You are planning an exhibition and need someone to photograph the installation. A fortnight ago an AI assistant gave you a long list of names to look into. Today a colleague sends a shortlist, and one name immediately feels like a safe choice. You cannot remember a portfolio, a conversation or a finished commission. The name just rings a bell. It is tempting to treat that feeling as a recommendation. What if all it tells you is that you read the name in the first list?
This is a problem of attribution. You recognise the name, but you cannot say where you encountered it or what you learnt about the person. For a creator or business owner who uses AI to begin research, the distinction matters. A name that appeared in a brainstorm and a name referred by someone who has worked with the person may both look familiar at the next meeting. They have earned different kinds of attention.
What a name can borrow from memory
In a 1989 study published in the Journal of Personality and Social Psychology, Larry Jacoby, Colleen Kelley, Judith Brown and Jennifer Jasechko examined what happens when people are asked to judge whether names are famous. Participants encountered non-famous names earlier in the experiment. After a delay of about a day, previously seen non-famous names were more likely to be judged famous than comparable names people had not seen earlier. The same false-fame effect did not occur in the immediate test described by the authors.
The result illustrates how past exposure can affect a judgement when its source is not readily recalled. It does not mean that every familiar name will seem famous, or that familiarity always outweighs evidence. The study tested judgements about fame under controlled conditions, not hiring, commissioning or the use of chatbots. Applying the distinction to an AI-assisted shortlist is a practical inference. The feeling can prompt a useful question: where did I see this name, and what did that encounter establish?
Separate discovery from endorsement
Imagine three entries in the exhibition shortlist. One came from an AI brainstorming session. A second came from a colleague who recognised the name from an industry directory. A third came from a colleague who commissioned similar work and can describe what the photographer delivered. All three may deserve a look. The first two routes, on their own, tell you little about whether anyone has seen the person do the work you need.
Notice that even the third route does not settle the question. The colleague might have commissioned indoor portraits, while you need photographs of reflective work in a difficult space. A referral carries more information when you know what it refers to. Write down the exact reason to investigate a name: evidence of a comparable project, a useful style match, or a recommendation tied to a specific job. Keep availability, fit and practical constraints as separate checks.
The same principle applies if you ask several AI tools for suggestions. Two answers may repeat a name without providing independent evidence about it. Ask what each source actually establishes. If both routes lead to the same portfolio, count that portfolio as one source of evidence. If a tool cannot show where a detail came from, record the detail as unverified or leave it out of the decision.
Give each name a short trail
You do not need to preserve every conversation. A compact record is enough if someone else can tell how a person reached the shortlist and why they stayed there. A row in a shared document can carry the candidate’s name, the route by which you found them, a link or reference you checked, and the reason for the next step. Write “found in brainstorm; portfolio not reviewed” when that is all you know.
- Preserve the first known route. If a name first appeared in a generated list, keep that fact even after you find better evidence. The later evidence can earn a different status.
- Record what the evidence supports. “Portfolio shows two installations in comparable spaces” tells a colleague more than “looks reputable”.
- Make the next decision small. If the evidence is incomplete, decide whether to review a relevant example, ask a colleague for specifics, or request a conversation.
This fits alongside our advice on keeping a record when an AI assistant remembers the project. The shortlist is one place where a few words about origin and purpose can help you reconstruct the decision later. If you cannot retrieve the first encounter, leave the origin marked unknown. There is no need to invent a backstory for the bell that rang.
Pay attention to the language when the shortlist moves from a private note to a shared document. A heading such as “Recommended photographers” can turn research leads into apparent endorsements before anyone has reviewed them. At the first meeting, divide the list into people to investigate and people whose relevant work someone has checked. Name who did the check and what they looked at. If the shortlist comes back a month later, a colleague can resume the work without mistaking the old heading for a decision.
Check what survives without the name
For a final comparison, try reading the evidence notes without looking at the names first. This is an editorial habit, not a test validated by the fame study. It can reveal that one row has a strong case and another has little beyond a confident adjective. Restore the names before making the decision, and check that the work and the person are correctly matched. A familiar name may still be an excellent choice. Let the available evidence explain why.
Source
- Jacoby, Kelley, Brown and Jasechko: Becoming Famous Overnight, Journal of Personality and Social Psychology (1989).
Discover more from Marychuks.com AI, Psychology, Business & CreativeVerse
Subscribe to get the latest posts sent to your email.