cclsnh

And yet it moves.

[Rob Henderson @robkhenderson]

"men are treated more harshly than women in the criminal justice system: They receive longer sentences for the same crime...Interestingly, male judges are more likely than female ones to go easy on female defendants." https://t.co/CqS3r2MbRl

[Phil Magness @PhilWMagness]

Here's something many people don't know about me -

Before I publicly dissected the long list of problems in the 1619 Project, I contacted the New York Times through their official channels to request a series of corrections to unambiguous factual errors in its content. The editor - Jake Silverstein - brushed me off and refused any correction - a pattern he also exhibited toward other critics from across the spectrum.

Before I publicly broke the story about Kevin Kruse's plagiarism in Reason, I contacted Princeton's academic integrity officer and alerted him to the problems I had found, giving them a chance to respond and address it internally. They ignored my email and later claimed to have lost my email after I went public.

Before I published my findings on Quinn Slobodian's habitual manipulation of source materials to alter its plain meaning through misquotation, I submitted an article to Contemporary European History (the journal where the worst examples appeared), highlighting the problems with the passages and asking for a correction through their official process. They desk-rejected it, brushed me off, and falsely claimed that Slobodian's piece had been thoroughly vetted in peer review. In fact, one of their own referees had flagged the same problems over a year earlier and recommended rejection of the article.

Before I published an expose on Nancy MacLean & Sandy Darity's similar manipulation of W.H. Hutt quotations in their article for History of Economics Review, I (along with 2 coauthors) submitted a response comment to this journal asking for a correction through its official processes. The editor gave us a complete runaround where he imposed an arbitrary length limit requiring us to cut the content, sent the trimmed version to a referee, then rejected the piece because the referee said we didn't sufficiently address the very same things we were forced by the editor to cut. When I then asked the editor to issue a simple corrigendum to the most egregious misquotation (one that transformed Hutt's explicit attack on the racism of white Afrikaners into a defense of Apartheid), he refused and tried to pass it off as a difference of "interpretation."

Before I published an expose of a leading covid masking model in the Wall Street Journal, I sent a comment to the medical journal that published it alerting them to a math error that changed their entire set of results. The journal acknowledged the error was real but refused to publish my piece on the grounds that the "next release" of the model would be updated to reflect it - even as politicians up to and including Joe Biden were trumpeting the erroneous results all over the news.

[Lola Viscera @lola_viscera]

one of the most inspiring interviews i’ve seen to date

[cclsnh @cclsnh]

@gabriel_zucman Honestly, I think the absolute amount of tax contribution is more important than the tax rate. 20 percent of a billion is much important in terms improving wellbeing of real people than 30 percent of 50,000.

[李老师不是你老师 @whyyoutouzhele]

5月19日，据《华尔街日报》报道：一项发表在《Nature》上的最新研究发现，中国官方媒体（新华社、人民日报等）的海量内容已被大量爬取并嵌入全球主流AI模型的训练数据中。结果导致ChatGPT、Claude、Gemini等AI在用中文回答涉华敏感问题时，立场明显更偏向北京官方叙事。

上周，在特朗普-习近平北京峰会成为全球头条的同时，《自然》（Nature）杂志发表了一项重磅研究：7位美国顶尖大学研究者首次用同行评审证据证明，中国官方媒体（新华社、人民日报、学习强国等）的海量内容，已大量进入全球主流AI模型的训练数据集。

研究团队通过分析开源数据集CulturaX（含1.89亿中文文档）发现：
中国官方媒体内容数量是中文维基百科的41倍；
涉及“党代会”“中央委员会”等敏感政治词汇时，官方内容占比高达1/4。

结果：ChatGPT、Claude、Gemini等AI在用中文回答涉华敏感问题时，立场明显更亲北京，能自然接上习近平标志性政治论述并“正面解读”；而英文回答则相对中立。中国自研的DeepSeek模型无论中英文都高度一致亲官方。
研究还发现类似现象存在于俄罗斯、朝鲜等国。更关键的是：这不需要黑客攻击——专制国家宣传完全免费、开放在网页上，而民主国家的独立媒体多采用付费墙，导致AI爬虫天然偏好“免费”内容。
研究覆盖37个国家得出结论：一个国家的新闻自由度越低，其语言的AI输出就越偏向政权叙事。由于大语言模型不透明，用户根本无法知道自己看到的“事实”来自哪里。
加州大学圣迭戈分校中国数据实验室联合主任Molly Roberts指出：“威权政府现在可以通过AI，跨越国界塑造全球信息消费。”

[cclsnh @cclsnh]

@shellenberger @mattwridley Apparently those booing kids don’t boo AI when they use it happily for their homeworks. This is exactly why they’re called the coddled generation because they insist that a coin should only have one side

[cclsnh @cclsnh]

@Overlap_Tech

[cclsnh @cclsnh]

@Dan_Jeffries1 Apparently those booing kids don’t boo AI when they use it happily for their homeworks. This is exactly why they’re called the coddled generation because they insist that a coin should only have one side

[Spotlight on China @spotlightoncn]

Last week, amid the global headlines surrounding the high-stakes summit between President Trump and Xi Jinping in Beijing, a quieter but profoundly consequential piece of research dropped in Nature. A team of seven researchers from major American universities published the first peer-reviewed evidence that China’s state-controlled media has successfully worked its way into the training data of AI chatbots that the world increasingly relies on.

The study demonstrates that scripted articles, official slogans, and party-line phrasings churned out daily by entities like the Xinhua News Agency and the Communist Party's study apps are now demonstrably embedded inside ChatGPT and other top models. A quick test of one of Xi Jinping's signature political doctrines shows that global chatbots seamlessly finish the phrases and offer to explain their political significance, reflecting an underlying saturation of state doctrine.

By combing through CulturaX, a massive open-source data set containing 189 million Chinese-language documents widely used to train AI models, the researchers found that state-media content is 41 times more abundant in the corpus than Chinese-language Wikipedia. While the overall overlap sits at a modest 1.64%, that share climbs to roughly one in four documents when filtering for politically sensitive terms like the Party Congress or the Central Committee.

“What is new here is now they are shaping the systems people increasingly ask to summarize, explain, and interpret the world for them,” explained Molly Roberts, a researcher on the team and co-director of the China Data Lab at the University of California San Diego. She noted that through this mechanism, authoritarian governments can now shape information consumption not just domestically, but across international borders.

When the team posed politically sensitive questions regarding Chinese governance to major commercial chatbots, the Chinese-language answers came back overwhelmingly more favorable to Beijing than their English counterparts. While Western models like GPT, Claude, Gemini, and Grok showed a distinct divergence between languages, China’s own DeepSeek model remained uniformly pro-Beijing across both English and Chinese, reflecting strict state regulatory control over its data.

The phenomenon extends beyond China, revealing a similar pattern for queries regarding Russia and North Korea. Crucially, this ideological slant did not require covert cyber operations; the propaganda is simply available on the open web in plain, unpaywalled HTML, making it free and easy for any AI lab's web crawler to scoop up and ingest.

This reality highlights an uncomfortable systemic asymmetry in global media ecosystems. While independent, high-quality journalism in democracies increasingly relies on paywalls to sustain its operations, state-run propaganda from authoritarian regimes remains entirely free, creating a massive economic imbalance in the textual material available for machine learning.

A broader audit spanning 37 nations confirmed that this trend is a global issue: the lower a country's press freedom, the more regime-friendly the local-language output of the AI becomes. Because large language models do not transparently cite their sources, users are left completely unable to decipher the true origins of the geopolitical narratives presented to them.

The Beijing summit generated a brief wave of international headlines, but this structural penetration of artificial intelligence demands a policy conversation that lasts years. While the scientific community has officially proven that authoritarian states are shaping global AI outputs, the question of how democracies will counter this invisible influence remains entirely unanswered.
https://t.co/kDVrilSS0P

[cclsnh @cclsnh]

@wangzhian8848 所以是不是可以用同样的逻辑证明你家暴了前妻

[The New York Times @nytimes]

The average pay of housekeepers in New York City hotels will increase to more than $100,000 a year as part of a contract settlement between an industry trade group and a powerful union. https://t.co/Jb95HIXQcL

[The Rabbit Hole @TheRabbitHole]

If diversity is a strength, then it’s a strength of the Right. In fact, the Right has a stronger claim than the Left does.

[Anatoli Kopadze @AnatoliKopadze]

the engineer who built Claude Code just dropped a 28-minute video on how to write prompts that actually work

I've seen $300 courses that don't cover what he shows in the first 10 minutes

CLAUDE.md files, memory shortcuts, parallel sessions, prompting patterns

all in one video and completely free

works whether you're a developer, a beginner, or someone who's been using Claude for months

based on this, I put together 18 things you can copy and use in Claude today

full guide in the article below

[cclsnh @cclsnh]

@SwipeWright Campuses are hopeless now, but the campuses are dying anyway in an era of AI

[Rob Henderson @robkhenderson]

YouGov: 25% of “very liberal” Americans say violence can be justified to achieve political goals compared with 3% of “very conservative” Americans.

[Anatoli Kopadze @AnatoliKopadze]

this is f*cking gold

the Claude setup most people will never find on their own

if I had this a year ago, I would've worked 5x faster

in the right hands, this changes everything:

[Steve Stewart-Williams @SteveStuWill]

"Relative to the general population, liberals are overrepresented among faculty by 186% - vastly more than men and considerably more than Asians. Meanwhile, conservatives are underrepresented by 70% - more than Hispanics, Black people, and Native Americans."

https://t.co/migUGJcTLQ

[cclsnh @cclsnh]

@MarlaFights @IanCopeland5 @SwipeWright I guess for him the fact that some people have six fingers proves finger number is a continuum…

[cclsnh @cclsnh]

@ItsEthanRay This woman is really good at making word salads that is based on zero reasoning but full of provocations