🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

I started teaching this semester and am taking every instance of being misrecognized as a student as a personal victory. Long may it continue!

As I hear about candidates on the academic job market getting calls for on-site interviews, I’m reminded of a head of search committee who asked if I’d gained weight since my headshot was taken—right before my first ever job talk. Thankfully, I didn’t end up working there. But

📢 Got early-stage research or exploratory ideas in CUIs? Submit a Short Paper or Work-in-Progress to #CUI2025 and share your insights with the community. ⏳ Submission deadline: 10 April 2025 (Less than 2 MONTHS away!) 🔗 Details: cui.acm.org/2025/submit/sh… #CfP #CUI #HCI

Did your Late Breaking Work on Conversational AI miss the cut for #CHI2025? Don’t worry—revise and submit it to #CUI2025! We’d love to see your work at ACM CUI.

Have you ever felt “left in the dark” while waiting outside the ICU? 📣If so, we built CareMap for you! An AI-powered tool to help families: 🧠 Understand medical updates ❓ Know what to ask ✅ Feel more in control 🔗 arxiv.org/abs/2502.05115 arxiv.org/abs/2502.05115

One of my most meaningful papers began with a participant laughing at Alexa and saying, “It’s like a silly child.” That metaphor led to years of research, multiple rejections, and a long relationship with uncertainty. I wrote about that journey here: smitdesai.net/chailab/2025/4…

A paper written by #iSchoolUI alum Smit Desai (MSIM '18, PhD '24) & Professor twidale has ranked in the top 10% of most-viewed papers published by "Proceedings of the Association for Information Science in Technology" in 2023. Congratulations! ▶️ bit.ly/4lPSQN4

I’ll be at #CHI2025 in Yokohama! I’m looking for PhD/Master’s students + collaborators interested in building conversational interfaces that support, not mimic, humans. Come say hi or DM me—happy to chat over coffee. More: smitdesai.net/chai-lab-index

🤖 LLMs can now roleplay as Daenerys, Obama, or your friend (to greater or lesser extent). Unlike traditional chatbots, these personas are dynamic & unpredictable—harder to design ethically, difficult to govern, and challenging to evaluate. To learn more, join our #CUI2025

Every researcher knows this feeling: staring at a screen, wondering what to work on next. Good news? Ideas can be generated systematically, not just discovered accidentally. Here's a framework I have developed for generating (good) research ideas: smitdesai.net/reflections/20…