The solo researcher with AI: a working map for assembling the new

⚡ The gist in 5 seconds - The point: a solo researcher with AI doesn't need a single "best methodology" — three abilities are enough: assembling a domain picture of the task, plugging in a strong specialized method at the right moment, and returning the result back into the overall picture. - Where available: the article was published on Habr on September 27, author ArgusXII, reading time — 12 minutes. - Limitation: this is the second part of a series; only a working map is presented so far, with no full methodology in the text, but the direction is set. ### 🔍 What was found ArgusXII, having compared approaches to invention, scientific discovery, and product development, concluded: a solo researcher is not a miniature research institute. Their route is cheaper to reconfigure: one can quickly step into an adjacent discipline and return. Research covering more than 65 million articles, patents, and software products showed that small and large teams occupy different niches: smaller ones more often open new directions, while large ones develop existing ones. Individual users create new products themselves when they encounter a need before the mass market does. The concept of a "facet" is introduced — an organized subject world with its own vocabulary, objects, relations, mechanisms, and proofs. Using the applied example of the ERP system "The Company's Brain: 2026," the author shows that to understand which options are viable under disruptions, one needs to assemble several facets — production planning, resilience, risks, the economics of trade-offs, system implementation — into a single working lens. The new is often found between worlds: Don Swanson described "undiscovered public knowledge" back in 1986 — published but unconnected fragments of knowledge. ### 💡 Why it matters AI dramatically lowers the cost of finding connections between disciplines, but creates a dangerous illusion: a similar word does not mean an identical object, and a beautiful explanation is not proof. The scheme gives the solo researcher a procedure: assemble the lens, find a bridge between facets, verify it, and return the result with a clear boundary of one's own verification — without mixing terminology from different fields and without losing proofs and limitations. ### 🧩 Context This is the second article in the series. The first one, "From the University to Artificial Intelligence: How the Generator of Inventions Accelerated," examined how humanity has successively amplified the generator of the new: universities, printing, scientific societies, laboratories, computing networks, and finally AI. Now the author moves to the practical side: what happens when a significant part of this power ends up in the hands of a single researcher or a small group.

🔗 Read on habr.com