The research and development agenda is public
All seven research pillars are now published on one page, each tied to the public artifacts behind it, alongside an honest map of how far the self-improvement work has gone.
The full ELI Labz research and development agenda is now on the Research page. Seven pillars sit between today's capable models and reliable, collaborative Intelligent Machine Systems: real-time reasoning, situated reasoning, multimodal reasoning, persistent memory, human action tokens, agent interoperability, and governed autonomy.
Each pillar states the research question it answers, why it matters to an operator, and where the work stands, with links to the public artifacts: case studies, the live Third Eye workspace, the review design documentation, Cursor & Keyboard Agent, and Labor Economics.
How the claims are bounded
The page closes with the Recursive Self-Improvement Claim Ladder, a plain statement of how far the self-improvement work has actually gone. Levels one through three are standing capability within bounds. Levels four through six are partial, each with a person approving the step that matters. Level seven is claimed by no one, including us.
The rule for the agenda is the rule for this blog: if there is no public artifact behind a statement, the statement does not get made.
Put this page in front of the right person
This page travels by forwarding, not by advertising. If it describes someone's week, send it to them directly.