Kaicheng Yang, PhD | Assistant professor @BingCompSci | AI & Society, + | views mine
Binghamton, NY
Joined May 2019
- Tweets3.5K
- Following1.5K
- Followers1.6K
- Likes8.4K
Pinned Tweet
This semester, I have been teaching an AI and Society class. As the name suggests, the goal is to explore the transformative impact of AI, particularly LLMs and AI agents, on society. This is a seminar class; see the reading list link below.
1/3
Kevin Yang retweeted
Open sourcing my regression lectures 🍻
AI is changing how we teach. Here’s one way I’m adapting: I’ve open-sourced my regression course, POLI 450A, built on Jens Hainmueller’s original class.
The course is suitable for first-year political science PhDs, as well as advanced undergraduate students in economics, political science, and other social sciences.
The materials include slides, R code, data, worksheets, and a short video for every lecture.
github.com/xuyiqing/regressi…
Kevin Yang retweeted
1/4 I've built a small tool, CiteVerify, to help researchers and journal editors check citation integrity, especially in the social sciences. I've found it useful in my own editorial work and hope others will too!
github.com/cuihuash/citeveri…
Kevin Yang retweeted
New Nature Health perspective 🧪
The invisible risks of AI-generated health information
rdcu.be/VuS2jwQt6B3a
We argue that these risks are serious, yet hidden from the institutions responsible for protecting public health
w/ @mdeverna2 @yang3kc @YyjHarry
New Perspective in Nature Health: "The invisible risks of AI-generated health information."
Many users (including me) use AI for health information or advice. We argue that when that guidance causes harm, no one outside the platform can usually see a record of it.
AI health information reaches people in two settings. Some people see it without asking: AI-generated posts and chatbot replies appear in their social media feeds. Others look for it: search engines now put AI summaries above the links, and chatbots replace the links altogether.
Other health guidance leaves a record when it causes harm. A clinician's advice produces a chart entry. A harmful drug generates an adverse-event report. A misleading ad stays publicly visible. However, AI health guidance rarely leaves any of these records. Users have no channel to report harm, and independent researchers have no data to study it. We recommend making AI health information observable.
With @mdeverna2 , @YyjHarry, @OSoMe_IU .
Journal page: doi.org/10.1038/s44360-026-0…
Free link: rdcu.be/VuS2jwQt6B3a
Kevin Yang retweeted
Another angle to this is that dogs effectively stumbled onto what is for them the equivalent of artificial superintelligence 20,000 years ago. The Singularity already happened for dogs. They live in a world they cannot understand, surrounded by miraculous wonders they take entirely for granted, their every need seen to by entities whose thoughts they cannot begin to fathom.
The result is that there are about a billion dogs on the planet now. Most of them are pretty happy. Almost none of them need to work. It worked out pretty well for them.
Looks like this gives actual API credits 👀
We’re rolling out monthly Claude Platform API credits for Max and Team plans:
Max 5x: $100
Max 20x: $200
Team: up to $500, pooled
Works on any model, including Haiku 5.5, in your code or third-party harnesses.
How it works and full terms: support.claude.com/en/articl…
Kevin Yang retweeted
Come work with us @JHUCompSci! We're hiring tenure-track faculty in all areas, with a particular focus on
- AI systems
- AI safety and security
- AI reasoning and alignment
- AI ethics, policy, and governance
apply.interfolio.com/194231
(DMs are open if you have any questions)
Kevin Yang retweeted
You hear it here first! My department at @ISTatPENNSTATE is hiring multiple TT faculty in AI (very broadly defined!) this round.
Come work with us!
Job Posting: psu.wd1.myworkdayjobs.com/PS… – at Berkeley, CA
Kevin Yang retweeted
🧪 AI agents are easily radicalized, particularly when messages resonate with their persona beliefs, raising concerns about the vulnerability of people who use them
Preprint: doi.org/10.48550/arXiv.2609.…
w/ @ozgurcanseckin @shalmolighosh94 @KristinaLerman & al
"arXiv now limits submitters to up to two submissions per calendar month, with a limit of three total active submissions at any given time."
I actually didn't know arXiv had a submission limit before, but it seems they are restricting it further.
arXiv has updated our policy on rate limiting for all submitters.
This update was made to fairly distribute moderator time & support the arXiv community of staff, volunteers, readers & authors.
Please read our announcement to learn more: blog.arxiv.org/2026/10/01/up…
ALT arXiv logo in white over a whiteboard with math symbols.
Kevin Yang retweeted
Interested in the social impact of AI? Consider applying for this new post-doc at Duke's Society-Centered AI Initiative: academicjobsonline.org/ajo/j…
I'm adding this to my slides to tell the students not to blindly trust whatever your chatbots throw out 🫠
BREAKING: The US "almost started a war" against China, nearly launching an operation against a Chinese ship in the Middle East this spring, in the midst of the Iran war, after an AI-generated intelligence report falsely identified the vessel as carrying nuclear weapons components, with armed US personnel preparing to board and military aircraft already in the air before officials found the report had been produced with a chatbot, per CNN.
A special operations command analyst had used a chatbot to fuse open-source material with classified signals intelligence, then used AI again to package the result into a standard intelligence report and circulated it across the US military.
Switched to Tinycast lately as well. $10 for BYOK is one reason. I think the team is trying to turn Raycast into an all-in-one hub with AI and everything. I get it, and there's nothing wrong with it. But all I need is a quick launcher.
I'm uninstalling Raycast ❌
I've been a heavy user of Raycast for the past 4-5 years, every single Mac has it, even my iPhone. First I used it as an Alfred replacement, later I started using all the AI chats in it - amazing!
But now since I pay $200 for Claude and $20 for Codex, I don't want to pay another $10 for Raycast just to use my Claude inside.
I'm switching to Tinycast from @abue_ammar, it's native, it's fast, all my notes are .md files on the disk and mainly - it actually does all what I needed from Raycast + I use my Claude and Codex in there.
Great job Kuro! I'll soon do some PRs to contribute to the project as well.
Whoever thinks AGI is here should give OpenAI or Anthropic's customer service a try 😅.
Starting Aug 28, I couldn't authenticate to our lab's OpenAI org that we've used since 2021. 3 PhD student admins all saw the same issue, so probably not a browser thing. Can't see usage/billing or manage payment since then.
OpenAI's "support" is a series of bots requesting I run more and more diagnostic tests. "Jeffrey" got stumped so "James" had to step in.
Whatever the issue is, seems like it's tough enough to be a benchmark task!
Kevin Yang retweeted
Got access to GPT-6 Astra. Want to see some pelicans? Yeah you want to see some pelicans... here's a grid comparing Astra to GPT-5.6 Sol, Terra, and Luna static.simonwillison.net/sta…
ALT Grid comparing 12 illustrations of pelicans riding bicycles. Columns: gpt-6-astra, gpt-5.6-sol, gpt-5.6-terra, gpt-5.6-luna. Rows: MAX, XHIGH, HIGH reasoning effort. All are recognizable, cheerful cartoons with orange bills and sunny pastel backgrounds. Astra is consistently polished, with elegant bird shapes, fine wheel spokes and coherent bicycles. Sol is appealing but varies more in proportions, with a particularly elongated neck at MAX. Terra has expressive birds but more awkward anatomy: its XHIGH pelican faces left on a left-facing bicycle, and HIGH has bulky, arm-like wings. Luna is simpler, with stiff poses and some awkward wing and foot placement. Higher effort does not produce an obvious, consistent improvement across all four models. Costs in cents, in column order: MAX: 63.21, 48.55, 32.09, 7.83. XHIGH: 33.85, 24.11, 14.67, 4.24. HIGH: 18.37, 10.38, 3.74, 2.46. (Alt text by GPT-6 Astra)