AI as an instrument, not an autopilot
AI as a creative tool is the idea that artificial intelligence should behave like an instrument in the studio or on set, extending human imagination, speed and reach without replacing the artist’s judgment, taste or authorship at any point in the process.
That distinction—AI as instrument versus AI as autopilot—is the real battle line in entertainment. Empire of the Sun sit squarely on the instrument side. Luke Steele describes AI not as a destroyer of creativity but as a “really smart guitar pedal,” a studio tool that bends ideas, throws them somewhere unexpected and loosens up stuck songs. Instead of prompting a model for a finished track, the duo have built custom AI studio tools that sit inside their existing workflow.
This is the most defensible version of AI music production: artists keep their hands on the desk while the machine becomes one more sound-shaping device, no different in spirit from a sampler or synth—only far more responsive.
Empire of the Sun’s ‘guitar pedal’ approach to AI music
Empire of the Sun have always traded in future-facing pop—ever since Walking on a Dream landed in 2008, they looked like they arrived ahead of schedule. Their AI stance matches that legacy: they treat AI as a strange new studio instrument rather than an automated hit factory.
Their custom AI music production pipeline includes a tool called Song Mountain, a lyric generator fed with ancient quotes, scripture and overheard street conversations. The point is not to outsource writing but to “stir the pot,” throwing together images that might crack open a fresh direction. They also use AI to blend wildly different musical references and drag abandoned 40‑second loops into more useful shapes.
Empire of the Sun’s model should be non‑negotiable for AI studio tools: human intent in, machine chaos in the middle, human selection out. Anything more automated than that risks turning artists into prompt engineers and flattening the weirdness that makes songs worth hearing.
Hollywood’s quiet AI takeover behind public scepticism
While musicians experiment in the studio, Hollywood is in a louder, more anxious phase. Generative AI was central to the 2023 strikes, as writers and actors protested its use for scripts and digital replicas of performers. Directors like Kane Parsons say they would “snap my fingers” to make AI disappear, and Christopher Nolan has called it a Trojan Horse.
Yet behind that scepticism, AI filmmaking is spreading through the entire pipeline, from script development and production planning to shot generation and editing. Netflix bought AI production firm InterPositive, led by Ben Affleck, for almost USD 600m (approx. RM2.76b) in March, and says AI has already touched about 300 titles this year. Another studio player backed by a major tech lab received about USD 75m (approx. RM345m) for AI tool development.
This is not neutral infrastructure; it is a strategic bet that AI will cut costs and speed up production, even as the creative class fears being squeezed out of the decision‑making loop.
Human‑centered AI versus automation in film and streaming
The most revealing moves are happening inside legacy giants. One entertainment conglomerate has explicitly named Industrial Light & Magic and Pixar as teams it is investing in to develop AI and machine learning, promising shareholders that the work will remain “human-centered, artist-driven, and creator-led”. In the same filing, the company stresses that AI is “not simply about efficiency” but about enhancing creative processes.
That language sounds reassuring until you read the fine print: the firm is “highly focused on reducing costs,” evaluating reductions in labor and other expenses, and has already recorded USD 88m (approx. RM404m) in severance for the quarter and USD 180m (approx. RM827m) for nine months. It has rolled out a J.A.R.V.I.S. AI tool to more than 2,000 Imagineers, putting over 70 years of institutional knowledge at their fingertips.
This is the fault line. Human-centered AI means artists steer the machine. Automation means executives use AI as cover to shrink crews. The industry is doing both at once, and workers know it.
Audience impact: faster content, more noise, and the fight for quality
For audiences, the most visible AI creative tool today is speed. One AI‑generated ten‑second scene of an invented actor’s arrest took four minutes to create; shooting it on location would have taken half a day. Netflix’s documentary series The American Experiment includes 17 minutes of AI‑enhanced battle footage that its co‑chief says were produced “twice as fast and at half the cost” of other options.
Studios are also using AI to chop catalog shows into smartphone‑friendly clips as they try to compete with micro‑dramas, YouTube and TikTok. One startup helps give old series “new life” by slicing them into scenes for social media distribution. Advocates argue this helps established studios win back attention; critics see a coming flood of low‑effort AI sludge, pointing to titillating micro‑dramas as a warning.
The future of AI filmmaking and AI music production will be decided not by whether AI is used—it already is—but by whether artists can keep human taste, consent, controls and compensation at the center. If Empire of the Sun’s guitar‑pedal model wins, audiences get stranger, richer work. If full automation wins, expect more content and less art.
