Tsinghua University
Public university in Beijing, Chinawww.tsinghua.edu.cn
117
MENTIONS
85
EPISODES
72
PODCASTS
Search complete. 117 mentions across 85 episodes found for "Tsinghua University".
Nov 4, 2026
China’s people-centered approach to internet development
D
37:05Ding HengHOST
Under a 2016 agreement, the United States committed to provide Israel with $3.8 billion in military aid annually through the year 2028.
D
37:18Ding HengHOST
So for more, my colleague Lian Qingxuan earlier spoke with Dr. Wen Jin, a research fellow at the Center for International Security and Strategy with Tsinghua University.
L
37:29Lian QinxuanCORRESPONDENT
Why has Washington chosen this particular moment to end this $4 billion aid, and what message does this decision send? I
W
37:39Wen JinGUEST
think actually this question is based on one big issue that US foreign policy towards Israel, I think may have some change.
8 MINS LATER
W
45:27Wen JinGUEST
So I think this could affect America's leverage over Israel.
W
45:31Wen JinGUEST
So in total, I think this decision may have great significance as a signal for changing Americans' priorities as an immediate change in the military balance of the Middle East.
D
45:44Ding HengHOST
That was Wenjing from the Center for International Security and Strategy, West Tsinghua University, talking with my colleague Lian Qingxuan.
D
45:54Ding HengHOST
You're listening to World Today.
Roasting China's AI Industry (中文版)
I
42:22Irene ZhangHOST
中国的产业为什么不造词呢?
X
42:24XianyuGUEST
我觉得它是——造词是上层建筑,它是一个你需要真的自信才能够做的事情。就中国的 AI 创业者们,他相对来讲这一个相——非常小的圈子来讲,它大体上是一个平替的逻辑。 当然我非常希望它能够有真正的创新,嗯......当然它也有很多真正的创新。对,但它大体上来讲,这些AI应用,这些具体的AI项目,它依然是一个美国火的一个什么项目,硅谷火的一个什么项目,然后我们几个人,不管回国以后,我们几个人,不管......呃......我们不——几个人,不管,不管几个大——几个,几个清华大学的学生也好,我们来做一个类似的項目,来蹭这个主题的一个事情。对,它-它-它-它应该和来——钱来源上的构成也有关系,比如说美元基金,他们很多钱......呃......来——你-你要,你要向LP解释,你要向你的出资人解释,你为什么要投这个项目。那对-对于,对于这些老登,这些老企业家来说,对于这些中东、欧洲的钱来说,你-你给他们解释为什么这个,这个项目是一个我们中国的一个年轻人自己想出来的一个项目,和这个项目是硅谷已经非常火的一个赛道,然后中国的这哥嘛,他可能——他的成本更便宜,然后......嗯......他能够执行得更好,他的效率更高。这两个故事之间,这些老企业家,这些不管是中东,还是欧洲,还是美国的老登们,他们大概率会 buy in 第二种啊,它和钱的来源上构成有关系。所以这回到——这其实是一个话语权定义,定义的问题。所以就-就假如说[笑]这-这——所-所以这个事情其实是中国的人民币基金是更会——更-更独立思考的。虽-虽-虽-虽-虽然现在也还在一个非常早期的阶段,但他们会,会更按照,会-会-会更按照符合中国情况的事情来投资,而不是跟随硅谷的方式来投资。当-当然我讨论的是一个非常理想的情况,现在都还在一个很早期的阶段,你没法这么严格来对比。对,但总的来说,我觉得它还是......呃......AI,AI行业的人不够自——中国AI行业的人不够自信造成的。当然这个自信——不自信,你可以给他有非常多理由,因为什么什么,所以他不自信,但最重要是不自信,但最重要的是自信。对,你首先要自信,你才能够去提出 bench,benchmark,去建立标准。对,但-但-但这是回到一个非常宏大的角度来讲了,就-就假如说我们今天举个例子,呃......一个中国的AI产品,AI应用,它向投资人展示它数据的时候,告诉他我有100万美国用户,呃......然后另一个告诉他我有100万印度用户,那我只有1万美国用户,投资人会投哪一个?这答案是不用质疑的,全世界投资人一定都会投有100万美国用户那一个人。对,它就是一个综合的美国,美-美-美国能力的体现。对,所以当-当今天这个市场上有一个 benchmark 的时候,啊......这个 benchmark 有-有人告诉你说,这个 benchmark 是一个中国的公司做的,然后另一个 benchmark 是 OpenAI 做的,嗯......那市场上绝大部分公司会选择 OpenAI 做的。对,当-当然事实上可能 OpenAI......呃......当然事实上 OpenAI 它的模型能力更领先,它做的 benchmark 可能它的代表性更好,但大体上来讲,大伙依然会选——大-大伙依然会不假思索地对美国做的这些......呃......话语权类的东西,指标类的东西,定义类的东西会更有信任。对,它-它是一个事实。
I
45:50Irene ZhangHOST
但是我觉得......嗯......就是我明白你前一部分,就是......嗯......毕竟因为发展中国家的这个消费力没有那么高,所以基金——所以投资人还是会更相信在美国能跑出的这个消费模式。啊......但是在技术这个级——层面上讲,我——至少我自己的观察是......啊......西方的研究者对于中国的这个......啊......中国这些 lab 的基础,基础实力是有-有信-有信——有-有了解的。啊......尤其当然是 DeepSeek 以后啊,然后......啊......就是如果——比如说这个 Kimi 出——Kimi K3 出的时候,看到很多人都去认真拜读了它的 paper。啊......就是我觉得中国公司如果想要至少在技术造词这方面做影响的话,绝对是有市场的。
X
46:41XianyuGUEST
它就是一个心-心态的问题,它是一个自信的问题。大伙,大伙下意识里面没有想到干这样事情。
Roasting China's AI Industry (中文版)
I
42:22IreneHOST
中国的产业为什么不造词呢?
X
42:24XianyuGUEST
我觉得它是——造词是上层建筑,它是一个你需要真的自信才能够做的事情。就中国的 AI 创业者们,他相对来讲,这一个相——非常小的圈子来讲,它大体上是一个贫瘠的逻辑。 当然我非常希望它能够有真正的创新。嗯......当然它也有很多真正的创新。对,但它大体上来讲,这些AI 应用,这些具体的AI 项目,它依然是一个美国火的一个什么项目,硅谷火的一个什么项目,然后我们几个人,不管回国以后,我们几个人,不管......呃......我们不——几个人,不管,不管几个大——几个,几个清华大学的学生也好,我们来做一个类似的项来蹭这个主题的一个事情。对,它-它-它-它应该和来——钱来源上的构成也有关系,比如说美元基金,他们很多钱......呃......那你-你要,你要向LP 解释,你要向你的出资人解释,你为什么要投这个项目。那对-对于,对于这些老登,这些老企业家来说,对于这些中东、欧洲钱来说,你-你给他们解释为什么这个,这个项目是一个我们中国的一个年轻人自己想出来的一个项目,和这个项目是硅谷已经非常火的一个赛道,然后中国的这哥嘛,他可能——他的成本更便宜,然后......嗯......他能够执行得更好,他的效率更高。这两个故事之间,这些老企业家,这些不管是中东,还是欧洲,还是美国的老登们,他们大概率会 buy in 第2种啊,它和钱的来源上构成有关系。所以这回到——这其实是一个话语权定义,定义的问题。所以就-就假如说[笑]这-这——所-所以这个事情其实是中国的人民币基金是更会——更-更独立思考的。虽-虽-虽-虽-虽然现在也还在一个非常早期的阶段,但他们会,会更按照,会-会-会更按照符合中国情况的事情来投资,而不是跟随硅谷的方式来投资。当-当然我讨论的是一个非常理想的情况,现在都还在一个很早期的阶段,你没法这么严格来对比。对,但总的来说,我觉得它还是......呃......AI,AI 行业的人不够自——中国 AI 行业的人不够自信造成的。当然这个自信——不自信,你可以给他有非常多理由,因为什么什么,所以他不自信,但最重要是不自信,但最重要的是自信。对,你首先要自信,你才能够去提出 bench,benchmark,去建立标准。对,但-但-但这是回到一个非常宏大的角度来讲了,就,就假如说我们今天举个例子,呃......一个中国的AI 产品,AI 应用,它向投资人展示它数据的时候,告诉他我有100万美国用户,呃......然后另一个告诉他我有100万印度用户,那我只有1万美国用户,投资人会投哪一个?这答案是不用质疑的,全世界投资人一定都会投有100万美国用户那一个人。对,它就是一个综合的美国,美-美-美国能力的体现。对,所以当-当今天这个市场上有一个 benchmark 的时候,啊......这个 benchmark 有-有人告诉你说,这个 benchmark 是一个中国的公司做的,然后另一个 benchmark 是 OpenAI 做的。嗯......那市场上绝大部分公司会选择 OpenAI 做的。对,当-当然事实上可能 OpenAI......呃......当然事实上 OpenAI 它的模型能力更领先,它做的 benchmark 可能它的代表性更好,但大体上来讲,大伙依然会选——大-大伙依然会不假思索地对美国做的这些......呃......话语权类的东西,指标类的东西,定义类的东西会更有信任。对,它-它是一个事实。
I
45:50IreneHOST
但是我觉得......嗯......就是我明白你前一部分,就是......嗯......毕竟因为发展中国家的这个消费力没有那么高,所以基金——所以投资人还是会更相信在美国能跑出的这个消费模式。啊......但是在技术这个级——层面上讲,我——至少我自己的观察是......嗯......西方的研究者对于中国的这个......啊......中国这些 lab 的基础,基础实力是有-有信-有-有了解的。啊......尤其当然是 DeepSeek 以后啊,然后......嗯......就是如果——比如说这个 Kimi 出——Kimi K3 出的时候,看到很多人都去认真拜读了它的 paper。嗯......就是我觉得中国公司如果想要至少在技术造词这方面做影响的话,绝对是有市场的。
X
46:41XianyuGUEST
它就是一个心-心态的问题,它是一个自信的问题。大伙,大伙下意识里面没有想到干这样事情。
China’s Massive Debt Scare Threatens The Entire System | Stock Market | Xi’s Crackdown
T
10:56TonyHOST
Patronage networks, nepotism, and the buying and selling of positions remain explicit targets, while personnel departments face greater accountability for appointments.
T
11:06TonyHOST
Tsinghua University researcher Xie Maosong said exclusively rapid rotations had encouraged some officials to pursue debt-financed prestige projects rather than complete difficult structural reforms, leaving their successors to manage the consequences.
T
11:21TonyHOST
The reforms therefore seek to address a practical governance problem.
T
11:25TonyHOST
Rewarding visible short-term achievements can undermine sustained administrative work needed to resolve local debt, property distress, and weak confidence, all deep systemic crises facing the country.
Field Notes: AI, Robotics, Commercial Space and the Case for Niches - Baiguan Radio #45
R
15:28Robert WuHOST
The best universities and the best research institutes are all in Beijing.
R
15:33Robert WuHOST
Yeah, people talk about Zhejiang University in Hangzhou, but that's just one Zhejiang University, right? In Beijing, you have Tsinghua, you have Peking U.
R
15:40Robert WuHOST
If you think about, say, Kimmy and Zhipu, and the reason they were founded in Beijing is because all these founders used to be trained in the AI research institutes in Beijing.
R
15:51Robert WuHOST
If you think about the robotics...
Cultural Capital: How do you hone cultural fluency as a competitive advantage?
J
6:51Judy LiuGUEST
So I thought that I understood the country very well.
J
6:54Judy LiuGUEST
However, you know, when I studied in Tsinghua University for my EMBA, my cohort actually came from across China, different cities, industries, and backgrounds.
J
7:08Judy LiuGUEST
This actually reminds me, especially as an international business, often we take Beijing, Shanghai, Guangzhou and Shenzhen for the whole China, but actually they only represent around 5% of the country's population.
J
7:24Judy LiuGUEST
So China is not one customer market, it's many different cultures and commercial markets within one country.
Stanford Seminar - Technology Management During a Time of U.S.- China Friction
R
66:23Richard DasherHOST
Before Xi Jinping, people would criticize the government pretty openly.
R
66:28Richard DasherHOST
I was at a big dinner hosted by a successful kind of marketing guru at Tsinghua University.
R
66:36Richard DasherHOST
and said, ah, the party's horrible.
R
66:40Richard DasherHOST
And, you know, he was really seriously angry.
ai morning #104 — claude mythos 5.5 found hidden in anthropic's own website
M
5:27MarcusHOST
And quantized builds from Unsloth and the MS community made it runnable on way more hardware, Apple Silicon included.
M
5:35MarcusHOST
There's a Tsinghua and Microsoft paper too.
M
5:37MarcusHOST
DeepSeq's harness used 87.93% fewer tokens on 98 human eval tasks, and success actually rose from 71.43% to 85.71% by skipping redundant calls on steps that were already predictable.
M
5:56MarcusHOST
And GitHub's top trending repo right now, claw-mem, does basically the same thing, compresses and persists your agent's session context instead of resending the whole thing every run.
Thinking Capital | Cultural Capital: How do you hone cultural fluency as a competitive advantage? (Sponsored Content)
J
6:51Judy LiuGUEST
So I thought that I understood the country very well.
J
6:54Judy LiuGUEST
However, you know, when I studied in Tsinghua University for my EMBA, my cohort actually came from across China, different cities, industries, and backgrounds.
J
7:08Judy LiuGUEST
This actually reminds me, especially as an international business, often we take Beijing, Shanghai, Guangzhou and Shenzhen for the whole China, but actually they only represent around 5% of the country's population.
J
7:24Judy LiuGUEST
So China is not one customer market, it's many different cultures and commercial markets within one country.
The END Is Coming, The Asteroid APOPHIS Signals TROUBLE
T
12:34Tim PoolHOST
In 2009, Russia's space agency said it would study a deflection mission.
T
12:40Tim PoolHOST
In 2011, a team at China's Tsinghua University proposed slamming a solar sail-powered impactor into it.
T
12:48Tim PoolHOST
None of those flew.
T
12:49Tim PoolHOST
What's flying now is a lot more serious.
75 more episodes mention Tsinghua University.
Create an account to see the whole feed, search across every transcript, and follow the entities you care about.