Shu Yang 杨澍

INTP | CS phd @KAUST_News ｜Msc in CL @UM_1981

[Shu Yang 杨澍 @shuyhere]

just relax and enjoy research

[Harley Finkelstein @harleyf]

People always ask me what it's like working with @tobi.

Here's @sama's answer: he writes the software himself, he sends OpenAI more detailed product feedback than anyone else on earth, and he stays six to eight months ahead of every other CEO.

Nothing in this clip surprised me. This is just what every day at @Shopify looks like.

[Shu Yang 杨澍 @shuyhere]

vibe coding无论是出品还是体验都犹如过山车，从最近的体验来看真的可以叫slop generator了 https://t.co/xZUnnrn4ir

[Shu Yang 杨澍 @shuyhere]

@utk7arsh Thank you!

[Shu Yang 杨澍 @shuyhere]

Four papers were accepted to EMNLP! But... another one with a good score was rejected because the meta-review response seemed to be completely for another paper (not in my context window). I still have no idea how that happened. Maybe I’ve become an octopus and did this in dream

[Shu Yang 杨澍 @shuyhere]

I am still waiting to review how fancy it is.

[Tibo @thsottiaux]

I was gifted a very fancy new reset button today

[Shu Yang 杨澍 @shuyhere]

true

[Nathan Lambert @natolambert]

I tell a ton of people that getting a PhD is not a good idea for your career, but it’ll likely be far higher in personal growth and fulfillment which is maybe worth way more anyways. I need to think about how to price that in.

[Shu Yang 杨澍 @shuyhere]

@jian_w3ng 我一直以为这叫胖虎

[Shu Yang 杨澍 @shuyhere]

@jian_w3ng 所以出门不要看手机

[Alexander Panfilov @kotekjedi_ml]

We can finally talk about it:

We found a way to extract hidden reasoning of frontier models using a vulnerability in the APIs of every frontier AI company.

We verified that our reasoning token count matches billed API thinking tokens 1:1 for most of the prompts we queried.

[Nathan Lambert @natolambert]

I'm about a year out from some of the worst burnout I've built up in my time building Olmo. I feel like just in the last few weeks I've turned a corner to being more chill again (huge yay). It's wild to me just how long recovering from burnout can take.

For many people when they ask me how to break into AI I tell them the actions are simple, but it takes longer than you think. Same goes for burnout.

Though I think burnout is harder, as breaking into AI is a much more "forward looking" thing. Burnout is really pernicious to nail down and address. I think it often takes a moderate change in your career and habits. It's hard to take a slight step back as a successful person and fix your burnout. It'll really take a while.

In my case the final straw maybe my book being done and some more clarity on what I'm doing next. On top of it surely taking months longer because of how sad the changes at Ai2 were for me personally.

I really wish all of my friends in AI who are feeling this even stronger than I was can catch a break. I've by a mix of my nature and my habits as an athlete been pretty in tune with stopping myself before overworking. Imo as AI researchers we should take care of our brains like professional athletes take care of their brain.

Touching grass is worth it, and I'm very glad to find myself being slightly bored again. It's nudged me to reading more, of all things!

Take care of yourself.

Ps I feel like there's a recurring behavior of people somewhat like myself where they write about burnout or try to help people when they're actually burnt out themselves. Surely this is part of why I felt like I could write about it so well.

[Shu Yang 杨澍 @shuyhere]

System prompts are the “operating system” behind today’s AI assistants, yet they’ve been difficult to study systematically. Here is the largest public collection of real-world system prompts to support research on AI agents, prompting, safety, and alignment.

[SystemPromptIndex @aisystemprompt]

https://t.co/vfc7H6ZZdQ

[David K 🎹 @DavidKPiano]

https://t.co/ziU0b3pBks

[Daniel Scalena @daniel_sc4]

for those worried that AI is writing all the reviews: I bring good news

[Shu Yang 杨澍 @shuyhere]

@jian_w3ng 为什么会发消息给你

[EMNLP 2026 @emnlpmeeting]

We are aware that it is no longer possible to post Official Comments on ARR May 2026 submissions, even though the Author Reponse period runs until July 14. This is being investigated and it will be fixed. Authors do not need to submit Author-Editor Confidential Comments.

[Shu Yang 杨澍 @shuyhere]

@jian_w3ng 同虚无中

[Yeuoly @Yeuoly1]

https://t.co/sEY8STy8lh

[Mira Murati @miramurati]

Today we're sharing our work on interaction models. A new class of model trained from scratch to handle real-time interaction natively, instead of gluing it onto a turn-based one.
https://t.co/MoS5s4cm60

[Shu Yang 杨澍 @shuyhere]

@jian_w3ng 哈哈哈哈哈哈哈哈哈哈哈 claude是真的把开发者当250