The Institutional Frame
Hamming gave his 'You and Your Research' talk dozens of times across institutions from Bell Labs to the Naval Postgraduate School. His core advice held constant: work on important problems, not just busy problems. Keep 10 important problems in mind. When a new technique appears, ask whether it solves any of those 10.
But throughout the talk runs a hidden assumption: you work inside an institution. Bell Labs paid Hamming's salary. He could spend Friday afternoons thinking without producing any billable output. He had colleagues across floors and buildings whose conversations he could wander into. He had a library with physical journals. He had computing resources accessible by signing a form.
When he said 'keep your door open,' he assumed a door connected to colleagues down the hall. When he said 'invest in yourself,' he assumed employer-subsidized conference travel. When he said 'compound your knowledge,' he assumed a stable employment context in which compounding had time to work.
In 1986, when Hamming first delivered the talk at Bell Communications Research, this was nearly universal for serious researchers. In 2026, open-source has broken this assumption completely. A researcher can produce consequential work from a home directory, a public git remote, & a community of strangers who share a problem.
This lesson extends Hamming's best ideas into that frame — not to replace him, but to update the environment his advice lands in.
Translating the Open Door
Hamming on the open door: 'I notice that if you have the door partly open, you get less work done, but you hear what is important. The great scientists tend to have open doors — not all the time, but some of the time.'
He meant it literally. A colleague walking past would mention a problem. Hamming would catch a fragment of conversation about a new technique. These collisions happened in physical space, at lunch, in hallways, at the coffee machine.
The 10-Problems Technique Outside an Institution
Hamming's 10-problems technique: maintain a list of the most important unsolved problems in your field. When a new method, tool, or result appears, ask whether it addresses any of the 10. This focuses attention & creates what look like lucky breaks: a new technique appears in a seminar and within minutes Hamming sees which problem it solves.
In open-source, the problems live in public places: issue trackers, security databases (CVEs, CWEs), conference talks, Stack Overflow threads that never get resolved, library changelogs that warn 'this is a known limitation.' A MOAD pipeline applies Hamming's technique systematically: scan for CWE-407 across ecosystems, match confirmed findings to upstream projects, file issues, submit patches.
The pipeline requires no salary. It requires: a problem list (MOADs), a scanning method (grep patterns, static analyzers), & upstream access (git, mailing lists, GitHub, GitLab). Anyone with a terminal & an internet connection can run it.
Hamming's compound knowledge: work on the most important problems & every new technique you learn potentially solves one. Open-source compounds differently: every patch merged upstream propagates to all downstream forks automatically. The fix spreads without additional effort from the original researcher. A patch submitted to Python's email library in 2020 reached every Python installation by 2021.
The institution provided: salary continuity, computing resources, library access, colleague network, prestige as validation. In 2026, most of these arrive free at the network edge: cloud compute, open journal archives, GitHub, Stack Overflow, academic Twitter. The remaining scarcity is attention & judgment, not access.
Apply the 10-Problems Technique
Hamming's question, directed at your domain:
What Institutions Provide, What They Do Not
Hamming: 'It takes courage to work on important problems. Most people do not work on important problems. If you do not work on important problems, it is not likely that you will do important work.'
Institutional backing provides a form of courage: tenure removes the threat of dismissal. Salary continuity removes income anxiety. Peer recognition validates that the problem is worth tackling. The institution absorbs the cost of failed attempts.
Working outside an institution removes each of these supports. A patch you submit may get ignored by maintainers who have other priorities. A disclosure you make may get dismissed as not a real vulnerability. A project you maintain for years may never attract contributors. No one guarantees your effort leads anywhere.
But open-source also removes one specific fear that institutions create: you cannot be fired from a project you maintain. No manager can redirect you to a less important problem because a client asked. No performance review penalizes you for working on something that took five years to bear fruit. A public-domain patch does not need permission to exist. It needs only to be correct.
Permacomputer principle: ship the patch as public domain. The patch does not need credit to survive. It does not need institutional affiliation to be adopted. It needs to be correct & reachable. If an upstream maintainer ignores it, fork the repo & ship the fix in the fork. The correctness persists regardless of the reception.
The Open-Source Closed Door
Hamming observed that scientists who close their office door get more done in the short term but fall behind long term because they stop hearing what matters.