40% with just 1 try per task: SWE-agent-LM-32B is the new #1 open source model on SWE-bench Verified. We built it by synthesizing a ton of agentic training data from 100+ Python repos. Today we’re open-sourcing the toolkit that made it happen: SWE-smith.

We're hiring at the Stanford Digital Economy Lab! Seeking a Postdoctoral Research Fellow to join our project on developing user-centered AI agents in consumer settings. Learn more: digitaleconomy.stanford.edu/about/post-doc…

Come and work with us on consumer-authorized AI agents!! Happy to chat about this opportunity!

AI Shopping/Sales Agents sound very cool! But what if both the buyer and seller use AI agents? Our recent study found that stronger agents can exploit weaker ones to get a better deal, and delegating negotiation to AI agents might lead to economic losses. arxiv.org/abs/2506.00073

🚨 70 million US workers are about to face their biggest workplace transmission due to AI agents. But nobody asks them what they want. While AI races to automate everything, we took a different approach: auditing what workers want vs. what AI can do across the US workforce.🧵

Many companies are trying to automate human workflow with AI Agents. But what kinds of tasks do human workers really want to automate, and can AI agents really automate everything? Check out our paper 👇

AI can get very persuasive these days, but what happens in a future where AI agents negotiate with each other? New research from Jiaxin Pei et al showed that when your agent sucks, you could lose money, and this could lead to a new kind of digital gap technologyreview.com/2025/06/17/111…

AI companions aren’t science fiction anymore 🤖💬❤️ Thousands are turning to AI chatbots for emotional connection – finding comfort, sharing secrets, and even falling in love. But as AI companionship grows, the line between real and artificial relationships blurs. 📰 “Can A.I.

What happens when both consumers and merchants have AI agents acting on their behalf? Scholars, including Stanford Stanford Digital Economy Lab faculty lead Prof. Alex Pentland and Postdoctoral Fellow Jiaxin Pei, explore the art of the automated negotiation. hai.stanford.edu/news/the-art-o…

I love the meme lol

Are AI scientists already better than human researchers? We recruited 43 PhD students to spend 3 months executing research ideas proposed by an LLM agent vs human experts. Main finding: LLM ideas result in worse projects than human ideas.

What do workers want from AI? Researchers from Stanford HAI and Stanford Digital Economy Lab undertook a comprehensive study involving U.S. workers and AI experts. Here's what they found: stanford.io/3IsmHfg

After sharing our preprint on the Future of Work with AI Agents, we received strong interest in the WORKBank database. Today, we’re excited to release it publicly—along with a visualization tool to explore occupational and sector-level insights🧵

I will be at #acl2025 to present "Beyond Demographics: Fine-tuning Large Language Models to Predict Individuals’ Subjective Text Perceptions" ✨ Heartfelt thank you to my collaborators Jiaxin Pei Paul Röttger Philipp Cimiano David Jurgens is now on BlueSky only Dirk Hovy more below

I am looking for a postdoc to join my group at Harvard Graduate School of Education (HGSE) this fall. We'll work together to conduct empirical research on the impacts of AI on child development and advance child-centered AI design. More info and apply: apply.interfolio.com/170921 About the group: ying-xu.com/research/

Why don't you just say "this message is for Chinese researchers"? Besides, I am also amazed by your superpower to recognize the ethnicity of anonymous reviewers. Otherwise, how could one just assume a negative review is from a WeChat user?

When more and more people start to use AI agents to interact with the web, website maintainers need better ways to define the rules to guide the behaviors of these agents. We propose agent-permissions.json to better support coordination between agents and web pages!