東話作為「低資源語言」的說法，Tan Lee 認為已經不再準確。
也正是因此，在他看來，市面上機器的廣東話表現反映的並非是技術的能力，而是市場與商業的考慮。「假設現在全中國一起學廣東話，那肯定可以做起來；又比如，現在香港跟內地越來越融合，假設有一天教育政策變成，香港的中小學不能用廣東話，只能說普通話，那就又會是另外一個故事了。」
「吃下什麼便吐出什麼」的深度學習展現出的口音，實際上是廣東話在現實空間受到的擠壓。

諗諗下都係，當然唔係技術問題，技術層面英文、中文嘅性能明顯好好多，同埋文中都有介紹一啲已經好好效果嘅AI。所以其實係社會資源嘅選擇：有更大嘅需求，就會鼓勵更好嘅質量，現實層面都的確粵語嘅地位越嚟越式微同唔受重視，而其他語言就更係如此（強調多一次唔係方言！詳見語言學嘅定義）

語言的幻象不僅存在於粵語中。Reddit 論壇與 OpenAI 的討論區，來自世界各地的用戶都反映了 ChatGPT 在說非英語語言時存在類似表現：
「它的意大利語語音識別非常好，總是能聽懂且表達流利，就像一個真人。但奇怪的是，它有英國口音，就像一個英國人在說意大利語。」
「本英國人表示，它有美國口音。我很討厭這一點，所以我選擇不用。」
「荷蘭語也是，很煩人，彷彿它的發音是用英語音素訓練出來的。」
語言學上，將口音定義為一種發音方式，每個人受到地理環境、社會階層等因素影響，都或多或少會有發音選擇上的差異，這常常體現在音調、重音或詞彙選擇上的不同。有趣的是，過去被廣泛提及的一些口音，大多源於世界各地的人試圖掌握英語時從母語中攜帶而來的習慣，比如印度口音、新加坡口音、愛爾蘭口音——這反映了世界語言的多樣性。但人工智能展現出的，則是主流語言對區域性語言的曲解和反向入侵。

幾年前，因為個人精力有限，黃冠能停止了 Ekho 對安卓系統版本的維護，但停了一段時間，突然又有用戶跑來希望他將其恢復。他才得知，如今安卓系統已經沒有免費的粵語 TTS 可用了。
用當下的眼光看來，黃冠能開發的 Ekho 採用的已經是完全落伍的技術，但仍具有獨特之處。作為本土的獨立開發者，他在設計時帶入了對於這個語言的切身經驗。他記錄的廣東話包含了七個聲調，其中第七個是香港語言學會提出的 Jyutping （粵拼）中不存在的一個發音。「『煙』這個詞在『抽煙』和『煙火』中，會發出不同的聲調，也就是第一聲和第七聲。」
在整理發音字典時，他曾請教過 Jyutping 的研發者，得知隨着時代變化，年輕一代的香港人不再分辨第一聲與第七聲的區別，這個音也因此逐漸消失了。但他仍選擇將第七音納入，這並非出於公認的標準，只是他個人的情感記憶，「土生土長的廣州人是可以聽出來的，現在使用還是非常普遍」。
只聽到這個音，老廣便能分辨，你是本地人還是外來的。

慘，第一時間諗緊自己識唔識分……但一下子就反應過嚟，確實係唔同㗎，笑死，差啲以為自己唔識

Mexico Elects

Each country has its own dynamics, of course, but we will highlight some strands that connect Mexican politics to politics elsewhere. Regular readers of The Morning will notice a familiar theme: Populism is ascendant.

Beyond economic results, AMLO governs in a feisty style that voters seem to enjoy. “AMLO has forged a connection with many Mexican voters, which is rare in the country’s recent history,” our colleague Simon says. “He’s a combative communicator who seizes control of the narrative with a folksy, us-versus-them style.”
He cut government salaries and cracked down on corporate tax avoidance. He has taken the public bus as president, speaks in an informal version of Spanish and harshly criticizes Mexico’s economic elites, Viri Ríos, a political scientist, has noted. AMLO also increased oil production, rejecting environmental concerns.

政客本身就係要討好選民，所以噉做都好合理？而問題就在於其他民粹言行：

AMLO has called the judiciary “rotten” and claimed some judges were part of a conservative conspiracy “dedicated to plundering the country.” He disparages journalists whose coverage he doesn’t like and runs weekly sessions devoted to “lies.” He has given the military additional powers, including policing many parts of the country.
His reliance on the military has contributed to his biggest failure because it has been ineffective at protecting public safety. Crime has risen, and cartels control parts of the country. A Times analysis found that 36 political candidates had been killed since last summer, as had some relatives of candidates.

She will also inherit some problems AMLO created, such as a big budget deficit. And she lacks his personal popularity. AMLO founded his and Sheinbaum’s political party, known as Morena, and then enacted popular policies as president.
As Natalie Kitroeff, The Times’s Mexico City bureau chief, has written about Sheinbaum: “Will she dare to stray from those policies if she wins office, inviting the reproach of Mr. López Obrador and the movement that got her there? Or will she dedicate herself to cementing his legacy, even if it means stifling her own vision?”

呢個都係選舉成功之後嘅問題，係要延續政策以回饋支持自己嘅選民但同時等於綁實咗冇進一步發揮嘅空間，定係做多啲自己想做嘅事？

民國革命場景

革命，乃變革天命，甚至逆水行舟。可惜無論辛亥革命時期、國民革命時期還是共產革命時期，都未有如此想像力。於是革命，在上層變成了皇位輪替，在下層恐懼的目光中，變成了殺頭與掠奪。也許未來，革命將愈加成為一個幽靈，一個遠離中國想像的幽靈。

真係噉咩？革命者中應該唔缺理想主義者，個問題似乎在於，中國未試過真正由底而上成功革命過？

三體訪問

“In the realm of science fiction, Arthur C. Clarke’s ‘2001: A Space Odyssey’ stands out. As for literature in general, Leo Tolstoy’s ‘War and Peace’ left an indelible impression. Its sweeping portrayal of historical epochs and the profound weight of its narrative influenced my creative endeavors greatly. It’s a novel with a panoramic view of depicting an era. Nowadays, many writers become immersed solely in their circles, focusing even exclusively on personal experiences. They lack the ability to unfold grand narratives and possess narrow perspectives.”

所謂蘿蔔青菜各有所好，又唔覺邊樣更好或更高尚，當然，我更加反對隱瞞甚至洗白前事嘅做法

“It was necessary to mention the event to develop the story. The plot required a scenario where a modern Chinese person becomes completely disillusioned with humanity, and no other event in modern Chinese history seemed appropriate except the Cultural Revolution. It is disappointing that most participants of the Cultural Revolution did not repent, and the reason behind it is unclear.”

恐怕永遠都唔會解密唔會有答案……

後真相時代、私域流量

有陣時都幾難明點解呢啲會成為爆點新聞（今次仲要係假嘅），我無意關注但都有畀接觸到……

但这也许并不影响很多人寻找下一个胖猫和谭竹。
他们并不关心胖猫，只是关心自己可不可以借机泄愤；他们也不关心谭竹，只是关心自己有没有一个靶子。
他们不是在胖猫事件上充满正义，只是需要让自己的廉价正义感表演有附着物——刚好炒作与算法顺着「共情」网络将胖猫事件推了出来，让死去的胖猫成了「二次消费」标的而已。

专栏作家戴维·布鲁克斯说；互联网并没有改变人类无知的状态，反而让人们进入了一个「新无知时代」——无数未经提炼和归纳的信息碎片，正在被不断创造出来，犹如「一条浑浊的信息河流」，大量缺乏判断力的受众浸泡其中，对偏见、虚假消息不加辨别地吸收。
这段话，一下子点了太多人的穴。
对他们来说，吃一堑连着的往往不是长一智，而是再长一堑……长一智是不可能长一智的，这辈子都不可能长一智的，多看点书又不会，就只有去喷这样子才能够维持精神生活。

我都要講呢，其實我都……無幾耐之前先肥佬咗——關於某球員媽媽去咗宏遠主場嘅fake news。sigh，大概都真係冇乜靠得住，自媒體肯定唔理會，大平台都唔一定查證過，連官方都可能報啲唔報啲嘅，都係老老實實返做好fact check……奇怪，係幾時變到咁懶嘅呢？

村上春树在短篇小说《列克星敦的幽灵》中写道：
「我真正害怕的是那些毫无批判地接受和全盘相信别人说法的人们，是那些不制造也不理解什么而是一味地随着别人听起来顺耳的容易接受的意见之鼓点集体起舞的人们。他们半点都不考虑，哪怕是一闪之念，自己所作所为是否有错，根本想不到自己能无谓地致命地伤害一个人。我真正害怕的是这些人。」
这些偏听盲信者如同墙头杂草，性别对立的心态则是一把火，能快速将其点燃，最终吞噬无数个胖猫和谭竹。
这带来的直接问题是舆论场的「垃圾场化」，深层次弊害则是「真问题失焦」——正如法学家劳东燕所说的，公共事件逐渐从公域里消隐，越来越多的私人事务被拉到公域，在显微镜下被围观被细看，出现任何瑕疵都会被上纲上线。当「监督」的箭头一边从某些地方退缩一边向私生活领域延伸，社会注定会被一场又一场的舆论泡沫粘住。
一通没有意义的狂欢，留下一地没有价值的鸡毛，就是我们公共生活里的常态景象。
回到胖猫事件上，它就成了后真相时代被众人加工的标本。虽然许多人并不在意真相，但它暴露出了几重真相。
如果说第一重真相是警方通报还原的胖猫生前有情感寄托、死后被拿来炒作，第二重真相是无数哄客不关心胖猫也不关心谭竹，只是用他们来为自己的伪正义癖好和表演欲献祭。
那胖猫事件背后真相的真相的真相，就是真相濒死——它可能被掩埋在畸形信息供给的深坑里，也可能溺死在情绪洪流之下。
这，或许就是我们不得不直面的最残酷的真相。

我只希望少啲接觸到呢啲垃圾私域公開流量……

但係係點樣開始嘅呢？

微博時代開始，一大批公共知識分貢獻了這個平台最初的流量，也是他們最後的時光，隨着中共對互聯網監督的“成熟”，公共知識分子在輿論場上被清除，甚至“公知”這個詞也被嚴重污名化了。“圍觀”不再總是有效，而是變得危險。
在微博上，取代公共知識分子的是娛樂明星。儘管大部分明星的微博，可能都是由助理打理，並不是他們本人。但是他們粉絲眾多，流量更大，這也讓新浪微博看到“盈利”的曙光：與“娛樂”有關的信息可以被看成是中性的，它甚至是健康生活的一個標誌。明星作為公共人物，讓渡一部分個人隱私，八卦和緋聞滿天飛，有人甚至為了博取關注度，故意炒作不存在的緋聞。觀眾對此無所謂，因為他們並不損失什麼，獲得的則是“開心一刻”。這就是“吃瓜”文化。
但是，人們很快會發現，即便是人畜無害的娛樂明星，一旦粉絲和影響力達到一定級別，也會成為黨打擊的對象。儘管明星們已經習慣於在“七一”、“十一”這樣的“節日”表忠心，但是忠誠度總嫌不夠。如果回顧2015年以來的微博，會發現已經有很多明星消失了，原因可能是税務問題，也可能是言論不當。
接下來走上祭壇的是普通人。官員的腐朽生活，當然也引人關注，但是它不再像過去那樣引來調查，而是通往404——不但從熱搜中撤下來，也會成為敏感詞。有關注的社會名人、公司高管，逐漸成為“吃瓜”的對象。先是“名人”的私生活，接下來是普通人的“極端生活”，陸續填補到流量的真空。
人們的關注度，最終鎖定普通人。他們的“極端生活”，本來是應該受到“絕對保護”的個人隱私，此事不得不獻祭出來。“瓜眾”對個人生活的無限窺探，當然有心理基礎，但是更重要原因是，這樣做很少受到法律追究和權力干預，在移動互聯網時代，這成為新的流量密碼。

世界上並不存在一個恆定的公共資源，討論一個事情，也並不影響到其他事件的討論——如果沒有外力干預的話。中國的“公共領域”確實在萎縮，但是這並不是任何個體的錯。
公共領域的發生和擴展，離不開對權力的監督。權力被監督得越厲害，公共空間就會越大。“把權力關進籠子裏”是一個很好的比喻，如果權力在籠子裏，整個“籠”外的世界就都是公共領域，理論上可以無限大。
當然，這只是一個比喻，它沒有在現實中發生。實際上，中國人公共生活的不斷萎縮，恰恰是“被”權力干預的結果。敏感詞越來越多，不能討論的問題越來越多，人們就只能聚焦於那些沒有公共價值的個人生活，全民“窺淫”背後，有一隻無形的大手。

在當下的媒體環境裏，他們只會不斷調適自己，認為“一切責任”都在監管者，自己只是奉命行事。
中國並沒有存在過真正獨立的媒體。市場化媒體最好的時候，曾經擁有一定的空間，但是“主管部門”依然存在。現在，這個空間已經喪失殆盡，媒體領導者幾乎全部“官員化”，他們已經習慣於嚴格的自我審查，甚至完成了和宣傳部“思想同構”，不再有創造空間做一點報道的想法。
每次類似“胖貓跳江”這樣的事件，公權力總是勝利的一方。如果用一個表情符號來形容它們，應該是“偷笑”。這次重慶警方以真相發佈者的面目出現，網友要麼感激（終於看到了兩人的詳細轉賬記錄），要麼道歉（此前對譚某的網暴草率），而真正值得注意的是，一些人又被封號了。
這是常見的結局和“成果”。這是一個事件的結束，也是下一個事件的開始。在封號的陰影下，對公權力的猜忌和監督，又弱了一分，而對個人生活的審判，又會嚴厲一分。

歸根到底，就係冇獨立自由嘅公共領域土壤

3M with PFAS

係時候將所有真相公開，因為遲啲就趕唔切，化石燃料公司隱瞞温室氣體全球變暖，而家嘅氣候反常到冇得再反常；而3M隱瞞PFAS呢種永久性物質有害，危害咗幾多人

She reminded herself of what he had said — that the chemical wasn’t harmful in factory workers. But she couldn’t be sure that it was harmless. She knew that PCBs, for example, were mass-produced for years before studies showed that they accumulate in the food chain and cause a range of health issues, including damage to the brain. The most reliable way to gauge the safety of chemicals is to study them over time, in animals and, if possible, in humans.
What Hansen didn’t know was that 3M had already conducted animal studies — two decades earlier. They had shown PFOS to be toxic, yet the results remained secret, even to many at the company.

right，需要求證，唔可以直接相信權威；而事實上，大公司又真係做過實驗但為咗自身利益而罔顧公眾利益選擇隱瞞甚至宣稱無害，xxxx

Rats that had more fish meal in their diets, she discovered, tended to have higher levels of PFOS, suggesting that the chemical had spread through the food chain and perhaps through water. In male lab rats, PFOS levels rose with age, indicating that the chemical accumulated in the body. But, curiously, in female rats the levels sometimes fell. Hansen was unsettled when toxicology reports indicated why: Mother rats seemed to be offloading the chemical to their pups. Exposure to PFOS could begin before birth.

This year, a study discovered 13 forever chemicals, including PFOS, in weeks-old fetuses from terminated pregnancies and linked the chemicals to biomarkers associated with liver problems. A team of New York University researchers estimated in 2018 that the costs of just two forever chemicals, PFOA and PFOS — in terms of disease burden, disability and health-care expenses — amounted to as much as $62 billion in a single year. This exceeds the current market value of 3M.

恐怖……

It may be tempting to think of Creacey and his peers as unwitting research subjects; indeed, recent studies show that PFOS is associated with an increased risk of thyroid cancer and, in Air Force servicemen, an elevated risk of testicular cancer. But it is probably more accurate to say that we are all part of the experiment. Average levels of PFOS are falling, but nearly all people have at least one forever chemical in their blood, according to the Centers for Disease Control and Prevention. “When you have a contaminated site, you can clean it up,” Elsie Sunderland, an environmental chemist at Harvard University, told me. “When you ubiquitously introduce a toxicant at a global scale, so that it’s detectable in everyone ... we’re reducing public health on an incredibly large scale.” Once everyone’s blood is contaminated, there is no control group with which to compare, making it difficult to establish responsibility.

sad but true，再冇純淨嘅樣本，舊時嘅樣本遲早會用晒。and so sad that it's forever (as least in our life)

Insulin Empire

The potential for insulin’s market exploitation was almost presciently understood by Banting and his team at the University of Toronto, so in 1923, when Banting and Best were awarded the U.S. patents for insulin and the method for making it, they swiftly sold them to the university for $1 each. “Insulin does not belong to me, it belongs to the world,” Banting explained, believing that profiting off such an essential treatment was not only immoral but detrimental to ensuring universal affordability and access. He and Macleod were awarded the Nobel Prize in Physiology or Medicine later that year.

respect，發明人係理想嘅，但現實……

Despite their efforts, things have not gone as the scientists envisioned. The major reason for this can be traced back to the capture of insulin manufacture worldwide by an oligopoly of multinational pharmaceutical firms known as the Big Three—Eli Lilly, Novo Nordisk, and Sanofi—which control more than 90 percent of global insulin markets and the entirety of the U.S. insulin market. At times involved in lockstep collusion schemes, and at others, in transparent attempts to copy one another’s products, the Big Three have thrived in no small part because of obsequious regulations, loopholes in intellectual property law, and an odious health care system, pushing a miracle drug out of reach for millions of people who need it.

Things began to change in the 1970s, as the pharmaceutical industry embarked on its glorious struggle to become the most profitable sector of our economy. It was within this political-economic reconfiguration that a new parasitic industry in the pharmaceutical pricing regime emerged: the pharmacy benefit manager, or PBM.
PBMs are entities working with drugmakers and insurance companies to determine which drugs will receive insurance coverage or be included on formularies (lists of drugs covered by health insurance plans), and to negotiate price cuts in the process. Initially formed to provide a much-needed solution to insurers that were unable to efficiently respond to claims as drug coverage increased, the arrangement became a privatized scheme in which drugmakers and insurance companies could acquire and control PBMs of their own. Rarely are the discounts they negotiate passed along to the customers (i.e., the people who need said medicines to live); they’re more often pocketed by these managers instead.

Insulin producers looked to the general drug industry, where academic and industry labs focused on “orphan diseases”—rare diseases affecting smaller patient populations—including those linked to specific genetic mutations that new drugs could target. “Scientists began dividing broad disease categories into various sub-types. . . . Treatment varied accordingly,” Goozner writes. It was elegant science, but it also meant the “patient population for any given drug shrank. That’s why most of the targeted medicines developed over the past two decades have come from small, venture capital-funded biotech firms started by scientists whose original research was funded by the National Institutes of Health (NIH), charitable foundations, patient groups, or some combination of those resources.” Big drug companies then buy these biotech firms from venture capitalists at inflated prices, often with drugs already deep in the FDA-approval process. To compensate for the inflated cost, these new owners hike prices for drugs targeting small populations.
It’s in this environment that our insulin oligopoly found its groove. The evergreening of Eli Lilly, Sanofi, and Novo Nordisk’s synthetic insulin has allowed prices for the hormone to soar despite there being no evidence that each new version is any better than the last. They’ve also been successful in rallying opposition to legislation that supports biosimilar generic competition. Biosimilar firms are subject to additional regulatory scrutiny; they must demonstrate that there are no meaningful differences between their products and the original biologic—usually through costly clinical trials. While Europe approved its first “biosimilar” in 2006, the U.S. pharmaceutical industry stalled progress on biosimilar regulation until it was sufficiently favorable to incumbent firms, after which a spurt of “ferocious lobbying” shaped and passed the Biologics Price Competition and Innovation Act in 2009. It wasn’t until 2015 that the FDA approved the first biosimilar, and only by 2019 did it issue its final guidance on allowing biosimilar interchangeability at pharmacies, which is key to letting them replace brand-name drugs.
The crux, as Goozner points out, is that “biosimilar” is industry speak crafted by pharmaceutical firms to convince the public that biologic drugs like insulin—which are produced using a living organism—can’t be generics. In a 2017 research paper for the South Centre, Germán Velásquez observed, “In any debate on the impossibility of producing ‘identical’ drugs, it should be made clear that what is at stake is not identical products but therapeutic equivalents. What matters to the patient, as we have said, is whether or not a drug can prevent, cure or mitigate the effects of the illness.” We can understand the moves by the U.S. insulin cartel’s war on insulin, then, as a war to remake insulin into a more profitable form, its patients into a more profitable demographic, and its market into a noncompetitive oligopoly.

But just as insulin unmistakably, and almost miraculously, transforms a sick body, it has the potential to reconstitute our political economic realities. Medical anthropologist Samantha Gottlieb gestures towards the “paradigm-shifting” potential of the “fantastical empowered” insulin-dependent patient. Unlike the engaged, “compliant” patient of the regulatory-clinical-commercial public health imaginary, the insulin-dependent patient is an “actor-creator and disrupter.” Diabetics’ hands remain full as they traipse through multibillion-dollar mazes just to make it through each day, but, as diabetes activists on X have expressed so cogently, #WeAreNotWaiting. When insulin pumps and CGMs were faulty and incompatible, hackers developed a do-it-yourself device looping system for more personalized and precise blood glucose management; when pharma announced insulin shortages, Mutual Aid Diabetes and other groups have worked to give excess supplies to diabetics in need; when people died from rationing their insulin, protestors have demanded justice and gotten life-saving, emergency access provisions written into state legislation. While the machinations of the pharmaceutical empire and its profiteering bedfellows continue to privilege profits over people, we continue to keep ourselves cared for and alive; we continue to demand guaranteed, protected, and dignified access pathways for insulin and all other drugs and medical technologies people need—not just to survive but to live full and thriving lives.

love this ending, the hope exists because of community

高考

我还是恨高考，我希望没有人再经历它，但还是要目睹每年“超多少多少考生”往里跳，就好像那是一个什么庆典，所有的大人们谈笑风生，翘首以盼，唯独忘了自己是怎么从里面爬出来的——或许这也有点勉强，从中国的教育情况来看，虽然每年参加高考的人数惊人，但是在高考之前就被教育甩出去的人数只会更加惊人。

too bad but so true...事到如今，雖然都唔否認其公平性，但始終覺得係suck

我想中国今天这种局面，政治高压、经济下行，最痛苦的无疑是未成年人，所有的压力都向最弱者倾斜，毫不夸张地说，他们从睁开眼的那一刻起，就开始面对整个社会的压力。

既然这个社会能将整体的压力投射到孩子们身上，又怎么不允许孩子们的恨意存在。老实说我宁愿接受他们的恨，也不愿意看到他们自戕。
我受够了千百年来的哪吒叙事，以自戕实现反抗，这太惨痛了。然而没有意义，厚颜无耻的大人们还在等待新的孩童，还将观赏他们新一轮的大逃杀游戏。
孩子的自杀就和夭折一样普遍，且能够被“成熟”的大人们接受，为了某种“效益”的实现，他们还将生产出另一个孩子，另一些孩子——就像那个新闻里，弟弟自杀后，父母决定代孕再生一个儿子，最后被女儿举报而告终。

我自己来到这个世上都够烦了，我真的很难忍受所有不被爱的孩子们被制造出来的前前后后，只能看到人类男性糟糕的繁殖欲望。他们得到孩子，然后折磨孩子，就是这么回事。
你以为我在这里是写我与妹妹的关系，实际上我可能是在写我几乎素未谋面的爹的关系：不存在，但又几乎无处不在。只是他早已不值得我用“百感交集、层次复杂”来形容，索性到此为止。

OpenAI suck

Setting aside the legal questions here, such behavior would align with some of the harshest criticism of Sam Altman and OpenAI — that it’s a company with little regard for the value of creative work led by a scheming, untrustworthy operator. This episode also complicates the company’s preferred narrative of unstoppable inevitability: You’re either the company harnessing the barely controlled phenomenon of imminent self-replicating machine intelligence, leading humanity into its next technological epoch, or you’re a mid-stage start-up that for some reason really needs to copy that voice from that movie to market an incremental product upgrade.
OpenAI’s lopsided public dispute with one of the most recognizable human beings on earth is the latest in a string of episodes in which OpenAI and Sam Altman have struggled to keep their stories straight. Shortly after OpenAI’s demo, news broke that the leaders and much of the staff associated with its “superalignment” team — a group that was established last year and tasked with figuring out how to “steer and control AI systems much smarter than us” — had resigned, among them OpenAI co-founder and former chief scientist Ilya Sutskever. With the exception of superalignment head Jan Leike, who said that he had been “disagreeing with OpenAI leadership about the company’s core priorities for quite some time,” most of the resignations were quiet or tersely announced — by necessity, it turned out, as Vox reported that they’d signed an “extremely restrictive off-boarding agreement that contains nondisclosure and non-disparagement provisions” that “forbids them, for the rest of their lives, from criticizing their former employer.” Altman responded that he was “genuinely embarrassed” about how this had played out, but also pled ignorance.

完全可以講，已經偏離咗初衷。無法否認，營銷手法確實幾好，但初時嘅安全性、謹慎同道德呢？nowhere

If you’re fully invested in the narrative that OpenAI is the tech company to end all tech companies — the one firm that can bring about artificial general intelligence, a rightful steward of a future that most people can’t comprehend — then you might worry that OpenAI has turned away from safety and alignment toward uninhibited AI development, consequences be damned. But the company’s recent behavior, and that of its CEO, who was briefly deposed by a board that accused him of a lack of “candor” before clawing his way back to power, is also consistent with that of a firm that’s facing a lot of competition whose product is less differentiated than ever, and which pronounced its commitment to safety mainly just to imply how powerful it would one day become: a company that has raised huge amounts of money not just on the strength of its technology or a clearly defined business model, but with a series of grand — and suspiciously familiar! — stories about the future, drawn from fiction of the past.

唔知未來會點，但好不容樂觀……so sad to know that

祖姑崇拜

實際上，早在何仙姑信仰流傳以前，廣東地區自古就有女子不嫁人的風俗，很多家族歷史上都曾出現不婚的女性成員。
在傳統農業社會，女性是家中重要的勞動力，尤其在不富裕的家庭，很多女性為了維持家計、或照顧幼年失怙的弟妹而終身不嫁。至於菁英家庭的女兒也未必要進入婚姻才能讓後半輩子有所依托。
早期這些未嫁女姓的兄弟及後人為了感念其貢獻，或不想讓親人魂無所依，會在她們往生後供奉香火。然而，到了明清時期，當地逐漸成為宗族體制盛行的社會，想繼續祭祀未嫁女性就沒這麼簡單了。
在宗族體制下，這些女性違背了以夫家為依歸的理想，連原生家庭都不能在家中神龕或宗族祠堂供奉牌位。為了讓地方上既有的未嫁女性祭祀傳統，得以在宗族體制下延續，勢必得想出一套不違反宗族價值觀的邏輯。

換句話說，地方社會想到的理由是，這些女子因成功維繫家族男性血脈，或因其他足以光宗耀祖的事蹟，才破例以「祖姑」身分納入宗族祭祀。
何仙姑信仰的正當性，便是在泛珠江三角洲地區特有的祖姑崇拜傳統中得到強化，更在晚清民變中，成為凝聚增城周遭不同何姓宗族的關鍵力量，藉由共同祭祀何仙姑，結成同姓互助聯盟，一起挺過人心惶惶的混亂時局。
對於宗族以符合禮法的話語，合理化地方社會祭祀未嫁女性的傳統，廖小菁提出她的看法：
歷史的行動主體始終是人，雖然身處既定的社會結構、面對特定的歷史情境，人們還是可以做出選擇，而「信仰」一直提供人們行動的資源。

yep, always have the choice, always

無體温關係

陪伴是亙古不變的需求，只是，好好維持一段讓人舒服自在的親密關係，已是一件非常奢侈的想望。
現代人的生活如此壓縮，都會區裡的物理空間日益逼仄，我們花在工作的時間不斷延伸，還得自我調適無所不在的社群凝視，更別說近年外在世界的戰爭與瘟疫讓人不安，在克服上述所有的挑戰後，的確有為數不少的人難以有餘裕去經營一段親密關係，去消化關係裡各種的情緒勞動。
於是有不少人決定從挫敗的關係裡退場，或是避免磨合的煩擾，不再委屈自己。
2023年年度流行詞彙之一的“situationship（情境關係）”，指的是一種不明確的浪漫或性關係，此情境下人們僅是朋友，隨性且沒目的性地彼此陪走一程，不牽涉承諾，也不一定建立長期關係。

當我們大力談高未婚率、少子化問題，將之視為國安問題，動輒要別人「好好愛一場」、把成家當最重要目標、要生兒育女傳宗接代，其實忽略了人們最根源的需求其實是個人的身心安適。在這個高度耗神的時代裡，一味地鼓吹傳統或神聖的親密關係，可能是帶有負面作用的道德教條。
目前全台灣916萬家戶裡，已有322萬是單獨生活戶。這個時代有更多自我滿足的方式，除了結婚生子，懂得自處與自我照料、追尋屬於自己的快樂、建立多元的社會關係，也是值得尊重的人生選擇。

Boys struggle

We have spent the past half-decade wrestling with ideas of gender and privilege, attempting to challenge the old stereotypes and power structures. These conversations should have been an opportunity to throw out the old pressures and norms of manhood, and to help boys and men be more emotionally open and engaged. But in many ways this environment has apparently had the opposite effect — it has shut them down even further.
For many progressives, weary from a pileup of male misconduct, the refusal to engage with men’s feelings has now become almost a point of principle. For every right-wing tough guy urging his crying son to “man up,” there’s a voice from the left telling him that to express his concerns is to take airtime away from a woman or someone more marginalized. The two are not morally equivalent, but to boys, the impact can often feel similar. In many cases, the same people who are urging boys and men to become more emotionally expressive are also taking a moral stand against hearing how they actually feel. For many boys, it can seem as though their emotions get dismissed by both sides. This political isolation has combined with existing masculine norms to push a worrying number of boys into a kind of resentful, semi-politicized reclusion.

well，都係幾現實嘅問題，最重要都係真實嘅個體，而唔係度度都用到嘅廢話

Perhaps it’s not surprising that in the grip of the culture wars, caring about boys has become subtly coded as a right-wing cause, a dog whistle for a kind of bad-faith politicking. Men have had way more than their fair share of our concern already, the reasoning goes, and now it’s time for them to pipe down. But for boys, privilege and harm intertwine in complex ways — male socialization is a strangely destructive blend of indulgence and neglect. Under patriarchy, boys and men get everything, except the thing that’s most worth having: human connection.
Silencing or demonizing boys in the name of progressive ideals is only reinforcing this problem, pushing them further into isolation and defensiveness. The prescription for creating a generation of healthier, more socially and emotionally competent men is the same in the wider political discourse as it is in our own homes — to approach boys generously rather than punitively. We need to acknowledge boys’ feelings, to talk with our sons in the same way we do our daughters, to hear them and empathize rather than dismiss or minimize, and engage with them as fully emotional beings.
They are more than ready to talk. We just need to make sure we are listening.

需求係存在，人類本性如此。如果始終不被理會，上面就係一種解決方案——無體温關係。

NBA Foster

https://www.nytimes.com/athletic/5494360/2024/05/20/scott-foster-nba-referee

The gap between how Foster is viewed by some players and coaches, and how the league sees him, highlights an issue that has plagued the NBA for years and sharply escalated this season: the quality and consistency of the league’s officiating. The league this season levied $765,000 in fines to players and coaches for criticism and/or inappropriate conduct toward officials, up from the $385,000 fined last season. Postgame news conferences are often as notable for coaches’ denunciations of the referees as for commentary about the players.
Foster’s polarizing status derives from two of his definitive traits: His accuracy in calling a game and what is viewed as an unpleasant bedside manner. It has led to a complicated legacy, one that confuses and pains him.

其實個問題主要都係NBA，裁判係你請，判罰有問題點解唔可以公開批評同質疑呢？雖則從聯盟嘅角度嚟講，球員都係舞台上嘅演員，配合好打晒成場波就好，唔好計較咁多影響公信力同聲譽。但的確有人會着緊輸贏着緊公平囉，裁判咁縱法第日仲得了嘅？

The conflicts and confrontations have led to what Foster calls “the noise.” Criticism that he is arrogant. Complaints that he refuses to communicate with players and coaches. Insinuations that he cannot be trusted. And the insistence that he holds grudges.
He says he can handle “the noise,” in part because it’s part of the job and in part because he is held in the highest regard by his peers.

“Scott Foster is by far one of the top referees in the game, if not the top referee,” Crawford said. “As far as his personality and people not liking him? It’s because he takes no flack. If you come at him in an unsportsmanlike manner, you are going to pay for it.”

“I think Scott is an excellent official,” Billups said. “He’s not going to put up with any s—. He’s an old-school guy in the way that I like officials, like Joey Crawford, Steve Javie … they are not going to take nothing, whether you are the best player or a player up from the G League. It’s the same respect given. So I’ve always respected that about him.”
Billups said he thinks what hurts Foster is his body language and his natural affectation. Billups said Foster often looks aggravated or in a bad mood, and noted he rarely, if ever, smiles. It took years for him to conclude that it’s just Foster’s natural look.

以我睇比賽嘅差唔多十年嚟講，確實Foster絕對係專業、有經驗同埋尺度一致，但態度同肢體語言都的確唔幾令人喜歡

地方治理信用透支

冯川专门研究农村社会与基层治理，长期在基层进行调研，他在调查中发现，我国地方治理中的信任危机已经层层蔓延到村里，县政府失信于银行、企业和乡镇政府，乡镇政府再失信于乡镇干部、村干部和工程队，然后村干部失信于工程队、村民，村民也对基层干部失去信任。

中国法律规定，地方政府不能直接借债。各个地方政府绕开这一规定的方法，就是成立作为融资平台的国有企业，如城发集团、城投公司之类，以公司名义向银行贷款融资，弥补县城建设存在的巨大资金缺口。
县政府透支金融信用的表现之一，是“只还利息不还本”。在贵州省S县调研，当地负债的原因主要是搞了很多公共建设，比如大广场，而这种工程普遍没有收益。负债搞建设的动力何来？县长说，主要是来自政绩压力：其他县都这么搞，不搞就落后了。而且一届县长一届政绩，他们不考虑以后怎么样。虽然当地财政状况在黔东南地区算较好的，还没有拖欠公务员工资，但当地土地财政只有1亿，却已负债80亿，利息每年还需要偿还1-2亿。县长直言，负债是不可能还了，银行只能要利息，否则连利息都不给。
县政府透支金融信用的表现之二，是为争取银行放贷而采取手段“虚增”国有资产。银行放贷有一套严格的审核体系，对融资平台公司的资产总值、资金现金流、担保及抵押品折现值、公司信用等级、项目回报率等等都有明确要求，不能违规放贷。而政府成立的平台公司大多没有实际资产和现金流，增加融资平台公司资产的手段，通常是把政府投资建造的机井、渠道、门面房甚至公共厕所等统归为该公司的有效资产，但这些多是不良资产。

政績工程~土地財政都未夠畀利息，大癲

项目式供给是县域发展的普遍模式。项目竞争一方面需要地方治理主体进行资金配套，另一方面，许多项目的申请都要求手续“先建后补”、资金“以奖代补”，即先开工建设，再补办手续，建成验收合格后获得项目资金拨付。这种模式，使地方治理主体能够将一个大的项目拆分成很多小包，地方治理主体不仅有了选择由谁来承包工程的空间，还能赚得上级项目经费和实际建设经费的差价，这个差价也成为了地方治理主体自主配置的资源，为地方治理主体与工程队、企业之间的合谋提供了较大空间。为了推进项目的落地，地方治理主体有必要与包工头和企业家搞好关系，以个人名义担保包工头和企业家：先开工建设，等项目资金下来之后，将不光能收到工程款，还能从中获利，或者获得更多参与其他项目的机会。
比如在陕西调研了解到，村里的项目通过财务运作，不同项目之间经费可以挪用，其实里面空间很大。尤其是基础设施类项目，区里干部说，利润空间有时可达50%。那里的村干部用私人的名义与包工头签合同，因为包工头不愿意跟村里签。包工头知道给村里或者政府做工程，工程款要拖很久，最后找不到人。但是跟村干部签，就有具体的债权人。参与项目的包工头其实大多是与村干部交好的工程老板，村干部跟他们软磨硬泡，做工作，他们才同意接项目。所以这类村级债务实际是私人债务，但做的又都是村集体的事情。
然而，一旦本应按时下拨的项目经费迟迟难以兑现，或者项目经费难以覆盖工程款，地方治理主体就不得不再通过跟进项目来弥补前面的窟窿，逼得地方治理主体不断申请新的项目奖补。这种私人借债最终可能演化为新的县级债务或村级债务，不断透支着基层干部与工程队、企业之间的信用。

奇奇怪怪嘅方案，對於年年挖路整路有咗新嘅理解方式

据观察者网此前报道，曾有480多个人民政府被列为失信被执行人（包括村一级村民委员会），其中是地方政府的为100多个，不乏包括副省级城市、省会城市核心区等。有的地方政府上榜不止一次，郸城县宁平镇人民政府、郸城县李楼乡人民政府分别“上榜”3次。

若要从根源上解决地方治理中“无赖化”行为的蔓延问题，还需要跳出政府失信行为本身，着眼于失信行为背后的结构性问题。
第一，要为县域发展解套。首先是正确评估县城发展能力。在以县城为重要载体的城镇化推动过程中，县级财力薄弱的县城更容易产生高负债问题。作为人口净流出地区的县域经济体量有限，工业发展程度不高。此类县城不具备区位和资源优势，没有能力复制一线城市、省会城市和东部“百强县”的发展经验。其次是恰当定位县域发展目标。位于中西部欠发达地区的县域政府盲目招商引资、开展工业园区化和大城市化建设，政府投入无法带来相应产出，造成国家公共资源的大量浪费。此类县城的发展目标应顺应人口流动趋势、产业分布规律和社会民生需求，不应脱离实际、盲目贪大求全，诱发政府无序贷款融资。应在县级治理考核中弱化GDP比重，扩大生态保护、公共服务等综合指标比重。再次是恰当处理政治与行政的关系。缺口动员的逻辑起点是县域财政的结构性缺口，而缺口产生的主要原因之一是不切实际的政治目标。因此，县域政府要避免治理事务的泛政治化，上级政府只应负责把握大体方向，克制对行政过程和行政指标进行精细化安排的政治冲动，适度保留县乡政府主体性的运作空间。这样才能实现投融资市场化，从根源上解决县域财政缺口增量。
第二，要为村级组织解套。一方面，要弱化以亮点工作作为基层工作考核的主要依据之一。避免“看点”“拉练”“红黑榜”“观摩”等成为逼迫基层扩大乡村建设投入的常用手段。严禁以乡村振兴为名义，盲目上项目、铺摊子，通过借债而运动式打造“明星村”。要按照“一要吃饭，二要建设，量入为出”的原则，在安排村级基础建设和公益性设施建设时，坚持量财办事、量力而行。另一方面，要对成为村干部的村庄经济精英进行规范和引导。消除能人治村的灰色空间和超前发展偏好，加强村级民主建设，发挥村民代表作用，规范决策程序，坚持实行“一事一议”。进一步加强村级财务管理，全面锁定村级债权、债务和资产。
第三，要为基层干部解套。基于信任的需要而重构问责制度，使基层干部在信任的驱动下主动履行责任，是维护一线治理工作对基层干部产生内涵式激励效果的当务之急。我们需要重新思考信任的价值和公共利益的定位。只有重建组织信任，重塑公共建设的服务指向，才有可能使一线治理事务促成基层干部内涵式激励的达成，重构基层治理的信用生态。

啱㗎，呢個解決辦法合理且應該有效

taste

But the idea that specific tastes are confined to certain areas of the tongue is a myth that “persists in the collective consciousness despite decades of research debunking it,” according to a review published this month in The New England Journal of Medicine. Also wrong: the notion that taste is limited to the mouth.

Such taste receptors aren’t limited to the tongue; they are also found in the gastrointestinal tract, liver, pancreas, fat cells, brain, muscle cells, thyroid and lungs. We don’t generally think of these organs as tasting anything, but they use the receptors to pick up the presence of various molecules and metabolize them, said Diego Bohórquez, a self-described gut-brain neuroscientist at Duke University. For example, when the gut notices sugar in food, it tells the brain to alert other organs to get ready for digestion.
Dr. Breslin likens the system to an airport preparing for a plane landing.
“Think about if a plane landed at an airport terminal that wasn’t ready,” he said. No one would be prepared to guide the plane to the gate, clean it up or unload the luggage.
Taste, he said, gets things ready. It wakes up the stomach, stimulates salivation and sends a little insulin into the blood, which in turn transports sugars into the cells. Ivan Pavlov, a Russian physiologist who won a Nobel Prize for his studies on digestion in 1904, showed that lumps of meat placed directly into a hole in the dog’s stomach would not be digested unless he dusted the dog’s tongue with some dried meat powder to start things off.

評論留言区

← 夕陽，好暖！

聖誅 مقدس‎ →