The next important breakthrough may already have all its ingredients sitting in published research. The missing piece could simply be a connection that nobody has noticed — the unknown between the known.
That possibility sits at the heart of Intercognix, a concept developed by technology professional and author Jay Gontu to describe candidate insights found through previously unrecognized relationships among independently known things.
Beyond the Boundaries of Individual Disciplines
Modern civilization has accumulated an extraordinary amount of knowledge. Universities, laboratories, libraries, journals, databases, and researchers continuously add new findings to that collection. Yet knowledge remains organized largely around individual disciplines — like a library built one hall at a time, with no corridors connecting them.
That organization serves an important purpose. Specialization allows researchers to develop deep expertise. It also means observations made in one field may remain unfamiliar to people working somewhere completely different. A historian rarely reads an obscure pharmacology paper; a physicist rarely encounters an observation buried in an unrelated field.
Imagine two published findings sitting in separate corners of the intellectual world. Each is understood within its own discipline. When considered together, however, they might suggest a question, mechanism, or hypothesis that hasn’t appeared in either literature.
This isn’t purely theoretical.
Information scientist Don Swanson demonstrated a famous example during the 1980s. He examined separate medical literatures concerning dietary fish oil and Raynaud’s phenomenon. Relationships between findings in those independent bodies of research suggested a potential therapeutic connection that had gone unrecognized. Clinical investigation later provided support for the connection.
Swanson described the broader phenomenon as “undiscovered public knowledge.” Information could therefore be publicly available while an insight produced by connecting that information remained undiscovered — hidden not for lack of facts, but for lack of a corridor between them.
His work required painstaking manual investigation. Today’s artificial intelligence introduces a very different scale. AI systems can transform large collections of text into mathematical representations called embeddings. Concepts originating in different subjects can consequently be compared within a shared computational environment. Gontu believes this capability creates an opportunity to search systematically for meaningful relationships across fields that have historically developed apart.
Why Intercognix Focuses on Candidates
Gontu coined Intercognix using “inter” for between, “cogni” for knowing, and “X” for the unknown.
He defines an Intercognix as a candidate insight discoverable through previously unrecognized relationships among independently known things: two known things, and a bridge between them that implies something else entirely.
That definition deliberately stops at “candidate.”
An unexpected computational relationship isn’t automatically knowledge. AI can identify countless associations, including ones produced by coincidence, superficial similarities, incomplete information, or characteristics of the model itself. Genuine discovery demands considerably more.
A candidate must face prior-art research, examination by people with relevant expertise, and empirical validation. Researchers should also be able to identify conditions under which the idea would fail.
AI as a Guide Toward Unasked Questions
People generally approach artificial intelligence with a prompt. They already possess some idea of what they want explained, summarized, calculated, compared, or investigated.
Gontu believes that structure reveals an important limitation.
“Every question we ask a machine is limited by what we can already imagine,” he says. “We have built extraordinary systems for answering questions. The harder frontier is discovering the questions nobody has thought to ask.”
Intercognix shifts attention toward that earlier stage of discovery. Instead of waiting for a researcher to independently recognize an unusual interdisciplinary relationship, computational systems could help surface promising corridors between existing bodies of knowledge.
Human judgment then becomes crucial.
A researcher doesn’t need to become an expert in two unrelated disciplines. A candidate could bring together two specialists, each deeply familiar with one side of the proposed relationship. They could examine the evidence, challenge assumptions, search for earlier work, and determine whether testing is justified.
“The science on each side is often already done,” Gontu explains. “What remains is the bridge, and building a bridge is a smaller ask than building both shores.”
This collaborative model could give students and early-career researchers particularly interesting opportunities. AI can help reveal an unusual starting point, while curiosity and domain expertise determine where the investigation goes next.
About Jay Gontu
Jay Gontu is a technology professional based in the Chicago area and the author of Close Your Eyes to See, a work of philosophical nonfiction. In 2026, he coined Intercognix — the unknown between the known — to give a distinct identity to candidate insights emerging from previously unrecognized connections among independently known things.
Where the Next Search Could Begin
Intercognix proposes an unusual kind of frontier. It doesn’t require traveling farther into physical space. Its raw material already exists throughout humanity’s accumulated knowledge.
As new findings enter that collection, additional connections become possible. The territory therefore expands with every advancement — new corridors opening in a library that keeps growing.
AI can help expose those potential pathways, while researchers provide the skepticism, expertise, experimentation, and evidence required to determine which ones matter. For Jay Gontu, that growing space between what humanity already knows may contain questions worth asking and discoveries still waiting to be recognized.
Knowledge grows when shared. Please Pass it on!
Discover more from Paxjones
Subscribe to get the latest posts sent to your email.