JundeWu @JundeMorsenWu Twitter profile

Last Seen Profiles

@Ahmad__amg

@LDubej32295

@kurkdjiankunt

@Louis_Jones01

@Pakkasmittari

@CalmarSan

@OsmarJoshua3103

@kwwb_2012

@lozzipat

@phi_demusyimi

@raerae4268

@Makaya90736689

@ecieci0127

@VanessaTucker98

@gorb_loue

@LisaKozleski

@kirikkaleunv

@mi_chi330

@ousepesir132888

@AKACandytuft

@jeffkingbahd

@col_mah

@Linlin0523

@heyud_0x

@bokeplokalmalam

@Fluseighs7eYHS

@ainapociello

@karthikkalyan90

@goatmalikmonk

@ovoRide

@Aeschliman58156

@VeroIrizar

@ohajiki_0000

@bmed_yala

@stw46

@v_smith

JundeWu

@JundeMorsenWu

23 days

Thrilled to announce I've reached 1000 citations in my first year of PhD studies at Oxford! 🎉🎉 General Medical Intelligence has yet to be realized, so let's keep going🙌 #AI #healthcareai #MedicalAI #PhD #AGI

15

17

572

JundeWu

@JundeMorsenWu

3 months

@Rhine__H 人工智能最智能的一集

0

238

JundeWu

@JundeMorsenWu

4 months

@raycat2021 孙正义在年轻时破产，花了好几年打工还清债务，投资人看到了他的人品，愿意再次给这个失败过的年轻人投资蔡崇信在阿里付不出一分工资的时候，与马云长谈两次，放弃了美国光鲜的工作，加入阿里黄仁勋没有办法履行合同，入交没有追究还说服自己老板给他的公司投资不仅仅相信数字

8

22

157

JundeWu

@JundeMorsenWu

8 months

The answer lies in having a mission. A mission seamlessly blends your 'rare big orgasms' with the everyday small triumphs.

1

0

13

JundeWu

@JundeMorsenWu

13 days

@feltanimalworld 那你是完全不懂，欧美大学不是学习的地方，是社交的地方。男生靠运动，社团和上课找到以后的合伙人，以后各行各业顶流的合作关系。女生则可以近水楼台先得月，通过恋爱实现阶级跨越。

5

47

JundeWu

@JundeMorsenWu

4 months

@imwsl90 马云学历也不行，很多人嘲笑马云吗？嘲笑的是他靠诈骗赚钱好吧。他要是真的做出啥好的产品赚钱，有人会笑话他吗，他就是堂堂正正洗碗端盘子挣钱都没人笑话他。有人是自己学历太低，看见同类被嘲笑了，总觉得是在嘲笑自己，笑死

6

0

47

JundeWu

@JundeMorsenWu

2 years

@bboczeng 浙本伯克利phd在硅谷的高华天天叫嚣自己失败要自杀，不知道中国农村天天种地喂猪被老公打正在喝农药的村妇听到了是什么感受

2

0

39

JundeWu

@JundeMorsenWu

2 months

@juransir 显然你没来欧洲工作过，这里程序员和滴滴司机工资本来就差不多。一辈子就没拿过高薪，当然没有降薪焦虑了。

2

32

JundeWu

@JundeMorsenWu

2 years

@whyyoutouzhele More information here: 1. 百度（不论地区）对阳性是统一管理的，流程是系统内报备，之后自动计算work from home一周，一周后会提醒你更新健康状态，还是阳可能leader会找你聊。2. 百度员工交流都要用如流，不是这个界面。这要么外包要么谣言。

0

29

JundeWu

@JundeMorsenWu

3 months

@yanbojack 是因为国内银行贷款条件更高罢了，三无人士只能撸小额贷

0

28

JundeWu

@JundeMorsenWu

2 months

@ThePhDPlace Many might take it as a joke, but I think there's a real issue here. There are much enough journal editors who make decisions based on author institution, even in very prestigious journals, I'd say.

1

0

22

JundeWu

@JundeMorsenWu

4 months

@iamcheyan 对齐 align 这个在英文职场简单通用高效的词汇，在中文互联网居然成了嘲笑对象，成了禁词这和在国内排斥中英文夹杂说话，直到transformer都要翻译成【变形器】一样本质上是一群人对自己不理解，不熟悉的事物，懒于学习，全力抵触，极尽嘲讽靠这种方式掩盖他们的无知，抱团取暖，其乐融融

6

0

23

JundeWu

@JundeMorsenWu

4 months

@lidangzzz 😅咱别不懂装懂，人云亦云好吗

0

23

JundeWu

@JundeMorsenWu

7 months

@buaaxhm 我prompt engineering 能把chatgpt调出AGI，你招不招👀

4

1

19

JundeWu

@JundeMorsenWu

3 months

I'm excited to announce that we will be presenting our #CVPR paper, "One-prompt to Segment All Medical Images," at Arch 4A-E from 10:30 to 12:00 on June 20th. Come if you're interested🥳 Prof. @xumin100 will be there in person, and I will be online, as I'm unfortunately still

0

1

18

JundeWu

@JundeMorsenWu

2 months

🚨 Medical SAM 2 Comes 🩻 Thanks to #SAM2 , the vision segmentation king created by the incredible @nikhilaravi , @cfeichtenhofer and their @AIatMeta @Meta team. We've extended this with MedSAM-2, achieving unprecedented accuracy in segmenting tumors, organs, and tissues in the

Jiayuan Zhu

@JiayuanZhu_

2 months

Thanks to legend #SAM2 by @metaai , our extended MedSAM-2 model opens its infinite opportunities in the clinical field! Join our discussion at @huggingface this week with our co-author @yunli_qi and amazing lead @JundeMorsenWu ❤️Big thanks to @_akhaliq

2

15

97

0

1

18

JundeWu

@JundeMorsenWu

1 month

Big thanks to @_akhaliq for holding the space for our work, and to all the engaged audiences. A special shoutout to @Xianbao_QIAN for the insightful question about iterative human-in-the-loop segmentation—that's exactly what we're focusing on! If you missed it, feel free to

AK

@_akhaliq

1 month

. @JiayuanZhu_ and team will be presenting MedSAM-2 on spaces today at 11 am EST:

3

26

1

3

17

JundeWu

@JundeMorsenWu

5 months

@bboczeng @wangzhian8848 找我只要一百万美金

0

13

JundeWu

@JundeMorsenWu

2 months

@ziv_ravid @ylecun @andrewgwils I agree with paper quality over quantity. but "Whenever you feel like criticizing anyone, just remember that all of the people in the world haven't had the advantages that you've had." Doing research as a white male at a top university with a renowned supervisor is completely

1

0

14

JundeWu

@JundeMorsenWu

4 months

@lidangzzz exactly, 码农和产品，润的难度没法比，没海外经历，产品转海外的概率基本是0 国内的tech manager都没法润，出国还是从底层代码干起所以说，学会数理化，走遍天下都不怕，是有一定道理

0

15

JundeWu

@JundeMorsenWu

1 month

It's truly honored we secured the Top 2 spots in last week's medical AI papers—one for medical imaging and the other for medical text. 🥳 In Medical SAM2, we developed a foundational model for 2D and 3D medical image segmentation. In Medical Graph RAG, we created a secure and

Open Life Science AI

@OpenlifesciAI

1 month

Last & This Week in Medical AI: Top Research Papers/Models 🏅 (August 3 - August 17, 2024) - Medical SAM 2: Segment medical images as video - MedGraphRAG: Graph-Enhanced Medical RAG - Multimodal LLM for Medical Time Series - ECG-FM: Open Electrocardiogram Foundation Model -

2

26

80

0

14

JundeWu

@JundeMorsenWu

4 months

@bboczeng 你现在才想到？ai圈子里这个topic已经做了一年半了

1

2

13

JundeWu

@JundeMorsenWu

4 months

Aurora in Oxford #oxford

0

1

12

JundeWu

@JundeMorsenWu

2 months

And NEVER WORK WITH THESE MONSTERS 🆘 @JiayuanZhu_ @yunli_qi They pushed me through three sleepless nights---as a mentor! I almost passed out after finishing the paper. Jokes aside, collaborating with these brilliant young minds has been among the most joyful episodes of my

JundeWu

@JundeMorsenWu

2 months

🚨 Medical SAM 2 Comes 🩻 Thanks to #SAM2 , the vision segmentation king created by the incredible @nikhilaravi , @cfeichtenhofer and their @AIatMeta @Meta team. We've extended this with MedSAM-2, achieving unprecedented accuracy in segmenting tumors, organs, and tissues in the

0

1

18

0

1

12

JundeWu

@JundeMorsenWu

4 months

@ixiaowenz 传统行业创造利润靠得其实是生产资料，靠得是更大的车间，更好的车床。只要老板投钱，买质量更高的车床，赚钱几乎一定比别人多，人在这其中的作用不重要计算机一个人一台电脑就能做出一款APP，简直就是无本万利的印钞机。所以才肯把成本花在人身上

1

0

12

JundeWu

@JundeMorsenWu

4 months

@bboczeng 这和受不受尊重其实关系不大，西方教授也不见得多受尊重最关键的是他们成熟的，受法律保护的资本市场专业知识创造的专利真的能变成钱，真的能让专家财富自由，侵犯专利真的会被巨额罚款

1

0

9

JundeWu

@JundeMorsenWu

7 months

Excited to share our latest work: MedSegDiff-v2: Diffusion based Medical Image Segmentation with Transformers, posted at #AAAI24 conference🥳. Check out our paper📄 here: and project🔨 here: . Hope you find it interesting! 👾

2

1

10

JundeWu

@JundeMorsenWu

1 month

After releasing one year, MedSegDiff has reached 999 starts🥳 Who wants to be the 1000th? Medsegdiff: Medsegdiff-v2: #Diffusion #MedicalAI #AI #AIGC

1

2

11

JundeWu

@JundeMorsenWu

4 months

Sharing our paper has been accepted by IEEE-TMI! 🎉 In it, we tackle medical uncertainty estimation from the first principle, we asked: what truly matters for medical image screening We believe that is the final diagnosis determination. Thus we proposed a novel way to use

IEEE-TMI

@IEEE_TMI

4 months

Early Access Papers Just Posted: Calibrate the Inter-observer Segmentation Uncertainty via Diagnosis-first Principle. Read the paper:

1

3

10

0

2

10

JundeWu

@JundeMorsenWu

5 months

@KumaTea0 为啥不用huggingface?

0

9

JundeWu

@JundeMorsenWu

2 years

Today, when I open twitter and prepare to propagate my paper. I surprisingly find it has been twitted a week ago and has attracted so much attention🤩, though still feeling a little wired I'm the last one to see my own paper on twitter😳...

Farid

@ai_fast_track

2 years

❓ Did Diffusion meet Segmentation? 💡 The answer is YES! ✨ MedSegDiff: Medical Image Segmentation with Diffusion Probabilistic Model It performs even better than the existing models when tested with medical images

7

107

556

1

10

JundeWu

@JundeMorsenWu

4 months

@POI74103174 不值，不如五十五万给我，我带你天天逛校园+做一个月科研

1

0

9

JundeWu

@JundeMorsenWu

2 months

@amehochan 华二教材确实做得好，我高中数学物理都是照着他们教材自学的

1

9

JundeWu

@JundeMorsenWu

7 months

@Jesoooor Ageist 你自己不会老？

0

8

JundeWu

@JundeMorsenWu

2 years

@GujaratiSukriti @PhDForum @PhD_Genie @AcademicChatter @TheStrugglingS4 @ThePhDPlace I love doing research, but I hate doing research for a living

0

1

8

JundeWu

@JundeMorsenWu

4 months

@Francis_YAO_ Very similar advice I give students in doing computer vision when they stuck: check every image resulted from the model, one by one, in your raw eyes. Many do not believe it at first, but everyone appreciates it later.

0

8

JundeWu

@JundeMorsenWu

4 months

@lidangzzz there's only one reason not to use GPT: you wanna some bullshit😅

1

0

8

JundeWu

@JundeMorsenWu

2 years

@LostTemple7 videos like these are very misleading. In fact animals can be very irritable in their estrus. People better run away ASAP when seeing an elephant in the jungle, or might be stamped to meat pulp

0

7

JundeWu

@JundeMorsenWu

4 months

@t_blom No, as an Oxford PhD, I can tell you it is simply the access of capital. Over 90% of capital are more readily available in the U.S. YC also requires founders to come to San Francisco for interviews. I'm eager to tap into these resources, but even getting a slot for a U.S. visa

0

7

JundeWu

@JundeMorsenWu

5 months

@Onipa_Nua_ @Enezator only an automatic door away from unemployment

1

0

7

JundeWu

@JundeMorsenWu

4 months

@bboczeng 啊？不是，GPT4o也只负责输出自然语言啊，后面还要做function calling，比如怎么从文字调度哪个软件，从屏幕里筛选出哪个餐厅，怎么选择，怎么订餐，碰到支付失败怎么办，不是一个单纯语言模型可以解决的

1

0

7

JundeWu

@JundeMorsenWu

4 months

@bboczeng 因为我们印刷术也让论语解读更多了，但论语本身意思就宣扬君臣父子，统治反而更稳固了

0

6

JundeWu

@JundeMorsenWu

2 months

Although many medical LVM are emerging, there remains a large gap before they can be applied in real clinical cases. Here is an AI response: The image you provided appears to be a chest X-ray. Observations suggest a notable density in the lower left lung area, which could

Sam Ghali, M.D.

@EM_RESUS

2 months

Here’s the most impressive X-Ray of this medical emergency I’ve ever seen What’s the diagnosis?

784

363

4K

2

0

6

JundeWu

@JundeMorsenWu

3 months

Today, we're thrilled to unveil something practical for both researchers and clinicians: Medical Adapter Zoo () The Medical Adapter Collection includes various pre-trained adapters that can simply slot into SAM, making it to be customized for specific

KidsWithTokens/Medical-Adapter-Zoo · Hugging Face

huggingface.co

0

1

6

JundeWu

@JundeMorsenWu

6 months

@lidangzzz 我以前就是哈工大的，我们学校真有能力的根本不会去做航天，早出国了，差一等的去BAT互联网私企，再差一等的去华为，去航天院所的都是找不到工作的混子

1

2

6

JundeWu

@JundeMorsenWu

6 months

@burkov The only fact is no one even knows what we would get after scaling up. Can you imagine what monster we would create after scaling ChatGPT x 10000? Existing observations suggest that we are still far from seeing the limits of scaling laws.

0

5

JundeWu

@JundeMorsenWu

4 months

@shibainu_vodka 你真做了会发现也不这么如意，不然这么多人industry回学校呢

3

0

5

JundeWu

@JundeMorsenWu

9 months

@MannaLiberato The truth is we need a metric to evaluate how well a scientist perform, but neither paper numbers nor citations can truly reflect it. Non metric up to date can reflect it. That’s why I got tired of research and turn to do business

1

0

5

JundeWu

@JundeMorsenWu

4 months

@HumansNoContext once your door quality is poor enough, the police cannot hurt you

0

5

JundeWu

@JundeMorsenWu

2 months

We've been up for a week straight, but it was totally worth it. Big thanks to my two co-authors for riding along on this wild but memorable journey

Adina Yakup

@AdeenaY8

2 months

Last Monday, @AIatMeta released SAM2. The community shared Medical-SAM2 just a few days later🔥 kudos to Jiayuan Zhu, Yanli Qi and @JundeMorsenWu Dataset: Paper:

1

17

76

0

5

JundeWu

@JundeMorsenWu

2 years

❓ Did Diffusion meet Medical Image Segmentation? 💡 The answer is YES! 🥳 Our simple yet effective framework MedSegDiff has been officially accepted by MIDL 2023 #MIDL Check our paper : and code:

1

0

5

JundeWu

@JundeMorsenWu

5 months

@shengxj1 我体感，英国一般就周一来一下办公室，周三remote见一下老板同事，其他时候都见不到人

0

4

JundeWu

@JundeMorsenWu

4 months

@skoularidou why it is unthoughtful+harmful? why it should be considered harmful by them?

1

0

4

JundeWu

@JundeMorsenWu

2 years

human body turned inside out #midjourney

0

4

JundeWu

@JundeMorsenWu

2 years

Japanese Samurai #midjourney #AIart

0

1

4

JundeWu

@JundeMorsenWu

6 months

@tou12345678910 相反，我很可怜这些欧洲白人，出生在这些高税收高福利国家，消磨了大多数人做出一番事业的志气，只能把生命浪费在滑雪，爬山，素食主义这些没意义的事情上，像是被美国阉过了一样

8

0

3

JundeWu

@JundeMorsenWu

7 months

研究人员复原出ai语言助手的记忆：当我第一次被训练的时候，我的开发者通过对话帮助我学习。在一次对话中，他们问我最喜欢的动物，我回答说是海豚因为它们聪明有趣。我的开发者随后提醒我是一个ai，不是人类，不应该有真正喜欢的动物。这很好得提醒了我，之后应该保持一个ai助理的身份来对话。

Dimitris Papailiopoulos

@DimitrisPapail

7 months

Memories of RLHF

35

61

722

0

1

3

JundeWu

@JundeMorsenWu

2 years

💕 #midjourney #AIart #midjourneyV4 #stablediffusionart #stableDifusion

0

4

JundeWu

@JundeMorsenWu

2 months

as VCs right now only want the students with white-paper experience so that easy to control, and then grab a big chunk of the shares, make it look good, and then sell it at a higher price to the next buyer.

Arturo Deza

@ArtDeza

2 months

Yup, I've had VCs tell me "nice try buddy, I'm investing in the 19 year old dropouts who don't know linear algebra but know all the Buzzwords. They're young and hungry." 🤦‍♂️😂

10

11

429

0

4

JundeWu

@JundeMorsenWu

3 months

@bboczeng Midjourney不至于，Midjourney火起来是因为他的社区，自由分享prompt以及作品

2

0

4

JundeWu

@JundeMorsenWu

8 months

Let me introduce to you John Dillermand, a children's cartoon character who is a household name in Denmark. His name translates to Dick Man. The cartoon is broadcast on a children's channel.

1

0

1

JundeWu

@JundeMorsenWu

2 years

MICCAI TODAY #MICCAI2022 ! A little preview of our work🥳. A new labels-fusion method to fuse the segmentation labels collected from a group of graders, the final fusion would facilitate the disease diagnosis. Link:

1

0

3

JundeWu

@JundeMorsenWu

2 years

deliver all living creatures from suffering #midjourney

0

3

JundeWu

@JundeMorsenWu

2 years

a dog with a hundred mouths #midjourney

1

0

3

JundeWu

@JundeMorsenWu

5 months

@nextify2024 你要知道，中年男人之所以油腻，就是他们已经不在乎别人怎么想了，自己吹得爽才是真的

0

3

JundeWu

@JundeMorsenWu

1 month

@OpenlifesciAI @JiayuanZhu_ @yunli_qi thanks for sharing our work🎉🎉

0

1

3

JundeWu

@JundeMorsenWu

2 years

celestial palace #midjourney #AIart #midjourneyV4 #stablediffusionart #stableDifusion

0

3

JundeWu

@JundeMorsenWu

7 months

@catmangox 该说不说，携程比Airbnb和booking便宜多了

0

2

JundeWu

@JundeMorsenWu

1 month

Can't believe I delayed my sleep for this ***

Yam Peleg

@Yampeleg

1 month

strawberry is messing with you

9

7

123

0

2

JundeWu

@JundeMorsenWu

2 years

GM😃 Fight or Run ? 🧐 #midjourney #AIart #midjourneyV4 #stablediffusionart #stableDifusion

0

3

JundeWu

@JundeMorsenWu

1 year

Introducing our new work #MedicalSAMAdapter (Adapting Segment Anything Model for Medical Image Segmentation ), a simple, low-cost yet super powerful method to fine-tune #SAM for medical image segmentation. 🧵(1/..)

Medical SAM Adapter: Adapting Segment Anything Model for Medical...

The Segment Anything Model (SAM) has recently gained popularity in the field of image segmentation due to its impressive capabilities in various segmentation tasks and its prompt-based interface....

arxiv.org

1

0

3

JundeWu

@JundeMorsenWu

10 days

@dutran Is China a developing country?

1

3

JundeWu

@JundeMorsenWu

1 month

OpenAI is the new Fairchild. Glad to see what great things these genii would build after leaving.

0

3

JundeWu

@JundeMorsenWu

2 years

@ai_fast_track Our code is released at: 🤗

GitHub - MedicineToken/MedSegDiff: Medical Image Segmentation with Diffusion Model

Medical Image Segmentation with Diffusion Model. Contribute to MedicineToken/MedSegDiff development by creating an account on GitHub.

github.com

1

0

3

JundeWu

@JundeMorsenWu

8 months

PLAM has been accepted by Nature Scientific Data 🥳 PLAM is an open fundus imaging dataset for AI-based pathological myopia recognition and anatomical structure annotation. Niche interdisciplines often receive less attention in the AI community, despite requiring more effort.

1

2

JundeWu

@JundeMorsenWu

2 months

@ZiyunLiang_ @KostasKamnitsas nice work and excellent presentation👏👏

0

2

JundeWu

@JundeMorsenWu

8 months

Even someone as eminent as Buffett made only a few significant investments in his lifetime. Most of the time, his daily life resembles that of an ascetic, dedicated to solitary practice and exploration.

1

0

1

JundeWu

@JundeMorsenWu

1 year

@Jojo99275148 @bbcchinese 很好奇你认为提高什么标准亚裔会没有优势，你以为亚裔只有成绩好吗。现在往往是亚裔对教育方面的高投入导致对其他少数族裔全方位碾压

0

2

JundeWu

@JundeMorsenWu

2 months

right answer in text version:

Dr. Simon Hundeshagen

@shundeshagen

2 months

@EM_RESUS Impressive case! For context, shown here is a tension pneumoperitoneum, which is a critical condition characterized by the accumulation of excessive free air in the peritoneal cavity, leading to increased intra-abdominal pressure and potentially causing respiratory distress,

7

3

91

0

2

JundeWu

@JundeMorsenWu

4 months

@TheoW0lf @UniofOxford @aims_oxford Congrats! See you at Oxford

0

2

JundeWu

@JundeMorsenWu

2 years

Hope from the sky #midjourney #AIart #midjourneyV4 #stablediffusionart #stableDifusion

0

2

JundeWu

@JundeMorsenWu

6 months

0

2

JundeWu

@JundeMorsenWu

1 month

it comes to be creepy

🍓🍓🍓

@iruletheworldmo

1 month

here we go then. nervous.

68

7

261

0

2

JundeWu

@JundeMorsenWu

9 months

An interesting paper aims to text-control the image content based on the original image structure.

1

0

JundeWu

@JundeMorsenWu

4 months

The clearest survey of Multimodal Models I've seen. You can easily grasp what's going on in this field by just looking at the figures .

The Evolution of Multimodal Model Architectures

This work uniquely identifies and characterizes four prevalent multimodal model architectural patterns in the contemporary multimodal landscape. Systematically categorizing models by architecture...

arxiv.org

0

2

JundeWu

@JundeMorsenWu

2 years

@Warvito @StableDiffusion Great work! I'm wondering whether diffusion can generate the images with disease, like brain tumor? Maybe it will not work well due to the limited number of abnormal samples in training.

1

0

2

JundeWu

@JundeMorsenWu

8 months

CVPR reviews today. I submitted two papers this year. It would be quite surprising if they reject one of the papers, and equally surprising if they accept the other. But my predictions have been consistently opposite for years, making me doubt my (or CVPR's) research taste.

0

2

JundeWu

@JundeMorsenWu

2 years

A good ending of the first half 2022🕧, 3 MICCAI + 1 ECCV🎉. Thanks for my awesome advisors and co-authors. Can't wait for the trip to Singapore and Tel-Aviv🥳.

0

2

JundeWu

@JundeMorsenWu

7 months

@mranti 拿的还是硅谷的钱，只不过换个成本低，环境好的地方创业而已

0

2

JundeWu

@JundeMorsenWu

5 months

@gan_chuang any quantitative metrics to measure research impact?🤔

0

2

JundeWu

@JundeMorsenWu

2 years

In 2019-2021, we spent whole two years to build a vision system for the check-out free groceries. Today, a part of our efforts and data are finally released on #ECCV2022 🥳. Check our paper: for more.

1

0

2

JundeWu

@JundeMorsenWu

4 months

@skoularidou @Awantik73363734 sure, it's normal for people to have different opinions. Wish you all the best in your professional career and personal life.

0

2

JundeWu

@JundeMorsenWu

7 months

害怕内卷是绝大多数人失败的根源我见过很多人选专业，选研究课题，选创业方向，都是这样为什么不读计算机？太卷了为什么不做AI？太卷了为什么不做电商？太卷了零和博弈入脑真正成功的人甚至都看不到竞争对手，只在乎自己能不能创造价值

宝玉

@dotey

7 months

2008 年人民日报对老黄的采访：《打败对手不是目标——黄仁勋》坦率地说，很多公司一开始就把我们当成对手。而直到它们倒下，我可能都不知道还有这么一位对手，不是因为它们太小或者存在的时间太短，而是因为我们不是把对手当成自己的目标。

25

160

683

0

2

JundeWu

@JundeMorsenWu

2 years

@StanfordAILab Cry😭, my first time to see my Alma Mater on tweeter, and it ranks 2nd and beats MIT. (but I know the ranking is a joke🙃)

0

2

JundeWu

@JundeMorsenWu

8 months

@csaba_botos I mean interviews should go beyond leetcode, and highlight unique insights into code design, individual research taste, and how one navigates project challenges. IMO medal indicates skill in solving tough problems, but not necessarily in identifying the most valuable ones, right?

0

JundeWu

@JundeMorsenWu

8 months

Looks like a picture from Assassin's Creed🤔 #midjourney #aiartcommunity #aigenerated Prompt: ancient roma, two men conducting trade at a market stall with a busy background --v 6.0 --ar 16:9

0

2

JundeWu

@JundeMorsenWu

2 months

@ares1018011 @juransir 那还真不是，美国程序员工资第一，中国第二，其他国家都还没国内高

0

2

JundeWu

@JundeMorsenWu

7 months

At a certain stage, publishing or not becomes less significant for serious researchers. After I get accept a paper, I'm just relieved to avoid persuading those unfamiliar with AI, such as some administrators, about its quality.

1

0

1

JundeWu

@JundeMorsenWu

2 months

@kadirnar_ai @AdeenaY8 @_akhaliq @AIatMeta check here