Animation gives movement to things that have never moved and life to characters that have never existed. In this eighth essay in my 2024 series, I ask what AI might change in that craft—and what faster production cannot decide for us.
Animation has long combined artistic judgement with technical systems, from hand-drawn frames to computer-generated imagery. AI-assisted interpolation, lip-sync and motion-processing tools can reduce some repetitive work. Their usefulness depends on the task and the desired style. Smooth motion is not always expressive motion, and deliberately uneven timing can be part of a character’s personality.
Generated designs and movement studies can help a team compare ideas early. Turning them into a film is another task: characters must remain recognisable across shots, objects must behave consistently, and movement must serve the scene. Attractive isolated frames are not a substitute for direction, staging and continuity.
Animating a still image can make a storyboard easier to discuss or suggest a possible camera movement. When the source is a historical photograph, however, the added expression or gesture is imagined rather than recovered. That distinction should remain clear, especially when an animation is presented as educational or documentary material.
For independent artists and small studios, reducing some production effort could make experiments more feasible. Larger teams may also gain time, but correction and integration can consume those savings. Lower production costs do not guarantee distribution: recommendation systems and promotional budgets still shape who gets seen.
Promotional clips can be adapted to different formats, while interactive or real-time animation offers other ways to involve an audience. These are related but distinct tasks; not all require AI. A responsive character still needs an artistic framework, and a story that branches according to choices needs coherent consequences, not merely more variations.
In education and training, animation can explain a process that is difficult to observe directly. AI may assist its production, but the resulting simulation needs subject-matter review. Visual plausibility is not proof of physical or procedural accuracy, particularly where learners might act on what they see.
There are also questions about whose work makes these systems possible and how production roles change. The use of source material, voices and performances needs careful consideration, as do credit and working conditions. Stylistic variety will not arise automatically if teams keep choosing the same readily available visual patterns.
I see AI as a possible expansion of an animator’s working methods, not a definition of animation’s future. The central choices remain familiar: where to place an emphasis, how long to hold a gesture, and what makes a character feel alive. The value of the tools lies in what those choices allow an author to express.

AI & Visual Culture: 9. Adult Media, Digital Identity and Consent ↗

