How We View AI

Artificial intelligence is reshaping the education sector. Universities are scrambling to craft policies on how AI should be used in classrooms and research. However, the prevailing narrative remains rooted in fear—particularly around plagiarism—rather than vision. This reactive posture not only stifles innovation but misses a critical opportunity: to explore how AI can meaningfully democratize the educational space for creativity, collaboration, and discovery.

Neither AI denialism
nor AI hopium

A growing concern is that AI in education is being driven largely by external technology firms eager to capture the lucrative education market. The trend may produce two unintended consequences. First, it may deepen the ideological and technical divide between early AI enthusiasts and AI skeptics, creating unnecessary friction that hinders the development of safe and inclusive AI. Second, big tech firms lack lived experience in the realities of higher education—the diverse needs of students, instructors, researchers, and administrators. As a result, they often promote AI tools that act as shortcuts to efficiency but risk short-circuiting the deeper values of education: intellectual agency, learning from failure, and the joy of discovery.

Our alternative is simple: work with educators to develop AI tools that augment—not replace—learning and tools that are localized to their environment. Our work is built in collaboration with educators, by educators. As university faculty ourselves, we bring insider knowledge of how pedagogy, research, and academic governance work in practice. This allows us to design and deploy AI systems that are aligned with institutional goals, responsive to evolving policy needs, and grounded in the day-to-day realities of academic life.

Closing the Gap:
Supporting Inclusive AI Adoption

We recognize that AI adoption in education comes with resistance,. Faculty, researchers, and administrators have widely varying levels of technical fluency and comfort with new tools. While AI experts and developers move quickly at the frontier, many others are still responding to technologies like ChatGPT with limited understanding. This knowledge gap risks leaving much of the academic community behind.

To bridge this divide, we are building inclusive, user-centered environments where faculty, students, and staff can safely explore, experiment with, and adapt AI tools to their own needs. Our platform empowers end-users to control the tools—not the other way around. This creates a space for trial and error, internal testing, and iterative development, encouraging genuine engagement across skill levels and academic roles.

AI as potential equalizers (only when used well)

Beyond elite institutions and large research labs, AI has the potential to democratize access to powerful capabilities for small or under-resourced teams. We’ve seen this firsthand in our work creating child education content in Cantonese—a language underserved in mainstream educational materials and in generating Mandarin media literacy animated series for Chinese immigrants in the US. By leveraging large language models and multimodal tools, we enabled culturally specific content development with a fraction of the traditional investment. This same potential exists for any team working on niche or local knowledge domains.

It doesn't have to be 'AI slop'

Put humans back in the loop of AI-driven curiosity and creativity

'The-last-mile'

AI isn't truly empowering until it’s easy, personal, and built for you

AI Whoppers??

we’re not just stacking patties—we help you bring your homegrown tomatoes and secret sauces to the table.

No sugarcoating of AI harms

We are deeply aware of the risks associated with AI: hallucinations, environmental costs, ethical uncertainties, and inequitable access. Some of these are systemic and beyond any single organization’s control, but others can be mitigated through responsible design. We engage in AI auditing, data curation, and ongoing translational research (through LampBotics Lab) to ensure the outputs meet academic standards. Most importantly, our tools are open-source for our clients. Clients retain high degree of control over implementation, data management, and model usage.