
postdoc @kaistpr phd @UofMaryland | prev @MSFTResearch @MITIBMLab @fujitsulabs @Dataminr
College Park, Maryland
Joined May 2020
- Tweets186
- Following1.5K
- Followers1.8K
- Likes443
AI’s rat race: serving the paying 0.2% of the world and becoming one of the 5,000 people at frontier labs. The real blue ocean is serving the other 99.8%. The models are already good enough. This is an HCI opportunity.
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.
So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
1
6
887
Still prompting your AI for the "next best research idea"? Or hoping it will single-handedly crack humanity's hardest problems—but watch it fall short?
Imagine a memory palace: a 3D space where researchers and AI jointly weigh trade-offs and surface the most promising ideas.
Introducing ResearchCube: Multi-Dimensional Trade-off Exploration for Research Ideation → arxiv.org/pdf/2604.11538
1
7
857
April 15th is the day — thrilled to share that three students I've had the privilege of mentoring have made their PhD decisions:
@ZoeWang0213 → CS PhD @TAMU 🎉
@Carol_Zhang1027 → CS PhD @FloridaState 🎉
@georg11a1 → DS PhD @UVA 🎉
Watching them grow as researchers — from refining research ideas to navigating the application process — has been one of the most rewarding parts of my own PhD journey. Mentoring is a two-way street, and I've learned so much from each of them along the way.
1
785
🚨 When people are worried about language models’ linguistic intelligence surpassing humans, we should instead harness other human strengths such as spatial reasoning.
💡Thrilled to share our research concept accepted by UIST'25: Direct Intent Manipulation (DIM).
Zijian "Jason" Ding retweeted
🚨 Do stronger reasoning models guarantee better real-world coordination and performance?
Not always.
🔔 Sharing our new paper: From MAS to MARS: Coordination Failures and Reasoning Trade-offs in Hierarchical Multi-Agent Robotic Systems within a Healthcare Scenario
🎉 Officially ABD! My dissertation proposal "From Automation to Augmentation: Scaffolding and Structuring Human-AI Collaboration through Intent-oriented Interaction Paradigms" is approved by my committee. Time to hit the job market 🚀
As agents increasingly operate UIs, users can express intents instead of stepwise GUI manipulation. We've been exploring how to combine the strengths of intent-based interaction and direct manipulation and proposing Direct Intent Manipulation (DIM), accepted as a UIST'25 poster!
5
935
Not sure which career path to take in the age of AI automation—PhD, dev, ML, or startup? Introduce CareerPooler 🎱: turn career planning into a pool game—pocket milestones, skills & surprises to simulate multiple career paths! 🚀Try it out—join our study: forms.gle/8WqDx7ZbFsyDdU5U6
Zijian "Jason" Ding retweeted
🎓💰🔬 Want to learn machine learning, win a cash prize (USD 50K in total!!), and help drive real progress in discovering new polymer materials?
All available at our NeurIPS 2025 Open Polymer Challenge: open-polymer-challenge.githu… 🚀
Join now (Kaggle): kaggle.com/competitions/neur… 📈
Seeing other researchers adopt, apply, and cite a concept I contributed to is a truly rewarding moment: dl.acm.org/doi/full/10.1145/…
1
9
1,636
Zijian "Jason" Ding retweeted
🧪 Want an AI-generated paper draft in just 1 minute? Or dreaming of building auto-research apps but frustrated with setups?
Meet tiny-scientist, a minimal package to start AI-powered research:
👉 pip install tiny-scientist
🔗 github.com/ulab-uiuc/tiny-sc…
#AIAgent #pythonpackages
"History doesn't repeat itself, but it often rhymes."
We introduce the Analogical Timeline for the Hype Cycle as an initial approach to tackling a question that captivates everyone yet remains elusive: What will the future be like?
1
9
1,320
Forecasting tech advancements has always been challenging. Yet, historical parallels can offer insights into the patterns of innovation. We present the concept of an analogical timeline—a method of using LLM and one tech trajectory to predict the future evolution of another.
1
603
Try out the interactive demo at anatime.netlify.app and check out the open-source repository at github.com/JsnDg/Analogical-… :)
448
Excited to share that the full paper version of my intern work on GenAI hypothesis exploration @MITIBMLab has been accepted by @acm_cc, with an acceptance rate of 21.8%. This year was especially competitive—I deeply felt it as a PC member.
2
689
Next week at @ACMIUI, we’ll present my internship work at @MITIBMLab: "Structuring GenAI-Assisted Hypothesis Exploration with an Interactive Shared Representation." We applied Intent-AI Interaction to data analysis, allowing analysts to collaborate with AI at a conceptual level.
1
19
1,411













