See, seek, shift. Notice a fracture in reality and write it down without explaining it away; run one small test this week rather than convening a discussion; then change something and record what the change was for. The loop sharpens the more it is run, which is the part that cannot be taught in advance.
The answer, in one line
Three steps. See: notice an anomaly and write down what deviated without explaining it away. Seek: run one small test by the end of the week rather than holding a meeting about it.
Definition#
The Curiosity Loop: a three-step discipline for practising curiosity. See surfaces an anomaly and records it without adornment. Seek runs one small test. Shift changes what you do and ties the change to what was learned.
See#
The discipline of surfacing a fracture in reality. An anomaly, a friction, an outlier that does not fit the story. A pattern breaks, a number looks wrong, a silence feels heavy. The work is awareness without adornment: write down what deviated and resist the urge to explain it away, because the explanation arrives faster than the evidence and usually closes the question.
The speed at which you move from noticing to asking why is what sharpens the loop. Most anomalies are resolved into the existing story within seconds, by the person best placed to have investigated them.
Seek#
The habit of running one small test by Friday. Not a meeting about possibilities. A probe: a direct question, a quick cut of the data, a conversation with the person closest to the friction, a micro-change in the field. The evidence of Seek is a log of tests rather than a pile of slides, and the distinction is not rhetorical. One is a record of things that happened and the other is a record of things that were said.
Shift#
The courage to change what you do. Record the learning and tie it explicitly to the action: we learnt X, so we will Y. Change a threshold, kill a feature, rewrite a question, reorder a process. The proof is a question that became a test, not a conversation that became a document.
The sentence is also a test of the original question. If you cannot complete we learnt X so we will Y within an hour, the question was too vague, and the fix is to rewrite the question rather than to run a bigger test.
The 48-hour rhythm#
The smallest version that works. Begin with one question that could influence a decision this week. Put fifteen minutes in the calendar to explore it before Friday. Write the line you are trying to answer at the top of your working document and keep it where you can see it. No slides, no meeting.
One question carries more than a dozen slides, and the book offers a default when nothing obvious presents itself: what are we treating as fact that is still a guess? It is a useful anchor for the loop when urgency starts to outrun sense.
Why this matters more with machines in the room#
Tools accelerate whatever they are pointed at, so the question that sets the direction is doing more work than it used to. A system that answers instantly also removes the interval in which a person would have noticed that the question was the wrong one. See is the step under most pressure, because the fracture is now covered over by a fluent answer before anybody has recorded that it was there. The related argument is at AI and critical thinking.
What this has not been shown to do#
The loop has not been tested. No study has compared people or teams who run it against those who do not, and the page it sits under is explicit that procedures of this kind do not train the underlying disposition. What the loop does is convert a disposition into a record, so that its presence or absence becomes visible. That is worth having and it is not the same as producing curiosity in somebody who has none.
Source#
Rahim Hirji, SuperSkills: The Seven Human Skills for the Age of AI, Kogan Page, 2026, chapter one. Coined there. The stories and the worked cases are in the book.
Related SuperSkills research#
Essay · SS-2026-337 · How I use AI here
Hirji, R. (2026). The Curiosity Loop. The SuperSkills evidence base, SS-2026-337. https://thesuperskills.com/research/the-curiosity-loop. Last reviewed 26 September 2026.
An evidence review by Rahim Hirji, not peer-reviewed research. For a material claim, cite the underlying study as well; every study here carries its own permanent link.
How citations and IDs work