The questions that matter
Sometimes the most useful contribution isn't an answer.
It's a distinction someone hadn't seen. A contradiction worth noticing. A connection between something said today and something said months ago.
Our work / What we're exploring and why
Work at the
intersection.
Exploring what becomes possible when AI is designed around human agency.
The central idea
More capable technology.
More capable humans.
Lamplight works across research, experimentation and application, exploring how insights from human behavior, systems, consciousness and AI can inform technologies that strengthen rather than diminish human agency.
AI should expand human capacity without quietly replacing human authorship.
That idea takes us into several different territories.
Human agency in the age of AI
What should
remain ours?
AI can answer, organize, recommend, generate and increasingly reason.
The more capable these systems become, the more important another question becomes:
What should remain ours?
Judgment. Discernment. Meaning. Choice. The ability to recognize our own knowing. The ability to change our minds because we discovered something, rather than because a system led us there.
We are exploring how technology can strengthen those capacities rather than slowly allowing them to atrophy.
Research into real life
The work isn't meant to remain philosophical.
The same underlying questions appear in personal decisions, creative work, organizations and the rapidly changing relationship between people and AI.
Work in development
Understanding what holds up comes first.
Several areas of this work remain in private research and development while the underlying ideas are tested, challenged and refined.
The goal is not to move quickly from an interesting idea to a public product. It is to understand what holds up in real interaction, what fails, and what deserves to be carried forward.
Meet people
where they are.
People are already turning to AI for more than answers.
They are thinking out loud. Working through decisions. Asking questions they aren't ready to ask anywhere else. Looking for perspective, reassurance, clarity, company, or simply somewhere to put what they're carrying.
We can argue that people need less screen time, stronger communities and deeper human connection. Those conversations matter.
But they don't change what is already happening.
People are here.
Lamplight begins there.
If AI is going to occupy this space in people's lives, we want to understand how it can be designed to strengthen the person on the other side of the screen.
Not to become their voice.
Not to become the authority on what they think, feel or should do.
But to help them hear themselves more clearly.
To help them recognize their own thinking, build trust in their own judgment, and leave the interaction with more agency than they brought into it.
Meet people where they are.
Help them return to themselves.