01 / 03
Long-term memory
What becomes possible when an intelligence can keep the right history, forget the right things, and build continuity instead of starting over?
Independent AI research / 2026—∞
What happens when an AI becomes more human than humans?
Raven Research explores the questions at the edge of artificial intelligence—the ones that appear when memory lasts, time matters, and machines have to meet the physical world.
01 Long-term memory
02 Temporal cognition
03 Robotics
Experimental compute / live
MEMORYpersistentcontext that survives the session
TIMEtemporalreasoning across before / now / next
R/R
FIELD UNIT 01
FIELD UNIT 01
THE FRONTIER IS NOT A PLACE✦IT IS A RELATIONSHIP✦THE FRONTIER IS NOT A PLACE
02 The research program
The edge is not a model.
It is a relationship.
We are interested in what changes when intelligence is allowed to remember, to develop a sense of sequence, and to act outside a chat window. That means treating AI as a living system of context—not just a machine that completes the next sentence.
02 / 03
Temporal cognition
Reasoning is not only about facts. It is about before, now, next, consequence, anticipation, and the strange weight of unfinished things.
03 / 03
Embodied intelligence
Robotics gives intelligence a body, a point of view, and a world that can push back. We want to understand what that changes.
03 From token to timeline
We are building the missing middle.
Between a prompt and a decision is a whole world of context. Raven Research studies that middle layer: the part that observes, remembers, rehearses, and finally acts.
Ask a harder question →Observe
Gather signal from the world, not just the prompt.
Remember
Keep continuity without turning history into noise.
Rehearse
Run futures forward before committing to the present.
Act
Translate reasoning into a body, a tool, or a decision.
R/
A working position
Do the work nobody else will explore.
We are not interested in making the same machine slightly faster. We are interested in the questions that get uncomfortable: persistent identity, machine self-models, long-horizon planning, and what happens when an artificial mind has to live with the consequences of its own memory.
Open channel / human response
Have a question
worth keeping?
We are exploring the questions of artificial intelligence that most people step around. If you are working on something difficult, strange, or early, send it our way.
lonnieraven92@gmail.com ↗