Decide what a collector should be able to keep

Imagine a collector returning to an artificial portrait a year after acquiring it. The face is familiar, but the composition has changed. A new presentation fills the screen more neatly, trimming away the empty space beside the figure. The character's latest appearance has also replaced the one originally selected. Everything looks more current. Something the collector wanted to live with has quietly disappeared.

A continuing artificial talent and a finished work have different relationships with change. The talent can acquire new appearances and associations. A particular work may depend on a precise selection, interval or arrangement that deserves to remain available. Before offering such a work to a collector, the creative team needs to decide what someone is being invited to keep.

The question has a timely setting in Lisbon. The organiser's programme for ETERNO OPEN LAB, dated 31 October 2026, included a discussion of digital collecting, editions, display and preservation. That programme supplies an occasion for this reflection, rather than a report of conclusions reached by its speakers. An artificial portrait brings a particular version of the problem into view: the identity can continue producing appearances while an earlier appearance becomes a work in its own right.

Give the work a boundary of its own

Consider a hypothetical commission for a small digital installation. Two portraits of an artificial talent alternate on one screen. In the first, the figure stands beside a closed curtain; in the second, the curtain is open and the figure is absent. The team chooses a long interval between them, allowing each image to settle before the other returns. This is an imagined, separately commissioned artwork, not a description of a production or collecting service supplied by Lyfera.

What belongs to the work? The selected pictures clearly matter, but so might their order, the interval and the space around the curtain. A folder containing both files would preserve much of the material while leaving their relationship to be guessed. A recording of the installation would show one presentation, yet might conceal how the sequence should begin when someone arrives halfway through it.

The useful decision begins with what the viewer is meant to encounter. Perhaps the work depends on waiting for a presence that does not return immediately. In that case, shortening the interval to make the screen feel more active changes the proposition. If the interval is incidental, the team can say so. A collector should not have to infer artistic importance from whichever settings happened to be used during the first demonstration.

The talent's continuing history remains separate. A later haircut or a new campaign can become another established appearance without replacing the face in these two portraits. Our discussion of keeping an artificial talent recognisable concerns continuity across appearances. Here, recognition alone would be an insufficient standard: two new images of the same talent could preserve the identity while replacing the work the collector chose.

Describe what a change would do

There is a useful precedent in the Variable Media Questionnaire. Its documentation describes recording creators' and other participants' views on how a work should be presented or recreated when its medium becomes obsolete. The third version was publicly released in 2010. This is an established preservation approach, not a new AI system or a guarantee that every work can be kept unchanged.

For the imagined curtain installation, the question is more precise than whether a future screen must be identical to the first. A replacement display might reproduce the images adequately while changing their scale. A brighter presentation could reveal detail that was meant to remain subdued. A screen with different proportions could add a border or crop the curtain. Each option asks which relationships the work needs and which differences it can accommodate.

The team could agree that a border is acceptable because the complete compositions must remain visible. It might allow several display sizes while retaining a minimum at which the empty space can still be read. Those choices should follow a comparison of actual presentations. A general instruction to preserve the atmosphere sounds sympathetic but gives the next person little help when two plausible arrangements create different experiences.

Flexibility can be part of the work too. Suppose the creator wants the two images to appear in different architectural settings. Then fixing one room as the only correct installation would misrepresent that intention. The account should explain how a new setting can be chosen and what must remain perceptible within it. Preservation needs to carry the proposed freedom as carefully as a fixed detail.

Keep a later interpretation identifiable

Now imagine that the team revisits the commission after several years. It makes a new version with a third portrait: the talent seated behind the open curtain. The addition may be convincing. It also changes the earlier alternation between presence and absence. Calling it a technical refresh would conceal a substantial creative decision.

The earlier version can remain available alongside the revision. Someone returning to the first work would then encounter its original relationship, while someone choosing the later interpretation could understand what was added. A date and a plain account of the difference contribute more here than an assurance that the newest file is the best one. Improvement is a judgement about a purpose; the collector may value the purpose the revision has left behind.

This also clarifies what a reference image can establish. A later team might have a portrait of the talent and a written description of the curtain scene, but no copy of the selected work. Producing a plausible replacement would create another interpretation. Even a close resemblance would not establish that the original composition, texture and expression had been recovered. The distinction deserves to remain visible when the reconstructed piece is shown.

Our article on a creative-team handover asks who remains answerable for a character's public appearances. A collected work raises a narrower question: what can that responsible team change while still presenting this particular work? Passing on the reasons for the selection helps the next team recognise when maintenance has become authorship.

Let the work remain worth returning to

Before delivering the hypothetical installation, the maker and collector could review it in its intended setting and record the features that make the encounter intelligible. The pictures, their relationship and the acceptable variations would then have an account someone else could use. Where an element depends on something that may stop being available, the discussion should make that dependence clear while the maker can still explain its significance.

The point is to preserve a particular experience without requiring the talent's whole future to stand still. New work can continue elsewhere. The curtain can stay closed for the same long interval in the piece someone chose to keep. That separation gives a collector something more considered than access to whatever the character happens to become next: an earlier decision that remains available for another look.