Justin D. Weisz
Human-Centered AI Researcher · AI Startup Advisor · ACM Senior Member
Hi, I’m Justin! I study how to design AI systems that amplify and augment human capabilities. One lesson runs through my work: AI doesn’t have to be perfect to be useful, but you can’t just hand people AI and expect them to become more productive. AI has to fit into people’s workflows, minimize harm, and adapt to the people using it. Those three needs shape my research:
- Useful AI: How do we build AI that fits into people’s workflows and makes them more productive, in areas such as software engineering, design, and data science?
- Safe AI: How do we help the people who build and use AI understand its risks and choose the right mitigations?
- Personal AI: How can AI model its users’ goals, knowledge, and intentions, and make those models visible so that people and AI develop a mutual understanding?
I’m a mixed-methods researcher and combine qualitative studies, prototyping, controlled experiments, crowdsourcing, and speculative design to envision the future and turn research insights into shipped products.
After my Ph.D. in Computer Science at Carnegie Mellon University, I spent over fifteen years at IBM Research, where I co-led our global research strategy on Human-Centered Trustworthy AI and managed the Human-AI Collaboration team. My contributions include:
- developing & evaluating conversational programming assistants, including the Programmer’s Assistant and the Agentic Code Explorer, which informed the design of IBM’s watsonx Code Assistant
- training 3,000+ IBM design practitioners in designing generative AI applications via design principles and strategies and new design thinking activities and workshops
- building research communities at the intersection of HCI and AI through workshop series on human-AI co-creation (HAI-GEN), human-centered explainable AI (HCXAI), and theory of mind in human-AI interaction (ToMinHAI), and the Human-Centered AI publication on Medium
- adding new risks from generative AI to the IBM AI Risk Atlas
- teaching hundreds of thousands of students and makers about AI and quantum computing through TJBot and Entanglion
I hold 35+ US patents and was named an IBM Master Inventor.
Today, I bring my experience in human-centered AI research, product strategy, intellectual property, and research leadership as an advisor to pre-seed and seed-stage startups.